S.P. Mulders
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
11 records found
1
Swarmster
Preliminary Design and Performance Assessment of an Unmoored Self-Propelled Floating Wind Turbine for Hydrogen Production
This thesis assesses the feasibility, behaviour, and operational viability of the Swarmster: an unmoored, self-propelled floating wind platform that generates power on the open ocean and stores it as hydrogen, drifting freely rather than being anchored or cabled to shore. Removing the mooring turns a fixed turbine into a mobile system whose motion continuously alters its own inflow, and whether such a concept can work has not previously been established. The concept is examined in three stages, corresponding to whether the platform can exist at a sensible size, how it behaves once set in motion, and whether it can complete a useful mission.
A preliminary sizing study shows that a minimum feasible size exists and is set by stability rather than storage. The requirement to limit pitch under rated thrust grows faster with installed power than the hydrogen storage requirement, so operational stability is the binding constraint at any rating of practical interest; the platform is fundamentally sized to stay upright under its own turbine loads. The analysis imposes a lower bound on size but no upper bound, and the resulting hull, built around a mainstream offshore turbine, falls within the size class of existing vessels, placing the concept inside conventional shipbuilding practice.
A steady-state performance model, coupling actuator-disk aerodynamics to a hydrodynamic force balance, characterises the platform’s behaviour. The defining feature is a feedback between thrust, drift speed, and apparent wind that collapses power near the downwind heading. Hydrokinetic turbines mounted beneath the hull recover this loss, since they are most effective where the wind turbines are weakest; scheduled by heading, they flatten the strongly directional power profile to a near-uniform capability across the forward headings. Hull hydrodynamic quality is found to govern the angular width of the operating window rather than the peak power, which is set by the turbines.
Finally, a time-domain case study driven by reanalysis wind over two contrasting thirty-day windows tests the platform as a mission. A viable cycle is achievable because a strategy that generates until the return threshold is reached allows the platform to return to a fixed offload point and deliver its stored hydrogen under both wind regimes without permanent mooring. Recovery is nonetheless costly. Requiring return roughly halves the delivered hydrogen relative to a strategy that maximises production alone. This trade-off arises from generating by drift rather than from a limitation of the particular controller.
The concept is therefore shown to be feasible, well understood, and operationally viable, with the central open question no longer whether it can work but whether, against that cost, it is preferable to moored alternatives. ...
A preliminary sizing study shows that a minimum feasible size exists and is set by stability rather than storage. The requirement to limit pitch under rated thrust grows faster with installed power than the hydrogen storage requirement, so operational stability is the binding constraint at any rating of practical interest; the platform is fundamentally sized to stay upright under its own turbine loads. The analysis imposes a lower bound on size but no upper bound, and the resulting hull, built around a mainstream offshore turbine, falls within the size class of existing vessels, placing the concept inside conventional shipbuilding practice.
A steady-state performance model, coupling actuator-disk aerodynamics to a hydrodynamic force balance, characterises the platform’s behaviour. The defining feature is a feedback between thrust, drift speed, and apparent wind that collapses power near the downwind heading. Hydrokinetic turbines mounted beneath the hull recover this loss, since they are most effective where the wind turbines are weakest; scheduled by heading, they flatten the strongly directional power profile to a near-uniform capability across the forward headings. Hull hydrodynamic quality is found to govern the angular width of the operating window rather than the peak power, which is set by the turbines.
Finally, a time-domain case study driven by reanalysis wind over two contrasting thirty-day windows tests the platform as a mission. A viable cycle is achievable because a strategy that generates until the return threshold is reached allows the platform to return to a fixed offload point and deliver its stored hydrogen under both wind regimes without permanent mooring. Recovery is nonetheless costly. Requiring return roughly halves the delivered hydrogen relative to a strategy that maximises production alone. This trade-off arises from generating by drift rather than from a limitation of the particular controller.
The concept is therefore shown to be feasible, well understood, and operationally viable, with the central open question no longer whether it can work but whether, against that cost, it is preferable to moored alternatives. ...
This thesis assesses the feasibility, behaviour, and operational viability of the Swarmster: an unmoored, self-propelled floating wind platform that generates power on the open ocean and stores it as hydrogen, drifting freely rather than being anchored or cabled to shore. Removing the mooring turns a fixed turbine into a mobile system whose motion continuously alters its own inflow, and whether such a concept can work has not previously been established. The concept is examined in three stages, corresponding to whether the platform can exist at a sensible size, how it behaves once set in motion, and whether it can complete a useful mission.
A preliminary sizing study shows that a minimum feasible size exists and is set by stability rather than storage. The requirement to limit pitch under rated thrust grows faster with installed power than the hydrogen storage requirement, so operational stability is the binding constraint at any rating of practical interest; the platform is fundamentally sized to stay upright under its own turbine loads. The analysis imposes a lower bound on size but no upper bound, and the resulting hull, built around a mainstream offshore turbine, falls within the size class of existing vessels, placing the concept inside conventional shipbuilding practice.
A steady-state performance model, coupling actuator-disk aerodynamics to a hydrodynamic force balance, characterises the platform’s behaviour. The defining feature is a feedback between thrust, drift speed, and apparent wind that collapses power near the downwind heading. Hydrokinetic turbines mounted beneath the hull recover this loss, since they are most effective where the wind turbines are weakest; scheduled by heading, they flatten the strongly directional power profile to a near-uniform capability across the forward headings. Hull hydrodynamic quality is found to govern the angular width of the operating window rather than the peak power, which is set by the turbines.
Finally, a time-domain case study driven by reanalysis wind over two contrasting thirty-day windows tests the platform as a mission. A viable cycle is achievable because a strategy that generates until the return threshold is reached allows the platform to return to a fixed offload point and deliver its stored hydrogen under both wind regimes without permanent mooring. Recovery is nonetheless costly. Requiring return roughly halves the delivered hydrogen relative to a strategy that maximises production alone. This trade-off arises from generating by drift rather than from a limitation of the particular controller.
The concept is therefore shown to be feasible, well understood, and operationally viable, with the central open question no longer whether it can work but whether, against that cost, it is preferable to moored alternatives.
A preliminary sizing study shows that a minimum feasible size exists and is set by stability rather than storage. The requirement to limit pitch under rated thrust grows faster with installed power than the hydrogen storage requirement, so operational stability is the binding constraint at any rating of practical interest; the platform is fundamentally sized to stay upright under its own turbine loads. The analysis imposes a lower bound on size but no upper bound, and the resulting hull, built around a mainstream offshore turbine, falls within the size class of existing vessels, placing the concept inside conventional shipbuilding practice.
A steady-state performance model, coupling actuator-disk aerodynamics to a hydrodynamic force balance, characterises the platform’s behaviour. The defining feature is a feedback between thrust, drift speed, and apparent wind that collapses power near the downwind heading. Hydrokinetic turbines mounted beneath the hull recover this loss, since they are most effective where the wind turbines are weakest; scheduled by heading, they flatten the strongly directional power profile to a near-uniform capability across the forward headings. Hull hydrodynamic quality is found to govern the angular width of the operating window rather than the peak power, which is set by the turbines.
Finally, a time-domain case study driven by reanalysis wind over two contrasting thirty-day windows tests the platform as a mission. A viable cycle is achievable because a strategy that generates until the return threshold is reached allows the platform to return to a fixed offload point and deliver its stored hydrogen under both wind regimes without permanent mooring. Recovery is nonetheless costly. Requiring return roughly halves the delivered hydrogen relative to a strategy that maximises production alone. This trade-off arises from generating by drift rather than from a limitation of the particular controller.
The concept is therefore shown to be feasible, well understood, and operationally viable, with the central open question no longer whether it can work but whether, against that cost, it is preferable to moored alternatives.
Optimal Control Strategy for the DOT500PRO Wind Turbine System
Using Data-Driven Wind Prediction and Comparing Control Strategies to Maximise Revenue
This thesis aims to develop an optimal control strategy for the DOT 500kW Pilot Reverse Osmosis (DOT500PRO) turbine system. The system integrates a 500 kW wind turbine with a reverse osmosis (RO) module to produce freshwater. The primary goal is to maximise revenue generation by optimising the turbine’s state transitions based on wind predictions.
The thesis begins with an analysis of the DOT500PRO and its state machine, identifying operational states, transitions, and constraints. A Markov model is used to model and predict wind speeds, which fits nicely with the Markov Decision Process (MDP) framework. The problem is formulated as an MDP, and multiple control strategies, including Threshold Control, Model Predictive Control (MPC), Stochastic Dynamic Programming (SDP), and Approximate Dynamic Programming (ADP), are evaluated.
MPC is found to be computationally intensive, making it less feasible for real-time control. SDP shows promising results, but is limited by the curse of dimensionality, restricting the use to higher order models. ADP, which approximates SDP solutions, can offer a potential controller for higher order models but requires further tuning and optimisation.
Simulations are conducted to compare the performance of these control strategies in several scenarios. While SDP demonstrates slight improvements over threshold control on the training dataset, its performance on different wind patterns is less consistent. The study concludes that while proactive control strategies such as SDP and ADP can offer improvements over reactive methods, their performance is dependent on the accuracy of wind predictions and the specific operational conditions.
Future work suggestions include refining the turbine and wind models, exploring adaptive control methods, and conducting real-life experiments to validate the control strategies, which are crucial for practical implementation and optimisation. ...
The thesis begins with an analysis of the DOT500PRO and its state machine, identifying operational states, transitions, and constraints. A Markov model is used to model and predict wind speeds, which fits nicely with the Markov Decision Process (MDP) framework. The problem is formulated as an MDP, and multiple control strategies, including Threshold Control, Model Predictive Control (MPC), Stochastic Dynamic Programming (SDP), and Approximate Dynamic Programming (ADP), are evaluated.
MPC is found to be computationally intensive, making it less feasible for real-time control. SDP shows promising results, but is limited by the curse of dimensionality, restricting the use to higher order models. ADP, which approximates SDP solutions, can offer a potential controller for higher order models but requires further tuning and optimisation.
Simulations are conducted to compare the performance of these control strategies in several scenarios. While SDP demonstrates slight improvements over threshold control on the training dataset, its performance on different wind patterns is less consistent. The study concludes that while proactive control strategies such as SDP and ADP can offer improvements over reactive methods, their performance is dependent on the accuracy of wind predictions and the specific operational conditions.
Future work suggestions include refining the turbine and wind models, exploring adaptive control methods, and conducting real-life experiments to validate the control strategies, which are crucial for practical implementation and optimisation. ...
This thesis aims to develop an optimal control strategy for the DOT 500kW Pilot Reverse Osmosis (DOT500PRO) turbine system. The system integrates a 500 kW wind turbine with a reverse osmosis (RO) module to produce freshwater. The primary goal is to maximise revenue generation by optimising the turbine’s state transitions based on wind predictions.
The thesis begins with an analysis of the DOT500PRO and its state machine, identifying operational states, transitions, and constraints. A Markov model is used to model and predict wind speeds, which fits nicely with the Markov Decision Process (MDP) framework. The problem is formulated as an MDP, and multiple control strategies, including Threshold Control, Model Predictive Control (MPC), Stochastic Dynamic Programming (SDP), and Approximate Dynamic Programming (ADP), are evaluated.
MPC is found to be computationally intensive, making it less feasible for real-time control. SDP shows promising results, but is limited by the curse of dimensionality, restricting the use to higher order models. ADP, which approximates SDP solutions, can offer a potential controller for higher order models but requires further tuning and optimisation.
Simulations are conducted to compare the performance of these control strategies in several scenarios. While SDP demonstrates slight improvements over threshold control on the training dataset, its performance on different wind patterns is less consistent. The study concludes that while proactive control strategies such as SDP and ADP can offer improvements over reactive methods, their performance is dependent on the accuracy of wind predictions and the specific operational conditions.
Future work suggestions include refining the turbine and wind models, exploring adaptive control methods, and conducting real-life experiments to validate the control strategies, which are crucial for practical implementation and optimisation.
The thesis begins with an analysis of the DOT500PRO and its state machine, identifying operational states, transitions, and constraints. A Markov model is used to model and predict wind speeds, which fits nicely with the Markov Decision Process (MDP) framework. The problem is formulated as an MDP, and multiple control strategies, including Threshold Control, Model Predictive Control (MPC), Stochastic Dynamic Programming (SDP), and Approximate Dynamic Programming (ADP), are evaluated.
MPC is found to be computationally intensive, making it less feasible for real-time control. SDP shows promising results, but is limited by the curse of dimensionality, restricting the use to higher order models. ADP, which approximates SDP solutions, can offer a potential controller for higher order models but requires further tuning and optimisation.
Simulations are conducted to compare the performance of these control strategies in several scenarios. While SDP demonstrates slight improvements over threshold control on the training dataset, its performance on different wind patterns is less consistent. The study concludes that while proactive control strategies such as SDP and ADP can offer improvements over reactive methods, their performance is dependent on the accuracy of wind predictions and the specific operational conditions.
Future work suggestions include refining the turbine and wind models, exploring adaptive control methods, and conducting real-life experiments to validate the control strategies, which are crucial for practical implementation and optimisation.
The increasing global container trade has intensified demands on container handling technology. Automation has addressed many challenges, but limitations persist, particularly in spreader design. These attachments, equipped with rotating twist locks, rely on remote-controlled actuators and sensors. However, current designs exhibit frequent maintenance, multiple failure points, and limited sensing capabilities. Additionally, dependence on external power restricts mobility, limiting their use in smaller terminals and remote operations. This research employs an iterative design approach to develop a novel spreader that reduces maintenance frequency by 75 percent, minimizes failure points by 50 percent, and integrates an advanced control algorithm with built-in error detection. This reduces reliance on remote operators, enhancing safety by mitigating operator fatigue-induced errors. An internal power source enables 20 hours of continuous operation, while the independent control system reduces downtime by 1 hour, minimizing delays from operator breaks.
...
The increasing global container trade has intensified demands on container handling technology. Automation has addressed many challenges, but limitations persist, particularly in spreader design. These attachments, equipped with rotating twist locks, rely on remote-controlled actuators and sensors. However, current designs exhibit frequent maintenance, multiple failure points, and limited sensing capabilities. Additionally, dependence on external power restricts mobility, limiting their use in smaller terminals and remote operations. This research employs an iterative design approach to develop a novel spreader that reduces maintenance frequency by 75 percent, minimizes failure points by 50 percent, and integrates an advanced control algorithm with built-in error detection. This reduces reliance on remote operators, enhancing safety by mitigating operator fatigue-induced errors. An internal power source enables 20 hours of continuous operation, while the independent control system reduces downtime by 1 hour, minimizing delays from operator breaks.
Master thesis
(2025)
-
E. Lemmen, J.W. van Wingerden, S.P. Mulders, E. Lourens, Joris Ottenheym, Jelle ter Braak
Monopiles are the most commonly used foundation type for offshore wind turbines. Traditionally, monopile foundations were installed using jack-up vessels, which can fix themselves to the seabed using extendable legs. However, as offshore wind farms expand into deeper waters and monopile sizes continue to grow, jack-up vessels are becoming less suitable. To overcome these challenges, SSCV have become viable alternatives. These vessels are equipped with a DP control system, allowing for precise positioning of the vessel without the need for anchors. This advancement significantly improves installation efficiency. However, because these floating vessels are not fixed to the seabed, their movement introduces a new challenge: maintaining the monopile in a vertical position during the installation process while experiencing vessel motions due to waves, currents, and wind. To mitigate this issue, a MCGF can be used.
A major challenge in monopile installation occurs when the tip of the pile contacts the seabed and establishes a connection with the lateral soil. This introduces an abrupt change in the overall system dynamics and increases the risk of DP instability problems of the vessel. These problems are not new to the offshore industry. However, with the introduction of the MCGF, the risk of DP instability increases, as there is now a much stronger dynamic coupling between the vessel, the monopile and the seabed. Therefore, it is important that the MCGF is properly controlled, as this directly influences the reaction forces applied to the vessel. However, properly tuning the MCGF controller is challenging, as it strongly depends on the lateral soil behavior. Currently, these controller gains are tuned using simulations in which the soil behavior is modeled using CPT data. Nevertheless, the estimated behavior based on this data contain large uncertainties and therefore the control settings might become suboptimal during the installation. Furthermore, since soil dynamics also changes during installation, it is beneficial to use a real-time estimation method that takes this into account. Therefore, this thesis investigates different identification methods to obtain real-time estimates of lateral soil behavior during the monopile installation process.
Two main approaches are explored; the augmented EKF approach and the GPLFM approach. The augmented EKF approach shows to be capable of estimating soil behavior by directly identifying the lateral and rotational soil stiffness values, given that the filter is carefully tuned. However, tuning the filter is computationally demanding due to the large number of tunable parameters, which makes this method impractical for site-specific tuning prior to installation using the first measurements obtained. This limitation is critical, as offshore installations face varying conditions at each site, and accurate estimation therefore requires site-specific tuning. To address this tuning challenge, an alternative method is introduced: the GPLFM approach. In this framework, the identification task is reformulated as a GP regression problem. An important advantage of this method is that the process covariance matrix is determined in a data-driven manner, providing a complete covariance structure governed by only a small number of tunable parameters. Consequently, the parameter space is greatly reduced, enabling efficient site-specific tuning. As a result, it is shown that it is possible to obtain accurate estimation results across varying conditions.
Therefore, it is found that the GPLFM method offers a promising solution for real-time estimation of lateral soil behavior during monopile installation. By providing real-time estimates, this study supports the development of more effective MCGF control strategie. This can in the future be used to improve the MCGF controller as it can now be adjusted automatically during the installation process rather than manually. Furthermore, it can help to prevent DP instability problems. ...
A major challenge in monopile installation occurs when the tip of the pile contacts the seabed and establishes a connection with the lateral soil. This introduces an abrupt change in the overall system dynamics and increases the risk of DP instability problems of the vessel. These problems are not new to the offshore industry. However, with the introduction of the MCGF, the risk of DP instability increases, as there is now a much stronger dynamic coupling between the vessel, the monopile and the seabed. Therefore, it is important that the MCGF is properly controlled, as this directly influences the reaction forces applied to the vessel. However, properly tuning the MCGF controller is challenging, as it strongly depends on the lateral soil behavior. Currently, these controller gains are tuned using simulations in which the soil behavior is modeled using CPT data. Nevertheless, the estimated behavior based on this data contain large uncertainties and therefore the control settings might become suboptimal during the installation. Furthermore, since soil dynamics also changes during installation, it is beneficial to use a real-time estimation method that takes this into account. Therefore, this thesis investigates different identification methods to obtain real-time estimates of lateral soil behavior during the monopile installation process.
Two main approaches are explored; the augmented EKF approach and the GPLFM approach. The augmented EKF approach shows to be capable of estimating soil behavior by directly identifying the lateral and rotational soil stiffness values, given that the filter is carefully tuned. However, tuning the filter is computationally demanding due to the large number of tunable parameters, which makes this method impractical for site-specific tuning prior to installation using the first measurements obtained. This limitation is critical, as offshore installations face varying conditions at each site, and accurate estimation therefore requires site-specific tuning. To address this tuning challenge, an alternative method is introduced: the GPLFM approach. In this framework, the identification task is reformulated as a GP regression problem. An important advantage of this method is that the process covariance matrix is determined in a data-driven manner, providing a complete covariance structure governed by only a small number of tunable parameters. Consequently, the parameter space is greatly reduced, enabling efficient site-specific tuning. As a result, it is shown that it is possible to obtain accurate estimation results across varying conditions.
Therefore, it is found that the GPLFM method offers a promising solution for real-time estimation of lateral soil behavior during monopile installation. By providing real-time estimates, this study supports the development of more effective MCGF control strategie. This can in the future be used to improve the MCGF controller as it can now be adjusted automatically during the installation process rather than manually. Furthermore, it can help to prevent DP instability problems. ...
Monopiles are the most commonly used foundation type for offshore wind turbines. Traditionally, monopile foundations were installed using jack-up vessels, which can fix themselves to the seabed using extendable legs. However, as offshore wind farms expand into deeper waters and monopile sizes continue to grow, jack-up vessels are becoming less suitable. To overcome these challenges, SSCV have become viable alternatives. These vessels are equipped with a DP control system, allowing for precise positioning of the vessel without the need for anchors. This advancement significantly improves installation efficiency. However, because these floating vessels are not fixed to the seabed, their movement introduces a new challenge: maintaining the monopile in a vertical position during the installation process while experiencing vessel motions due to waves, currents, and wind. To mitigate this issue, a MCGF can be used.
A major challenge in monopile installation occurs when the tip of the pile contacts the seabed and establishes a connection with the lateral soil. This introduces an abrupt change in the overall system dynamics and increases the risk of DP instability problems of the vessel. These problems are not new to the offshore industry. However, with the introduction of the MCGF, the risk of DP instability increases, as there is now a much stronger dynamic coupling between the vessel, the monopile and the seabed. Therefore, it is important that the MCGF is properly controlled, as this directly influences the reaction forces applied to the vessel. However, properly tuning the MCGF controller is challenging, as it strongly depends on the lateral soil behavior. Currently, these controller gains are tuned using simulations in which the soil behavior is modeled using CPT data. Nevertheless, the estimated behavior based on this data contain large uncertainties and therefore the control settings might become suboptimal during the installation. Furthermore, since soil dynamics also changes during installation, it is beneficial to use a real-time estimation method that takes this into account. Therefore, this thesis investigates different identification methods to obtain real-time estimates of lateral soil behavior during the monopile installation process.
Two main approaches are explored; the augmented EKF approach and the GPLFM approach. The augmented EKF approach shows to be capable of estimating soil behavior by directly identifying the lateral and rotational soil stiffness values, given that the filter is carefully tuned. However, tuning the filter is computationally demanding due to the large number of tunable parameters, which makes this method impractical for site-specific tuning prior to installation using the first measurements obtained. This limitation is critical, as offshore installations face varying conditions at each site, and accurate estimation therefore requires site-specific tuning. To address this tuning challenge, an alternative method is introduced: the GPLFM approach. In this framework, the identification task is reformulated as a GP regression problem. An important advantage of this method is that the process covariance matrix is determined in a data-driven manner, providing a complete covariance structure governed by only a small number of tunable parameters. Consequently, the parameter space is greatly reduced, enabling efficient site-specific tuning. As a result, it is shown that it is possible to obtain accurate estimation results across varying conditions.
Therefore, it is found that the GPLFM method offers a promising solution for real-time estimation of lateral soil behavior during monopile installation. By providing real-time estimates, this study supports the development of more effective MCGF control strategie. This can in the future be used to improve the MCGF controller as it can now be adjusted automatically during the installation process rather than manually. Furthermore, it can help to prevent DP instability problems.
A major challenge in monopile installation occurs when the tip of the pile contacts the seabed and establishes a connection with the lateral soil. This introduces an abrupt change in the overall system dynamics and increases the risk of DP instability problems of the vessel. These problems are not new to the offshore industry. However, with the introduction of the MCGF, the risk of DP instability increases, as there is now a much stronger dynamic coupling between the vessel, the monopile and the seabed. Therefore, it is important that the MCGF is properly controlled, as this directly influences the reaction forces applied to the vessel. However, properly tuning the MCGF controller is challenging, as it strongly depends on the lateral soil behavior. Currently, these controller gains are tuned using simulations in which the soil behavior is modeled using CPT data. Nevertheless, the estimated behavior based on this data contain large uncertainties and therefore the control settings might become suboptimal during the installation. Furthermore, since soil dynamics also changes during installation, it is beneficial to use a real-time estimation method that takes this into account. Therefore, this thesis investigates different identification methods to obtain real-time estimates of lateral soil behavior during the monopile installation process.
Two main approaches are explored; the augmented EKF approach and the GPLFM approach. The augmented EKF approach shows to be capable of estimating soil behavior by directly identifying the lateral and rotational soil stiffness values, given that the filter is carefully tuned. However, tuning the filter is computationally demanding due to the large number of tunable parameters, which makes this method impractical for site-specific tuning prior to installation using the first measurements obtained. This limitation is critical, as offshore installations face varying conditions at each site, and accurate estimation therefore requires site-specific tuning. To address this tuning challenge, an alternative method is introduced: the GPLFM approach. In this framework, the identification task is reformulated as a GP regression problem. An important advantage of this method is that the process covariance matrix is determined in a data-driven manner, providing a complete covariance structure governed by only a small number of tunable parameters. Consequently, the parameter space is greatly reduced, enabling efficient site-specific tuning. As a result, it is shown that it is possible to obtain accurate estimation results across varying conditions.
Therefore, it is found that the GPLFM method offers a promising solution for real-time estimation of lateral soil behavior during monopile installation. By providing real-time estimates, this study supports the development of more effective MCGF control strategie. This can in the future be used to improve the MCGF controller as it can now be adjusted automatically during the installation process rather than manually. Furthermore, it can help to prevent DP instability problems.
In recent years, the use of renewable energy sources has increased significantly. As a result, offshore wind energy has expanded due to its advantages compared to onshore, such as steadier winds, reduced visual impact and lower noise emissions. However, 80% of the global offshore wind resource potential is located in areas with water depths greater than 60 meters, making the installation of fixed-bottom wind turbines unfeasible and leading to an increase in floating wind turbines. Floating offshore wind turbines, due to their increased freedom of movement, experience more complex aerodynamic and hydrodynamic phenomena, resulting in different power generation compared to fixed-bottom wind turbines. The effect of six degrees of freedom on power production has not been extensively studied in full-scale operating wind turbines. Additionally, a better prediction of energy production could also reduce the investment risk of floating wind turbines, leading to a reduced levelised cost of energy and better integration of floating wind turbines into a fully renewable energy system.
The aim of this report was to investigate the power generation and response of the TetraSpar demonstrator project, the world’s first fully industrialised floating offshore foundation. To achieve this, a model of the TetraSpar demonstrator was created in OpenFAST. Due to confidentiality, some data, particularly regarding the wind turbine, were unavailable and were instead based on scaling the NREL 5 MW wind turbine. In addition, the ROSCO controller was incorporated into the model and the floating platform was modelled as a six degrees of freedom rigid body.
For the validation of the OpenFAST model, a comparison was performed with the on-site TetraSpar demonstrator. Data for the on-site demonstrator regarding motions, metocean conditions and power generation were provided by RWE. Further filtering and averaging the data in 10-minute periods for an entire year was performed. First, three one-hour simulations were performed for three different operating conditions: cut-in, below rated and above rated. Results from time series and power spectral density analyses showed a good agreement in platform motions. For the generated power, the mean average produced power was closest to the on-site data in the above rated region.
Next, for the power curve and AEP estimation, the probability of occurrence for each wind speed in the operating region was determined based on the on-site measured data. Furthermore, for every wind speed, the most representative significant wave height, wave period, turbulence intensity and other parameters needed as inputs in OpenFAST were found. The results showed that the AEP predicted by OpenFAST was 2.8% lower than the AEP measured at the on-site TetraSpar demonstrator. Moreover, different peak shaving levels of the controller contributed to the AEP percentage difference, ranging from about 4.4% to 2.1% compared to the on-site measurements. As far as platform motions are concerned, the model showed good agreement, capturing the trend of mean values and standard deviations of the two most dominant motions in power production, surge and pitch. In addition, the effect of wave height and wave period showed that the mean generated power is slightly affected. Similar results were also found for different current velocities. However, waves were found to influence the oscillation amplitudes of surge and pitch motions, while currents mainly affected the shift in mean values. Lastly, wind-wave misalignment also proved to have a negative effect on power performance.
In conclusion, the current OpenFAST model can be used to estimate the energy production and platform motion of the TetraSpar demonstrator. However, further improvements and reduction of the assumptions made, mainly due to the unavailability of data because of confidentiality, could enhance its accuracy. ...
The aim of this report was to investigate the power generation and response of the TetraSpar demonstrator project, the world’s first fully industrialised floating offshore foundation. To achieve this, a model of the TetraSpar demonstrator was created in OpenFAST. Due to confidentiality, some data, particularly regarding the wind turbine, were unavailable and were instead based on scaling the NREL 5 MW wind turbine. In addition, the ROSCO controller was incorporated into the model and the floating platform was modelled as a six degrees of freedom rigid body.
For the validation of the OpenFAST model, a comparison was performed with the on-site TetraSpar demonstrator. Data for the on-site demonstrator regarding motions, metocean conditions and power generation were provided by RWE. Further filtering and averaging the data in 10-minute periods for an entire year was performed. First, three one-hour simulations were performed for three different operating conditions: cut-in, below rated and above rated. Results from time series and power spectral density analyses showed a good agreement in platform motions. For the generated power, the mean average produced power was closest to the on-site data in the above rated region.
Next, for the power curve and AEP estimation, the probability of occurrence for each wind speed in the operating region was determined based on the on-site measured data. Furthermore, for every wind speed, the most representative significant wave height, wave period, turbulence intensity and other parameters needed as inputs in OpenFAST were found. The results showed that the AEP predicted by OpenFAST was 2.8% lower than the AEP measured at the on-site TetraSpar demonstrator. Moreover, different peak shaving levels of the controller contributed to the AEP percentage difference, ranging from about 4.4% to 2.1% compared to the on-site measurements. As far as platform motions are concerned, the model showed good agreement, capturing the trend of mean values and standard deviations of the two most dominant motions in power production, surge and pitch. In addition, the effect of wave height and wave period showed that the mean generated power is slightly affected. Similar results were also found for different current velocities. However, waves were found to influence the oscillation amplitudes of surge and pitch motions, while currents mainly affected the shift in mean values. Lastly, wind-wave misalignment also proved to have a negative effect on power performance.
In conclusion, the current OpenFAST model can be used to estimate the energy production and platform motion of the TetraSpar demonstrator. However, further improvements and reduction of the assumptions made, mainly due to the unavailability of data because of confidentiality, could enhance its accuracy. ...
In recent years, the use of renewable energy sources has increased significantly. As a result, offshore wind energy has expanded due to its advantages compared to onshore, such as steadier winds, reduced visual impact and lower noise emissions. However, 80% of the global offshore wind resource potential is located in areas with water depths greater than 60 meters, making the installation of fixed-bottom wind turbines unfeasible and leading to an increase in floating wind turbines. Floating offshore wind turbines, due to their increased freedom of movement, experience more complex aerodynamic and hydrodynamic phenomena, resulting in different power generation compared to fixed-bottom wind turbines. The effect of six degrees of freedom on power production has not been extensively studied in full-scale operating wind turbines. Additionally, a better prediction of energy production could also reduce the investment risk of floating wind turbines, leading to a reduced levelised cost of energy and better integration of floating wind turbines into a fully renewable energy system.
The aim of this report was to investigate the power generation and response of the TetraSpar demonstrator project, the world’s first fully industrialised floating offshore foundation. To achieve this, a model of the TetraSpar demonstrator was created in OpenFAST. Due to confidentiality, some data, particularly regarding the wind turbine, were unavailable and were instead based on scaling the NREL 5 MW wind turbine. In addition, the ROSCO controller was incorporated into the model and the floating platform was modelled as a six degrees of freedom rigid body.
For the validation of the OpenFAST model, a comparison was performed with the on-site TetraSpar demonstrator. Data for the on-site demonstrator regarding motions, metocean conditions and power generation were provided by RWE. Further filtering and averaging the data in 10-minute periods for an entire year was performed. First, three one-hour simulations were performed for three different operating conditions: cut-in, below rated and above rated. Results from time series and power spectral density analyses showed a good agreement in platform motions. For the generated power, the mean average produced power was closest to the on-site data in the above rated region.
Next, for the power curve and AEP estimation, the probability of occurrence for each wind speed in the operating region was determined based on the on-site measured data. Furthermore, for every wind speed, the most representative significant wave height, wave period, turbulence intensity and other parameters needed as inputs in OpenFAST were found. The results showed that the AEP predicted by OpenFAST was 2.8% lower than the AEP measured at the on-site TetraSpar demonstrator. Moreover, different peak shaving levels of the controller contributed to the AEP percentage difference, ranging from about 4.4% to 2.1% compared to the on-site measurements. As far as platform motions are concerned, the model showed good agreement, capturing the trend of mean values and standard deviations of the two most dominant motions in power production, surge and pitch. In addition, the effect of wave height and wave period showed that the mean generated power is slightly affected. Similar results were also found for different current velocities. However, waves were found to influence the oscillation amplitudes of surge and pitch motions, while currents mainly affected the shift in mean values. Lastly, wind-wave misalignment also proved to have a negative effect on power performance.
In conclusion, the current OpenFAST model can be used to estimate the energy production and platform motion of the TetraSpar demonstrator. However, further improvements and reduction of the assumptions made, mainly due to the unavailability of data because of confidentiality, could enhance its accuracy.
The aim of this report was to investigate the power generation and response of the TetraSpar demonstrator project, the world’s first fully industrialised floating offshore foundation. To achieve this, a model of the TetraSpar demonstrator was created in OpenFAST. Due to confidentiality, some data, particularly regarding the wind turbine, were unavailable and were instead based on scaling the NREL 5 MW wind turbine. In addition, the ROSCO controller was incorporated into the model and the floating platform was modelled as a six degrees of freedom rigid body.
For the validation of the OpenFAST model, a comparison was performed with the on-site TetraSpar demonstrator. Data for the on-site demonstrator regarding motions, metocean conditions and power generation were provided by RWE. Further filtering and averaging the data in 10-minute periods for an entire year was performed. First, three one-hour simulations were performed for three different operating conditions: cut-in, below rated and above rated. Results from time series and power spectral density analyses showed a good agreement in platform motions. For the generated power, the mean average produced power was closest to the on-site data in the above rated region.
Next, for the power curve and AEP estimation, the probability of occurrence for each wind speed in the operating region was determined based on the on-site measured data. Furthermore, for every wind speed, the most representative significant wave height, wave period, turbulence intensity and other parameters needed as inputs in OpenFAST were found. The results showed that the AEP predicted by OpenFAST was 2.8% lower than the AEP measured at the on-site TetraSpar demonstrator. Moreover, different peak shaving levels of the controller contributed to the AEP percentage difference, ranging from about 4.4% to 2.1% compared to the on-site measurements. As far as platform motions are concerned, the model showed good agreement, capturing the trend of mean values and standard deviations of the two most dominant motions in power production, surge and pitch. In addition, the effect of wave height and wave period showed that the mean generated power is slightly affected. Similar results were also found for different current velocities. However, waves were found to influence the oscillation amplitudes of surge and pitch motions, while currents mainly affected the shift in mean values. Lastly, wind-wave misalignment also proved to have a negative effect on power performance.
In conclusion, the current OpenFAST model can be used to estimate the energy production and platform motion of the TetraSpar demonstrator. However, further improvements and reduction of the assumptions made, mainly due to the unavailability of data because of confidentiality, could enhance its accuracy.
Optimisation of the Absolute Sine for Dynamic Induction Control
Power Maximisation of a Wind Farm Through Infinite and Finite Horizon Control
Master thesis
(2024)
-
J.A.G. Kuppers, J.W. van Wingerden, M.J. van den Broek, S.P. Mulders, A.A.W. van Vondelen
The clustering of wind turbines in a wind farm results in overall efficiency losses as downstream wind turbines operate in the wake of their upstream neighbours. Wind farm flow control (WFFC) strategies have emerged to reduce these wake effects with the goal of maximising overall performance. Dynamic induction control (DIC) aims to enhance the wake breakdown and restore the wake's energy content through dynamic thrust variations. The control signals are often found through the economic model predictive control (EMPC) method, which relies on an internal model to incorporate future system behaviour in the determination of the next optimal control input. These models are designed to capture the most dominant wake characteristics while remaining computationally efficient. We employ the two-dimensional free-vortex wake (FVW) model presented from [1], which models the wake through vortex element pairs released from the edges of the actuator disc. The power of the two-turbine wind farm is maximised through EMPC, improving performance by 9.64\% over greedy control simulations. However, the EMPC method inherits finite horizon effects, resulting in large control horizons to optimise. In this study, we address these limitations by employing an absolute sine parameterisation in the FVW model to limit the finite horizon effects and reduce the dimension of the optimisation problem. The significant dimension reduction allows for a grid search to find the optimal infinite horizon steady-state solution, improving the mean steady-state performance by 2.43\% over the baseline results from [1]. Additionally, we focus on converging towards this optimum through finite horizon EMPC optimising over the amplitude and offset. Grid search analyses reveal sensitivity towards initialisation due to the appearance of local minima around the infinite horizon optimum. A maximum success rate is realised for very large control horizons, maximising the probability of converging towards the infinite horizon optimum. Accounting for the inherited system delay in the objective function also realises a maximum success rate but for shorter control horizons, which significantly decreases the simulation time. The final controller design terminates simulations ten times faster through the absolute sine parameterisation compared to the baseline simulation from [1] while maximising the probability of convergence towards the infinite horizon optimum.
...
The clustering of wind turbines in a wind farm results in overall efficiency losses as downstream wind turbines operate in the wake of their upstream neighbours. Wind farm flow control (WFFC) strategies have emerged to reduce these wake effects with the goal of maximising overall performance. Dynamic induction control (DIC) aims to enhance the wake breakdown and restore the wake's energy content through dynamic thrust variations. The control signals are often found through the economic model predictive control (EMPC) method, which relies on an internal model to incorporate future system behaviour in the determination of the next optimal control input. These models are designed to capture the most dominant wake characteristics while remaining computationally efficient. We employ the two-dimensional free-vortex wake (FVW) model presented from [1], which models the wake through vortex element pairs released from the edges of the actuator disc. The power of the two-turbine wind farm is maximised through EMPC, improving performance by 9.64\% over greedy control simulations. However, the EMPC method inherits finite horizon effects, resulting in large control horizons to optimise. In this study, we address these limitations by employing an absolute sine parameterisation in the FVW model to limit the finite horizon effects and reduce the dimension of the optimisation problem. The significant dimension reduction allows for a grid search to find the optimal infinite horizon steady-state solution, improving the mean steady-state performance by 2.43\% over the baseline results from [1]. Additionally, we focus on converging towards this optimum through finite horizon EMPC optimising over the amplitude and offset. Grid search analyses reveal sensitivity towards initialisation due to the appearance of local minima around the infinite horizon optimum. A maximum success rate is realised for very large control horizons, maximising the probability of converging towards the infinite horizon optimum. Accounting for the inherited system delay in the objective function also realises a maximum success rate but for shorter control horizons, which significantly decreases the simulation time. The final controller design terminates simulations ten times faster through the absolute sine parameterisation compared to the baseline simulation from [1] while maximising the probability of convergence towards the infinite horizon optimum.
Master thesis
(2023)
-
J.D. van de Ketterij, S.H. Hossein Nia Kani, N. Karbasizadeh Esfahani, R.N. Buitenhuis, S.P. Mulders
Inherent limitations of linear control, including PID, are Bode's phase-gain relationship and the waterbed-effect. Literature showed that reset control can overcome these limitations. The 'Constant in Gain, Lead in Phase' (CgLp) reset element was developed, which can be used in combination with a conventional PID approach. The benefit is that CgLp increases the broadband phase without altering the gain.
Recently, a continuous reset (CR) CgLp was proposed to eliminate the discontinuous output of the CgLp, which for example, excites resonances in mechanical structures. Unfortunately, CR CgLp is more sensitive to measurement noise than CgLp due to CR CgLp taking the derivative of its input, amplifying noise. This causes resets to occur more frequently. Literature has shown that these excessive resets can deteriorate the system performance.
The main contribution of this work is to improve the noise robustness of CR CgLp by implementing a filter on the measurement line to attenuate measurement noise. The attenuation of noise causes a reduction of excessive resets. Two filters are designed and critically compared. First, a low-pass filter succeeded by a lead element (LL). Second, an accelerometer-aided Kalman filter (aaKF). These were validated in practice on a precision-positioning stage.
Simulations have shown that the proposed designs improve performance of CR CgLp. The overshoot of CR CgLp could be reduced by 32% and IAE could be reduced by 53%. Besides this, CR CgLp with aaKF showed to outperform a PID controller with similar bandwidth and phase margin. The overshoot could be reduced by 46% and the IAE by 34%.
Test results on a physical setup have shown that the LL filter could improve the IAE of a step response by 13% compared to a CR CgLp without filter. Compared to PID control, the CR CgLp with LL filter has a superior transient response. ...
Recently, a continuous reset (CR) CgLp was proposed to eliminate the discontinuous output of the CgLp, which for example, excites resonances in mechanical structures. Unfortunately, CR CgLp is more sensitive to measurement noise than CgLp due to CR CgLp taking the derivative of its input, amplifying noise. This causes resets to occur more frequently. Literature has shown that these excessive resets can deteriorate the system performance.
The main contribution of this work is to improve the noise robustness of CR CgLp by implementing a filter on the measurement line to attenuate measurement noise. The attenuation of noise causes a reduction of excessive resets. Two filters are designed and critically compared. First, a low-pass filter succeeded by a lead element (LL). Second, an accelerometer-aided Kalman filter (aaKF). These were validated in practice on a precision-positioning stage.
Simulations have shown that the proposed designs improve performance of CR CgLp. The overshoot of CR CgLp could be reduced by 32% and IAE could be reduced by 53%. Besides this, CR CgLp with aaKF showed to outperform a PID controller with similar bandwidth and phase margin. The overshoot could be reduced by 46% and the IAE by 34%.
Test results on a physical setup have shown that the LL filter could improve the IAE of a step response by 13% compared to a CR CgLp without filter. Compared to PID control, the CR CgLp with LL filter has a superior transient response. ...
Inherent limitations of linear control, including PID, are Bode's phase-gain relationship and the waterbed-effect. Literature showed that reset control can overcome these limitations. The 'Constant in Gain, Lead in Phase' (CgLp) reset element was developed, which can be used in combination with a conventional PID approach. The benefit is that CgLp increases the broadband phase without altering the gain.
Recently, a continuous reset (CR) CgLp was proposed to eliminate the discontinuous output of the CgLp, which for example, excites resonances in mechanical structures. Unfortunately, CR CgLp is more sensitive to measurement noise than CgLp due to CR CgLp taking the derivative of its input, amplifying noise. This causes resets to occur more frequently. Literature has shown that these excessive resets can deteriorate the system performance.
The main contribution of this work is to improve the noise robustness of CR CgLp by implementing a filter on the measurement line to attenuate measurement noise. The attenuation of noise causes a reduction of excessive resets. Two filters are designed and critically compared. First, a low-pass filter succeeded by a lead element (LL). Second, an accelerometer-aided Kalman filter (aaKF). These were validated in practice on a precision-positioning stage.
Simulations have shown that the proposed designs improve performance of CR CgLp. The overshoot of CR CgLp could be reduced by 32% and IAE could be reduced by 53%. Besides this, CR CgLp with aaKF showed to outperform a PID controller with similar bandwidth and phase margin. The overshoot could be reduced by 46% and the IAE by 34%.
Test results on a physical setup have shown that the LL filter could improve the IAE of a step response by 13% compared to a CR CgLp without filter. Compared to PID control, the CR CgLp with LL filter has a superior transient response.
Recently, a continuous reset (CR) CgLp was proposed to eliminate the discontinuous output of the CgLp, which for example, excites resonances in mechanical structures. Unfortunately, CR CgLp is more sensitive to measurement noise than CgLp due to CR CgLp taking the derivative of its input, amplifying noise. This causes resets to occur more frequently. Literature has shown that these excessive resets can deteriorate the system performance.
The main contribution of this work is to improve the noise robustness of CR CgLp by implementing a filter on the measurement line to attenuate measurement noise. The attenuation of noise causes a reduction of excessive resets. Two filters are designed and critically compared. First, a low-pass filter succeeded by a lead element (LL). Second, an accelerometer-aided Kalman filter (aaKF). These were validated in practice on a precision-positioning stage.
Simulations have shown that the proposed designs improve performance of CR CgLp. The overshoot of CR CgLp could be reduced by 32% and IAE could be reduced by 53%. Besides this, CR CgLp with aaKF showed to outperform a PID controller with similar bandwidth and phase margin. The overshoot could be reduced by 46% and the IAE by 34%.
Test results on a physical setup have shown that the LL filter could improve the IAE of a step response by 13% compared to a CR CgLp without filter. Compared to PID control, the CR CgLp with LL filter has a superior transient response.
Wind turbine installation vessels (WTIVs) are ships that are specifically designed to install offshore wind turbines. These WTIVs have four or six large truss-like legs that are lowered towards the seabed by means of jacking systems. These jacking systems are regulated by control systems to ensure the leg is lowered with a constant velocity regardless of external disturbances. The bottom of these legs are outfitted with spudcans, which have a conical shape to allow for penetration into the seabed. When these spudcans have settled in the seabed, the hull is lifted above the waves. Consequently, wave-exciting forces on the hull are prevented and the motions of the hull are near-zero so that the on-board crane can perform the installation operations with minimised disturbances.
The objective of this thesis is to develop and analyse a model that is able to describe and simulate the dynamics of these jacking systems in great detail in response to external loads and the dynamics of the WTIV. The control systems of the jacking systems are included in this model to simulate and evaluate the interaction between the control system and the dynamics of the WTIV. Two conventional control systems are considered: the Volts-per-Hertz (V/Hz) and the direct torque control (DTC) method. In the process of lowering the legs and subsequent platform lifting, a transient phase can be identified during which the spudcans are penetrating the seabed. Due to the periodical motions of the ships, multiple impacts with the seabed are expected. Additionally, the jacking systems and the leg undergo a change of load direction as initially the leg is in tension and the jacking systems are generating power, and afterwards the leg is in compression and the jacking systems are consuming power. This thesis is focused on this seabed penetration phase as this phase introduces complicated dynamics. In literature, no model is available that has the abilities to simulate the WTIVs and its jacking systems with control systems in such level of detail.
This research gap is addressed by developing such a simulation model. This model is written in Python and developed using finite element (FE) techniques and solved using numerical time integration. Seabed characteristics are derived using a detailed coupled Eulerian-Lagrangian (CEL) FE models. Multiple control strategies are simulated and evaluated, each differentiating how the velocity and torque setpoints of the jacking systems are calculated. From the simulation model, it is found that in order to achieve load sharing between jacking systems, torque and velocity require to be independently controlled which only the DTC method has the ability to. Furthermore, each of the jacking systems should be provided with its own power supply. Best performance and stability was achieved when each chord of the leg is given a common torque and velocity setpoint, which is equivalent to a common torque and setpoint per leg in reality. Moreover, load sharing can be improved without a control system by increasing the relative stiffness ratio between the chord and the mechanical contact between rack and pinion. ...
The objective of this thesis is to develop and analyse a model that is able to describe and simulate the dynamics of these jacking systems in great detail in response to external loads and the dynamics of the WTIV. The control systems of the jacking systems are included in this model to simulate and evaluate the interaction between the control system and the dynamics of the WTIV. Two conventional control systems are considered: the Volts-per-Hertz (V/Hz) and the direct torque control (DTC) method. In the process of lowering the legs and subsequent platform lifting, a transient phase can be identified during which the spudcans are penetrating the seabed. Due to the periodical motions of the ships, multiple impacts with the seabed are expected. Additionally, the jacking systems and the leg undergo a change of load direction as initially the leg is in tension and the jacking systems are generating power, and afterwards the leg is in compression and the jacking systems are consuming power. This thesis is focused on this seabed penetration phase as this phase introduces complicated dynamics. In literature, no model is available that has the abilities to simulate the WTIVs and its jacking systems with control systems in such level of detail.
This research gap is addressed by developing such a simulation model. This model is written in Python and developed using finite element (FE) techniques and solved using numerical time integration. Seabed characteristics are derived using a detailed coupled Eulerian-Lagrangian (CEL) FE models. Multiple control strategies are simulated and evaluated, each differentiating how the velocity and torque setpoints of the jacking systems are calculated. From the simulation model, it is found that in order to achieve load sharing between jacking systems, torque and velocity require to be independently controlled which only the DTC method has the ability to. Furthermore, each of the jacking systems should be provided with its own power supply. Best performance and stability was achieved when each chord of the leg is given a common torque and velocity setpoint, which is equivalent to a common torque and setpoint per leg in reality. Moreover, load sharing can be improved without a control system by increasing the relative stiffness ratio between the chord and the mechanical contact between rack and pinion. ...
Wind turbine installation vessels (WTIVs) are ships that are specifically designed to install offshore wind turbines. These WTIVs have four or six large truss-like legs that are lowered towards the seabed by means of jacking systems. These jacking systems are regulated by control systems to ensure the leg is lowered with a constant velocity regardless of external disturbances. The bottom of these legs are outfitted with spudcans, which have a conical shape to allow for penetration into the seabed. When these spudcans have settled in the seabed, the hull is lifted above the waves. Consequently, wave-exciting forces on the hull are prevented and the motions of the hull are near-zero so that the on-board crane can perform the installation operations with minimised disturbances.
The objective of this thesis is to develop and analyse a model that is able to describe and simulate the dynamics of these jacking systems in great detail in response to external loads and the dynamics of the WTIV. The control systems of the jacking systems are included in this model to simulate and evaluate the interaction between the control system and the dynamics of the WTIV. Two conventional control systems are considered: the Volts-per-Hertz (V/Hz) and the direct torque control (DTC) method. In the process of lowering the legs and subsequent platform lifting, a transient phase can be identified during which the spudcans are penetrating the seabed. Due to the periodical motions of the ships, multiple impacts with the seabed are expected. Additionally, the jacking systems and the leg undergo a change of load direction as initially the leg is in tension and the jacking systems are generating power, and afterwards the leg is in compression and the jacking systems are consuming power. This thesis is focused on this seabed penetration phase as this phase introduces complicated dynamics. In literature, no model is available that has the abilities to simulate the WTIVs and its jacking systems with control systems in such level of detail.
This research gap is addressed by developing such a simulation model. This model is written in Python and developed using finite element (FE) techniques and solved using numerical time integration. Seabed characteristics are derived using a detailed coupled Eulerian-Lagrangian (CEL) FE models. Multiple control strategies are simulated and evaluated, each differentiating how the velocity and torque setpoints of the jacking systems are calculated. From the simulation model, it is found that in order to achieve load sharing between jacking systems, torque and velocity require to be independently controlled which only the DTC method has the ability to. Furthermore, each of the jacking systems should be provided with its own power supply. Best performance and stability was achieved when each chord of the leg is given a common torque and velocity setpoint, which is equivalent to a common torque and setpoint per leg in reality. Moreover, load sharing can be improved without a control system by increasing the relative stiffness ratio between the chord and the mechanical contact between rack and pinion.
The objective of this thesis is to develop and analyse a model that is able to describe and simulate the dynamics of these jacking systems in great detail in response to external loads and the dynamics of the WTIV. The control systems of the jacking systems are included in this model to simulate and evaluate the interaction between the control system and the dynamics of the WTIV. Two conventional control systems are considered: the Volts-per-Hertz (V/Hz) and the direct torque control (DTC) method. In the process of lowering the legs and subsequent platform lifting, a transient phase can be identified during which the spudcans are penetrating the seabed. Due to the periodical motions of the ships, multiple impacts with the seabed are expected. Additionally, the jacking systems and the leg undergo a change of load direction as initially the leg is in tension and the jacking systems are generating power, and afterwards the leg is in compression and the jacking systems are consuming power. This thesis is focused on this seabed penetration phase as this phase introduces complicated dynamics. In literature, no model is available that has the abilities to simulate the WTIVs and its jacking systems with control systems in such level of detail.
This research gap is addressed by developing such a simulation model. This model is written in Python and developed using finite element (FE) techniques and solved using numerical time integration. Seabed characteristics are derived using a detailed coupled Eulerian-Lagrangian (CEL) FE models. Multiple control strategies are simulated and evaluated, each differentiating how the velocity and torque setpoints of the jacking systems are calculated. From the simulation model, it is found that in order to achieve load sharing between jacking systems, torque and velocity require to be independently controlled which only the DTC method has the ability to. Furthermore, each of the jacking systems should be provided with its own power supply. Best performance and stability was achieved when each chord of the leg is given a common torque and velocity setpoint, which is equivalent to a common torque and setpoint per leg in reality. Moreover, load sharing can be improved without a control system by increasing the relative stiffness ratio between the chord and the mechanical contact between rack and pinion.
Bathymetry SLAM using reduced rank Gaussian Processes and DVL range measurements
For real-time underwater position estimation
Underwater position estimation is challenging due to the absence of Global Navigation Satellite System (GNSS) signals. Underwater vehicles are typically equipped with a Doppler Velocity Log (DVL) that measures the velocity relative to the seafloor. Aside from the velocity, the DVL also measures the range of each of the four beams. When compared against a bathymetry height map, these measured ranges provide additional information enabling improving position estimates. Unfortunately, surveys often occur in areas where detailed bathymetry maps are unavailable. In these cases, bathymetry Simultaneous Localization and Mapping (SLAM) could be used to improve the position estimates compared to the velocity integration position. With SLAM, the map is being estimated during the mission while at the same time using the map for position determination.
In this thesis, reduced rank Gaussian processes (GPs) are used as map representation for SLAM. The downside of regular GPs is a time complexity of O(n^3) , reduced rank GPs improve the computation performance. GPs provide Gaussian distributions at any point, leading to a neat integration with probabilistic SLAM algorithms. To the best of the author’s knowledge, this has not been used in bathymetry SLAM. This report investigates how reduced rank approximated GPs, representing the bathymetry, can be integrated into a SLAM algorithm to improve the position estimates of an underwater vehicle equipped with a DVL and low-quality gyroscopes.
A squared exponential kernel is used as GPs model of the bathymetry. The reduced rank approximation is vital for real-time SLAM performance. This map representation is integrated with a Rao Blackwellized particle filter (RBPF) that estimates both the underwater vehicle’s trajectory and the bathymetry map.
The SLAM algorithm is evaluated using data from an underwater vehicle operated at the surface such that a GNSS reference position is available. Experiments of the SLAM algorithm show a reduced position error compared to the GNSS reference. The resulting algorithm has a computation time of up to 30 times faster than the Autonomous Underwater Vehicle (AUV) collects data while improving position estimates. This concludes that the RBPF using reduced rank GPs is capable of onboard improved position estimation on underwater vehicles. ...
In this thesis, reduced rank Gaussian processes (GPs) are used as map representation for SLAM. The downside of regular GPs is a time complexity of O(n^3) , reduced rank GPs improve the computation performance. GPs provide Gaussian distributions at any point, leading to a neat integration with probabilistic SLAM algorithms. To the best of the author’s knowledge, this has not been used in bathymetry SLAM. This report investigates how reduced rank approximated GPs, representing the bathymetry, can be integrated into a SLAM algorithm to improve the position estimates of an underwater vehicle equipped with a DVL and low-quality gyroscopes.
A squared exponential kernel is used as GPs model of the bathymetry. The reduced rank approximation is vital for real-time SLAM performance. This map representation is integrated with a Rao Blackwellized particle filter (RBPF) that estimates both the underwater vehicle’s trajectory and the bathymetry map.
The SLAM algorithm is evaluated using data from an underwater vehicle operated at the surface such that a GNSS reference position is available. Experiments of the SLAM algorithm show a reduced position error compared to the GNSS reference. The resulting algorithm has a computation time of up to 30 times faster than the Autonomous Underwater Vehicle (AUV) collects data while improving position estimates. This concludes that the RBPF using reduced rank GPs is capable of onboard improved position estimation on underwater vehicles. ...
Underwater position estimation is challenging due to the absence of Global Navigation Satellite System (GNSS) signals. Underwater vehicles are typically equipped with a Doppler Velocity Log (DVL) that measures the velocity relative to the seafloor. Aside from the velocity, the DVL also measures the range of each of the four beams. When compared against a bathymetry height map, these measured ranges provide additional information enabling improving position estimates. Unfortunately, surveys often occur in areas where detailed bathymetry maps are unavailable. In these cases, bathymetry Simultaneous Localization and Mapping (SLAM) could be used to improve the position estimates compared to the velocity integration position. With SLAM, the map is being estimated during the mission while at the same time using the map for position determination.
In this thesis, reduced rank Gaussian processes (GPs) are used as map representation for SLAM. The downside of regular GPs is a time complexity of O(n^3) , reduced rank GPs improve the computation performance. GPs provide Gaussian distributions at any point, leading to a neat integration with probabilistic SLAM algorithms. To the best of the author’s knowledge, this has not been used in bathymetry SLAM. This report investigates how reduced rank approximated GPs, representing the bathymetry, can be integrated into a SLAM algorithm to improve the position estimates of an underwater vehicle equipped with a DVL and low-quality gyroscopes.
A squared exponential kernel is used as GPs model of the bathymetry. The reduced rank approximation is vital for real-time SLAM performance. This map representation is integrated with a Rao Blackwellized particle filter (RBPF) that estimates both the underwater vehicle’s trajectory and the bathymetry map.
The SLAM algorithm is evaluated using data from an underwater vehicle operated at the surface such that a GNSS reference position is available. Experiments of the SLAM algorithm show a reduced position error compared to the GNSS reference. The resulting algorithm has a computation time of up to 30 times faster than the Autonomous Underwater Vehicle (AUV) collects data while improving position estimates. This concludes that the RBPF using reduced rank GPs is capable of onboard improved position estimation on underwater vehicles.
In this thesis, reduced rank Gaussian processes (GPs) are used as map representation for SLAM. The downside of regular GPs is a time complexity of O(n^3) , reduced rank GPs improve the computation performance. GPs provide Gaussian distributions at any point, leading to a neat integration with probabilistic SLAM algorithms. To the best of the author’s knowledge, this has not been used in bathymetry SLAM. This report investigates how reduced rank approximated GPs, representing the bathymetry, can be integrated into a SLAM algorithm to improve the position estimates of an underwater vehicle equipped with a DVL and low-quality gyroscopes.
A squared exponential kernel is used as GPs model of the bathymetry. The reduced rank approximation is vital for real-time SLAM performance. This map representation is integrated with a Rao Blackwellized particle filter (RBPF) that estimates both the underwater vehicle’s trajectory and the bathymetry map.
The SLAM algorithm is evaluated using data from an underwater vehicle operated at the surface such that a GNSS reference position is available. Experiments of the SLAM algorithm show a reduced position error compared to the GNSS reference. The resulting algorithm has a computation time of up to 30 times faster than the Autonomous Underwater Vehicle (AUV) collects data while improving position estimates. This concludes that the RBPF using reduced rank GPs is capable of onboard improved position estimation on underwater vehicles.
Master thesis
(2017)
-
Marco Stijnman, David Abbink, Jan-Willem van Wingerden, Roel Kuiper, Sebastiaan Mulders
Offshore excavators are large hydraulically driven machines which are difficult to control due to slow dynamics, inherent nonlinearities, a varying environment and complex kinematics. As digging is performed under water, only limited visual feedback of the task can be provided by means of a visualization interface. Operators require an extensive amount of practice before being capable of achieving sufficient and consistent performance. Often, automation is implemented as a means of reducing costs related to expensive operators and attaining consistent performance. However, automation struggles with adapting to unforeseen situations and a large task variety, which are areas human operators excel in. Instead of attempting to fully automate excavators, this thesis takes a more human-centered approach, and focuses on the design and evaluation of a human-like controller to partially automate excavator operations, while assuming a human operator is still present to trade or share control with. In order to simultaneously deal with the various nonlinearities in the system while providing human-like control this work proposes the use of an Adaptive Model Predictive Controller, whose underlying principles are similar to those of humans.To determine whether the controller is indeed human-like a complex excavator model including a realistic soil model was developed and used to implement and tune the controller. Finally, a simulator experiment was conducted to compare the subjects and the controller in terms of performance for various tasks and the control behavior similarity for a well-trained task. Eight subjects controlled the excavator model and performed four stages, starting with a familiarization stage in which the subject got accustomed to the system. The other three stages (easy, difficult, boulder) featured a 9 m long target path, with conditions of varying difficulty between stages. The controller showed 2 to 3 times lower tracking errors for both the easy and difficult stage while providing 1.5 to 5 times smoother inputs, but could not overcome the unforeseen boulder whereas all subjects could, showcasing the importance of having humans and automation complement each other. Furthermore, a high quality fit (VAF > 70%) was found between the boom inputs of the subjects and the controller in the well-trained easy stage, indicating human-like control.
...
Offshore excavators are large hydraulically driven machines which are difficult to control due to slow dynamics, inherent nonlinearities, a varying environment and complex kinematics. As digging is performed under water, only limited visual feedback of the task can be provided by means of a visualization interface. Operators require an extensive amount of practice before being capable of achieving sufficient and consistent performance. Often, automation is implemented as a means of reducing costs related to expensive operators and attaining consistent performance. However, automation struggles with adapting to unforeseen situations and a large task variety, which are areas human operators excel in. Instead of attempting to fully automate excavators, this thesis takes a more human-centered approach, and focuses on the design and evaluation of a human-like controller to partially automate excavator operations, while assuming a human operator is still present to trade or share control with. In order to simultaneously deal with the various nonlinearities in the system while providing human-like control this work proposes the use of an Adaptive Model Predictive Controller, whose underlying principles are similar to those of humans.To determine whether the controller is indeed human-like a complex excavator model including a realistic soil model was developed and used to implement and tune the controller. Finally, a simulator experiment was conducted to compare the subjects and the controller in terms of performance for various tasks and the control behavior similarity for a well-trained task. Eight subjects controlled the excavator model and performed four stages, starting with a familiarization stage in which the subject got accustomed to the system. The other three stages (easy, difficult, boulder) featured a 9 m long target path, with conditions of varying difficulty between stages. The controller showed 2 to 3 times lower tracking errors for both the easy and difficult stage while providing 1.5 to 5 times smoother inputs, but could not overcome the unforeseen boulder whereas all subjects could, showcasing the importance of having humans and automation complement each other. Furthermore, a high quality fit (VAF > 70%) was found between the boom inputs of the subjects and the controller in the well-trained easy stage, indicating human-like control.
Master thesis
(2017)
-
Sabyasachi Neogi, Jan-Willem van Wingerden, Wouter Engels, Jens Kober, Sebastiaan Mulders
The energy demand in current times has increased greatly in last few years. This increasing demand calls for a sustainable and clean energy resource that would reduce the load on non-renewable resources. Wind energy is a renewable resource which is harnessed by mankind from an ancient era. So as to meet this increasing energy demands, innovation in the field of wind energy is required.
A wind turbine generates electricity but achieving optimal power is a difficult task. In this case, the optimal power is defined as maximum power produced but at the constraints that the fatigue loading of the wind turbine structure should be as minimum as possible. Also, the wind turbine parameters such as rotor speed and pitch activity should be in a safe operational region. The problem in controlling wind turbines is that they work in a highly uncertain environment where managing so many factors at the same time are difficult. Optimal control of wind turbine has helped in achieving maximum power with-in safe working limits. Due to high uncertainty in the operating conditions of a wind turbine, it is quite a daunting task to find optimal gains for a wind turbine controller.
This thesis focuses on achieving optimal gain parameters for wind turbine controller by using an algorithm from machine learning community. In this thesis the problem is formulated as a supervised learning problem where input-output mapping has to be predicted. For this purpose, GPRT is used. The reason behind using GPRT is it takes fewer number of measurements to give good prediction compared to others. The property of GPRT where it deals with uncertain and non linear data with ease, making it a good choice for predicting wind turbine controller gains.
The second part of this thesis contains optimisation of the surrogate model achieved by performing regression. The optimisation is done by Monte Carlo Maximum distribution and improved results were generated by applying sequential sampling to this algorithm. This helps us to get a likelihood of optimal gains where the wind turbine gives out rated power with minimal fatigue loads, pitch activity and least deviation of rotor speed from rated.
The results obtained from the likelihood was tested for different operational wind speed and also tested for Extreme Operating Gusts as part of disturbance rejection and compared to current parameter used.
The comparison shows considerable improvement in the fatigue loads and pitch activity with having improvement in power production. In second case study, more parameters were predicted and optimised using the same algorithm so that the potential of this algorithm can be estimated. This was also performed successfully which proves that this technique can successfully be used to solve higher dimensional problems of wind turbine control. ...
A wind turbine generates electricity but achieving optimal power is a difficult task. In this case, the optimal power is defined as maximum power produced but at the constraints that the fatigue loading of the wind turbine structure should be as minimum as possible. Also, the wind turbine parameters such as rotor speed and pitch activity should be in a safe operational region. The problem in controlling wind turbines is that they work in a highly uncertain environment where managing so many factors at the same time are difficult. Optimal control of wind turbine has helped in achieving maximum power with-in safe working limits. Due to high uncertainty in the operating conditions of a wind turbine, it is quite a daunting task to find optimal gains for a wind turbine controller.
This thesis focuses on achieving optimal gain parameters for wind turbine controller by using an algorithm from machine learning community. In this thesis the problem is formulated as a supervised learning problem where input-output mapping has to be predicted. For this purpose, GPRT is used. The reason behind using GPRT is it takes fewer number of measurements to give good prediction compared to others. The property of GPRT where it deals with uncertain and non linear data with ease, making it a good choice for predicting wind turbine controller gains.
The second part of this thesis contains optimisation of the surrogate model achieved by performing regression. The optimisation is done by Monte Carlo Maximum distribution and improved results were generated by applying sequential sampling to this algorithm. This helps us to get a likelihood of optimal gains where the wind turbine gives out rated power with minimal fatigue loads, pitch activity and least deviation of rotor speed from rated.
The results obtained from the likelihood was tested for different operational wind speed and also tested for Extreme Operating Gusts as part of disturbance rejection and compared to current parameter used.
The comparison shows considerable improvement in the fatigue loads and pitch activity with having improvement in power production. In second case study, more parameters were predicted and optimised using the same algorithm so that the potential of this algorithm can be estimated. This was also performed successfully which proves that this technique can successfully be used to solve higher dimensional problems of wind turbine control. ...
The energy demand in current times has increased greatly in last few years. This increasing demand calls for a sustainable and clean energy resource that would reduce the load on non-renewable resources. Wind energy is a renewable resource which is harnessed by mankind from an ancient era. So as to meet this increasing energy demands, innovation in the field of wind energy is required.
A wind turbine generates electricity but achieving optimal power is a difficult task. In this case, the optimal power is defined as maximum power produced but at the constraints that the fatigue loading of the wind turbine structure should be as minimum as possible. Also, the wind turbine parameters such as rotor speed and pitch activity should be in a safe operational region. The problem in controlling wind turbines is that they work in a highly uncertain environment where managing so many factors at the same time are difficult. Optimal control of wind turbine has helped in achieving maximum power with-in safe working limits. Due to high uncertainty in the operating conditions of a wind turbine, it is quite a daunting task to find optimal gains for a wind turbine controller.
This thesis focuses on achieving optimal gain parameters for wind turbine controller by using an algorithm from machine learning community. In this thesis the problem is formulated as a supervised learning problem where input-output mapping has to be predicted. For this purpose, GPRT is used. The reason behind using GPRT is it takes fewer number of measurements to give good prediction compared to others. The property of GPRT where it deals with uncertain and non linear data with ease, making it a good choice for predicting wind turbine controller gains.
The second part of this thesis contains optimisation of the surrogate model achieved by performing regression. The optimisation is done by Monte Carlo Maximum distribution and improved results were generated by applying sequential sampling to this algorithm. This helps us to get a likelihood of optimal gains where the wind turbine gives out rated power with minimal fatigue loads, pitch activity and least deviation of rotor speed from rated.
The results obtained from the likelihood was tested for different operational wind speed and also tested for Extreme Operating Gusts as part of disturbance rejection and compared to current parameter used.
The comparison shows considerable improvement in the fatigue loads and pitch activity with having improvement in power production. In second case study, more parameters were predicted and optimised using the same algorithm so that the potential of this algorithm can be estimated. This was also performed successfully which proves that this technique can successfully be used to solve higher dimensional problems of wind turbine control.
A wind turbine generates electricity but achieving optimal power is a difficult task. In this case, the optimal power is defined as maximum power produced but at the constraints that the fatigue loading of the wind turbine structure should be as minimum as possible. Also, the wind turbine parameters such as rotor speed and pitch activity should be in a safe operational region. The problem in controlling wind turbines is that they work in a highly uncertain environment where managing so many factors at the same time are difficult. Optimal control of wind turbine has helped in achieving maximum power with-in safe working limits. Due to high uncertainty in the operating conditions of a wind turbine, it is quite a daunting task to find optimal gains for a wind turbine controller.
This thesis focuses on achieving optimal gain parameters for wind turbine controller by using an algorithm from machine learning community. In this thesis the problem is formulated as a supervised learning problem where input-output mapping has to be predicted. For this purpose, GPRT is used. The reason behind using GPRT is it takes fewer number of measurements to give good prediction compared to others. The property of GPRT where it deals with uncertain and non linear data with ease, making it a good choice for predicting wind turbine controller gains.
The second part of this thesis contains optimisation of the surrogate model achieved by performing regression. The optimisation is done by Monte Carlo Maximum distribution and improved results were generated by applying sequential sampling to this algorithm. This helps us to get a likelihood of optimal gains where the wind turbine gives out rated power with minimal fatigue loads, pitch activity and least deviation of rotor speed from rated.
The results obtained from the likelihood was tested for different operational wind speed and also tested for Extreme Operating Gusts as part of disturbance rejection and compared to current parameter used.
The comparison shows considerable improvement in the fatigue loads and pitch activity with having improvement in power production. In second case study, more parameters were predicted and optimised using the same algorithm so that the potential of this algorithm can be estimated. This was also performed successfully which proves that this technique can successfully be used to solve higher dimensional problems of wind turbine control.