J.A. Aviles Cedeño
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>
2 records found
1
Master thesis
(2026)
-
N. Rodriguez-Lafuente, J.L. Rueda Torres, J.A. Aviles-Cedeño, Sam Frohn, F.A. Muñoz Muñoz, A. Shekhar, P.P. Vergara Barrios
As the European grid transitions towards renewable energy targets, new system stability phenomena are being observed through the electrical power system. The share of renewable generation is growing quickly as synchronous machine-based power plants are phased out. Unlike traditional synchronous machine plants, these Inverter-Based Resources (IBRs) do not provide inherent synchronous mechanical inertia. Consequently, the overall system inertia is decreasing, posing a threat to the frequency stability of the system. This trend is observed worldwide and addressed by organisations such as ENTSO-E through the introduction of future minimum inertia thresholds.
Grid-forming (GFM) technologies have emerged as a solution, providing an immediate active power support after a disturbance, via control methods. These technologies enable IBRs to contribute to frequency stability and mitigate the Rate of Change of Frequency (RoCoF).
The share of distributed energy resources has also grown, with utility-scale renewable Power Park Modules (PPMs) connected to the Medium Voltage (MV) network. These PPMs could help unlock further active power frequency support from the MV network. However, whether this support effectively supports the transmission system remains insufficiently explored.
This thesis assesses the effectiveness of synthetic inertia provided to the transmission network by GFM-controlled IBRs connected at the MV level. The study explores the metrics to assess the effectiveness of the synthetic inertia at the MV-HV interface, investigates the technical constraints of the provision from the distribution-connected resources and offers mitigation measures. Additionally, the comparison of the synthetic inertia support from the HV and MV-connected GFM assets is done.
Firstly, using DIgSILENT PowerFactory, a test benchmark is developed, where the response of a storage PPM with grid-forming controls is evaluated under different disturbances, in order to tune the GFM dynamic models and ensure compliance with grid code requirements. To test the synthetic inertia support of PPMs in a realistic model, a combined transmission and distribution system model is developed from an initial model, which has been adjusted to represent the Dutch grid characteristics. Multiple scenarios and disturbances are simulated with the GFM PPM connected at various locations within the MV network to explore the limits of synthetic inertia support.
The results demonstrate that for a frequency disturbance, the RoCoF and the frequency nadir/zenith improve regardless of the PPM connection point. However, the magnitude of the improvement is dependent on the available active power headroom and the current capabilities of the PPM. A critical limitation was observed regarding the parallel provision of active and reactive power when the PPM is connected to the MV grid. For connection points far from the point of interconnection, voltage stability became more important, as MV buses are more sensitive to active/reactive power injection. During large frequency disturbances, the PPM's active power and reactive power demand grew, which led to current limits being hit, as the PPM sustains the active and reactive current. This suggested a need to enhance voltage and reactive power control or alternative voltage stability methods to support the voltage during frequency disturbances for distribution-connected grid-forming units. ...
Grid-forming (GFM) technologies have emerged as a solution, providing an immediate active power support after a disturbance, via control methods. These technologies enable IBRs to contribute to frequency stability and mitigate the Rate of Change of Frequency (RoCoF).
The share of distributed energy resources has also grown, with utility-scale renewable Power Park Modules (PPMs) connected to the Medium Voltage (MV) network. These PPMs could help unlock further active power frequency support from the MV network. However, whether this support effectively supports the transmission system remains insufficiently explored.
This thesis assesses the effectiveness of synthetic inertia provided to the transmission network by GFM-controlled IBRs connected at the MV level. The study explores the metrics to assess the effectiveness of the synthetic inertia at the MV-HV interface, investigates the technical constraints of the provision from the distribution-connected resources and offers mitigation measures. Additionally, the comparison of the synthetic inertia support from the HV and MV-connected GFM assets is done.
Firstly, using DIgSILENT PowerFactory, a test benchmark is developed, where the response of a storage PPM with grid-forming controls is evaluated under different disturbances, in order to tune the GFM dynamic models and ensure compliance with grid code requirements. To test the synthetic inertia support of PPMs in a realistic model, a combined transmission and distribution system model is developed from an initial model, which has been adjusted to represent the Dutch grid characteristics. Multiple scenarios and disturbances are simulated with the GFM PPM connected at various locations within the MV network to explore the limits of synthetic inertia support.
The results demonstrate that for a frequency disturbance, the RoCoF and the frequency nadir/zenith improve regardless of the PPM connection point. However, the magnitude of the improvement is dependent on the available active power headroom and the current capabilities of the PPM. A critical limitation was observed regarding the parallel provision of active and reactive power when the PPM is connected to the MV grid. For connection points far from the point of interconnection, voltage stability became more important, as MV buses are more sensitive to active/reactive power injection. During large frequency disturbances, the PPM's active power and reactive power demand grew, which led to current limits being hit, as the PPM sustains the active and reactive current. This suggested a need to enhance voltage and reactive power control or alternative voltage stability methods to support the voltage during frequency disturbances for distribution-connected grid-forming units. ...
As the European grid transitions towards renewable energy targets, new system stability phenomena are being observed through the electrical power system. The share of renewable generation is growing quickly as synchronous machine-based power plants are phased out. Unlike traditional synchronous machine plants, these Inverter-Based Resources (IBRs) do not provide inherent synchronous mechanical inertia. Consequently, the overall system inertia is decreasing, posing a threat to the frequency stability of the system. This trend is observed worldwide and addressed by organisations such as ENTSO-E through the introduction of future minimum inertia thresholds.
Grid-forming (GFM) technologies have emerged as a solution, providing an immediate active power support after a disturbance, via control methods. These technologies enable IBRs to contribute to frequency stability and mitigate the Rate of Change of Frequency (RoCoF).
The share of distributed energy resources has also grown, with utility-scale renewable Power Park Modules (PPMs) connected to the Medium Voltage (MV) network. These PPMs could help unlock further active power frequency support from the MV network. However, whether this support effectively supports the transmission system remains insufficiently explored.
This thesis assesses the effectiveness of synthetic inertia provided to the transmission network by GFM-controlled IBRs connected at the MV level. The study explores the metrics to assess the effectiveness of the synthetic inertia at the MV-HV interface, investigates the technical constraints of the provision from the distribution-connected resources and offers mitigation measures. Additionally, the comparison of the synthetic inertia support from the HV and MV-connected GFM assets is done.
Firstly, using DIgSILENT PowerFactory, a test benchmark is developed, where the response of a storage PPM with grid-forming controls is evaluated under different disturbances, in order to tune the GFM dynamic models and ensure compliance with grid code requirements. To test the synthetic inertia support of PPMs in a realistic model, a combined transmission and distribution system model is developed from an initial model, which has been adjusted to represent the Dutch grid characteristics. Multiple scenarios and disturbances are simulated with the GFM PPM connected at various locations within the MV network to explore the limits of synthetic inertia support.
The results demonstrate that for a frequency disturbance, the RoCoF and the frequency nadir/zenith improve regardless of the PPM connection point. However, the magnitude of the improvement is dependent on the available active power headroom and the current capabilities of the PPM. A critical limitation was observed regarding the parallel provision of active and reactive power when the PPM is connected to the MV grid. For connection points far from the point of interconnection, voltage stability became more important, as MV buses are more sensitive to active/reactive power injection. During large frequency disturbances, the PPM's active power and reactive power demand grew, which led to current limits being hit, as the PPM sustains the active and reactive current. This suggested a need to enhance voltage and reactive power control or alternative voltage stability methods to support the voltage during frequency disturbances for distribution-connected grid-forming units.
Grid-forming (GFM) technologies have emerged as a solution, providing an immediate active power support after a disturbance, via control methods. These technologies enable IBRs to contribute to frequency stability and mitigate the Rate of Change of Frequency (RoCoF).
The share of distributed energy resources has also grown, with utility-scale renewable Power Park Modules (PPMs) connected to the Medium Voltage (MV) network. These PPMs could help unlock further active power frequency support from the MV network. However, whether this support effectively supports the transmission system remains insufficiently explored.
This thesis assesses the effectiveness of synthetic inertia provided to the transmission network by GFM-controlled IBRs connected at the MV level. The study explores the metrics to assess the effectiveness of the synthetic inertia at the MV-HV interface, investigates the technical constraints of the provision from the distribution-connected resources and offers mitigation measures. Additionally, the comparison of the synthetic inertia support from the HV and MV-connected GFM assets is done.
Firstly, using DIgSILENT PowerFactory, a test benchmark is developed, where the response of a storage PPM with grid-forming controls is evaluated under different disturbances, in order to tune the GFM dynamic models and ensure compliance with grid code requirements. To test the synthetic inertia support of PPMs in a realistic model, a combined transmission and distribution system model is developed from an initial model, which has been adjusted to represent the Dutch grid characteristics. Multiple scenarios and disturbances are simulated with the GFM PPM connected at various locations within the MV network to explore the limits of synthetic inertia support.
The results demonstrate that for a frequency disturbance, the RoCoF and the frequency nadir/zenith improve regardless of the PPM connection point. However, the magnitude of the improvement is dependent on the available active power headroom and the current capabilities of the PPM. A critical limitation was observed regarding the parallel provision of active and reactive power when the PPM is connected to the MV grid. For connection points far from the point of interconnection, voltage stability became more important, as MV buses are more sensitive to active/reactive power injection. During large frequency disturbances, the PPM's active power and reactive power demand grew, which led to current limits being hit, as the PPM sustains the active and reactive current. This suggested a need to enhance voltage and reactive power control or alternative voltage stability methods to support the voltage during frequency disturbances for distribution-connected grid-forming units.
Master thesis
(2026)
-
A. Akaouche, J.L. Rueda Torres, Zian Qin, P.A. Procel Moya, J.A. Aviles Cedeño
The growing penetration of converter-interfaced generation, flexible demand, and prosumer assets is tightening the physical coupling between transmission and distribution networks. As a result, operators increasingly benefit from coordinated transmission-distribution (TSO-DSO) state estimation (SE) that provides a consistent view of bus voltages across voltage levels. In practice, however, TSOs and DSOs operate under strict data-governance constraints: measurements, historical archives, and detailed network information cannot be freely pooled across organizational boundaries. This limits the feasibility of centralized, data-hungry learning pipelines and complicates coordinated SE at scale.
This thesis proposes a federated learning (FL) framework for coordinated SE of an integrated TSO-DSO system. FL enables multiple operators to collaboratively train a shared neural estimator without exchanging raw data: each participant performs local training on its own measurement/state pairs and only model updates are aggregated on a central server. To evaluate the approach in a controlled setting, a steady-state digital twin of the Dutch transmission system is extended with three reduced-order distribution feeders (representing three DSO regions). A large synthetic dataset is generated through power-flow simulations, and measurement uncertainty is modeled via bounded uniform noise injected into the input features. The estimator predicts bus voltages in real and imaginary components rather than magnitude-angle form, avoiding phase-wrap discontinuities and improving numerical stability during training.
We establish a strong centralized baseline and study the effect of a warm-start strategy, where the model is initialized from a pre-trained checkpoint to accelerate convergence. We then implement a so called masked-target FL to reflect partial observability and data locality: each client trains on a restricted subset of inputs and supervises only its own target voltages. Three aggregation strategies are compared: FedAvg, FedProx, and FedAdam. This comparison will be in terms of convergence and global test accuracy over the full system state. The results show that FL achieves stable convergence and preserves data locality, while a substantial gap remains between federated and centralized test accuracy under partial labels and heterogeneous client distributions. Among the FL strategies, FedAdam provides the best global performance, indicating that server-side adaptive optimization can mitigate client drift in this setting.
Overall, this work demonstrates the feasibility of FL for multi-operator SE training in TSO-DSO systems and provides an end-to-end, reproducible benchmark covering digital-twin modeling, dataset generation, centralized baselines, and federated training. The findings also highlight key challenges, notably non-IID data, unequal target cardinality, and evaluation mismatch between masked validation and global testing, and motivate future work on personalization, hybrid neural WLS refinement, and privacy-preserving inference for real-time digital-twin applications. ...
This thesis proposes a federated learning (FL) framework for coordinated SE of an integrated TSO-DSO system. FL enables multiple operators to collaboratively train a shared neural estimator without exchanging raw data: each participant performs local training on its own measurement/state pairs and only model updates are aggregated on a central server. To evaluate the approach in a controlled setting, a steady-state digital twin of the Dutch transmission system is extended with three reduced-order distribution feeders (representing three DSO regions). A large synthetic dataset is generated through power-flow simulations, and measurement uncertainty is modeled via bounded uniform noise injected into the input features. The estimator predicts bus voltages in real and imaginary components rather than magnitude-angle form, avoiding phase-wrap discontinuities and improving numerical stability during training.
We establish a strong centralized baseline and study the effect of a warm-start strategy, where the model is initialized from a pre-trained checkpoint to accelerate convergence. We then implement a so called masked-target FL to reflect partial observability and data locality: each client trains on a restricted subset of inputs and supervises only its own target voltages. Three aggregation strategies are compared: FedAvg, FedProx, and FedAdam. This comparison will be in terms of convergence and global test accuracy over the full system state. The results show that FL achieves stable convergence and preserves data locality, while a substantial gap remains between federated and centralized test accuracy under partial labels and heterogeneous client distributions. Among the FL strategies, FedAdam provides the best global performance, indicating that server-side adaptive optimization can mitigate client drift in this setting.
Overall, this work demonstrates the feasibility of FL for multi-operator SE training in TSO-DSO systems and provides an end-to-end, reproducible benchmark covering digital-twin modeling, dataset generation, centralized baselines, and federated training. The findings also highlight key challenges, notably non-IID data, unequal target cardinality, and evaluation mismatch between masked validation and global testing, and motivate future work on personalization, hybrid neural WLS refinement, and privacy-preserving inference for real-time digital-twin applications. ...
The growing penetration of converter-interfaced generation, flexible demand, and prosumer assets is tightening the physical coupling between transmission and distribution networks. As a result, operators increasingly benefit from coordinated transmission-distribution (TSO-DSO) state estimation (SE) that provides a consistent view of bus voltages across voltage levels. In practice, however, TSOs and DSOs operate under strict data-governance constraints: measurements, historical archives, and detailed network information cannot be freely pooled across organizational boundaries. This limits the feasibility of centralized, data-hungry learning pipelines and complicates coordinated SE at scale.
This thesis proposes a federated learning (FL) framework for coordinated SE of an integrated TSO-DSO system. FL enables multiple operators to collaboratively train a shared neural estimator without exchanging raw data: each participant performs local training on its own measurement/state pairs and only model updates are aggregated on a central server. To evaluate the approach in a controlled setting, a steady-state digital twin of the Dutch transmission system is extended with three reduced-order distribution feeders (representing three DSO regions). A large synthetic dataset is generated through power-flow simulations, and measurement uncertainty is modeled via bounded uniform noise injected into the input features. The estimator predicts bus voltages in real and imaginary components rather than magnitude-angle form, avoiding phase-wrap discontinuities and improving numerical stability during training.
We establish a strong centralized baseline and study the effect of a warm-start strategy, where the model is initialized from a pre-trained checkpoint to accelerate convergence. We then implement a so called masked-target FL to reflect partial observability and data locality: each client trains on a restricted subset of inputs and supervises only its own target voltages. Three aggregation strategies are compared: FedAvg, FedProx, and FedAdam. This comparison will be in terms of convergence and global test accuracy over the full system state. The results show that FL achieves stable convergence and preserves data locality, while a substantial gap remains between federated and centralized test accuracy under partial labels and heterogeneous client distributions. Among the FL strategies, FedAdam provides the best global performance, indicating that server-side adaptive optimization can mitigate client drift in this setting.
Overall, this work demonstrates the feasibility of FL for multi-operator SE training in TSO-DSO systems and provides an end-to-end, reproducible benchmark covering digital-twin modeling, dataset generation, centralized baselines, and federated training. The findings also highlight key challenges, notably non-IID data, unequal target cardinality, and evaluation mismatch between masked validation and global testing, and motivate future work on personalization, hybrid neural WLS refinement, and privacy-preserving inference for real-time digital-twin applications.
This thesis proposes a federated learning (FL) framework for coordinated SE of an integrated TSO-DSO system. FL enables multiple operators to collaboratively train a shared neural estimator without exchanging raw data: each participant performs local training on its own measurement/state pairs and only model updates are aggregated on a central server. To evaluate the approach in a controlled setting, a steady-state digital twin of the Dutch transmission system is extended with three reduced-order distribution feeders (representing three DSO regions). A large synthetic dataset is generated through power-flow simulations, and measurement uncertainty is modeled via bounded uniform noise injected into the input features. The estimator predicts bus voltages in real and imaginary components rather than magnitude-angle form, avoiding phase-wrap discontinuities and improving numerical stability during training.
We establish a strong centralized baseline and study the effect of a warm-start strategy, where the model is initialized from a pre-trained checkpoint to accelerate convergence. We then implement a so called masked-target FL to reflect partial observability and data locality: each client trains on a restricted subset of inputs and supervises only its own target voltages. Three aggregation strategies are compared: FedAvg, FedProx, and FedAdam. This comparison will be in terms of convergence and global test accuracy over the full system state. The results show that FL achieves stable convergence and preserves data locality, while a substantial gap remains between federated and centralized test accuracy under partial labels and heterogeneous client distributions. Among the FL strategies, FedAdam provides the best global performance, indicating that server-side adaptive optimization can mitigate client drift in this setting.
Overall, this work demonstrates the feasibility of FL for multi-operator SE training in TSO-DSO systems and provides an end-to-end, reproducible benchmark covering digital-twin modeling, dataset generation, centralized baselines, and federated training. The findings also highlight key challenges, notably non-IID data, unequal target cardinality, and evaluation mismatch between masked validation and global testing, and motivate future work on personalization, hybrid neural WLS refinement, and privacy-preserving inference for real-time digital-twin applications.