Fault-Tolerant Flight Control (FTFC) systems reconfigure aircraft flight control laws to help restore a controlled flight and preserve acceptable performance under systemfaults. A typical approach to achieve fault tolerance is using inertial sensor measurementswhich can provide robustness to abrupt changes in the aircraft dynamics. This method does not rely on an accurate and full aerodynamic model, but only requires an estimate of the control effectiveness and an additional reconfiguration mechanism to the control laws such as switching, model following, matching, and adaptive compensation. With the inertial sensor being at the very heart of the control laws, to have proven and well-understood sensors is essential to increase the operational performance of a post-failure aircraft. Recent developments lead to FTFC systems that require angular acceleration feedback, which with the current conventional Inertial Measurement Unit (IMU) is obtained by taking the first derivative of the angular rates measured by gyroscopes. Because differentiation commonly intensifies noise and incurs a delay, a direct measurement of angular acceleration is anticipated to enhance the quality of the feedback signal. However, angular accelerometers are currently not part of the commercial aircraft system, which prompts the need to understand their characteristics. This thesis aims to investigate how an angular accelerometer can be evaluated and calibrated in a systematic fashion. @en

The calibration of Angular Accelerometers requires a controlled test with a known acceleration profile. However, current turn-tables have been designed primarily for generating a constant rotational velocity. To generate a profile with varying angular acceleration we propose using constant rotational velocity on the two axes of a 2-axes motion simulator. The proposed sequence was developed to obtain the required rotational acceleration signal quality using only rotational velocity input. The identified pattern is applied to an envelope of test conditions, resulting in test matrix. This method provides an alternative means to generate inputs that can be used to calibrate angular accelerometers using calibration hardware that is not primarily designed to provide accurate acceleration inputs. @en

A characterization of an angular accelerometer sensor is performed with a frequency- response experiment. To assess angular accelerometer dynamic properties, a position-based calibration table is utilized to provide the excitation input. The angular accelerometer specifications and calibration table limitations dfine a constrained test envelope for this particular setup. Measurement are obtained for a number of constant accelerations and varied with frequency. Analysis of the input-output relation in the frequency-domain leads to a frequency-response function model of the angular accelerometer. Time-domain data are used to validate the frequency-response model. A fourth-order structure is then preferred as the transfer-function for the 400 deg=s acceleration, with a 99:02 %fit. Besides representing the angular accelerometer dynamics, the angular accelerometer model is es- sential for the design of fault-tolerant light control system. @en

@en

@en

Angular acceleration feedback is a powerful concept when designing a model-free, state-of-the-art Fault-Tolerant Flight Control (FTFC) system. Such feedback is currently being obtained by differentiating the rate gyro signals [1], which amplifies the noise component and introduces delay [2]. Considering feedback is a valuable tool to adjust the performance and to ultimately meet the desired result, the feedback signal accuracy will determine the system’s successful control actions. To incorporate the angular acceleration feedback in the FTFC system design, analysis and simulation, a proper angular accelerometer model is of substantial concern. Creating a controller based on an incorrect angular accelerometer, for example, could result in the control not meeting the desired control variables. Therefore, establishing a proper angular accelerometer system identification method is crucial to achieving the correct model. The frequency response data generates the current angular accelerometer model in a transfer function form [3]. Frequency domain data provides the required accuracy to calculate the Frequency Response Function (FRF) in this approach. Specifying the sensor bandwidth as well as recognising the system’s phase are the additional advantages of this method. The Auto-Regressive with exogenous terms (ARX) model is employed as the first stage estimation due to its efficiency in estimating the polynomial because it is the result of solving linear regression equations in analytic form. Moreover, the result is distinctive; it always fulfils the global minimum of the loss function. The ARX model, therefore, is favoured, especially when the model order is high. @en

Angular acceleration feedback is a powerful concept when designing a model-free, state-of-the-art Fault-Tolerant Flight Control (FTFC) system. Such feedback is currently being obtained by differentiating the rate gyro signals [1], which amplifies the noise component and introduces delay [2]. Considering feedback is a valuable tool to adjust the performance and to ultimately meet the desired result, the feedback signal accuracy will determine the system’s successful control actions. To incorporate the angular acceleration feedback in the FTFC system design, analysis and simulation, a proper angular accelerometer model is of substantial concern. Creating a controller based on an incorrect angular accelerometer, for example, could result in the control not meeting the desired control variables. Therefore, establishing a proper angular accelerometer system identification method is crucial to achieving the correct model. The frequency response data generates the current angular accelerometer model in a transfer function form [3]. Frequency domain data provides the required accuracy to calculate the Frequency Response Function (FRF) in this approach. Specifying the sensor bandwidth as well as recognising the system’s phase are the additional advantages of this method. The Auto-Regressive with exogenous terms (ARX) model is employed as the first stage estimation due to its efficiency in estimating the polynomial because it is the result of solving linear regression equations in analytic form. Moreover, the result is distinctive; it always fulfils the global minimum of the loss function. The ARX model, therefore, is favoured, especially when the model order is high. @en