Showing papers by "Ron J. Patton published in 2013"

Active fault tolerant control for nonlinear systems with simultaneous actuator and sensor faults

Abstract: The goal of this paper is to describe a novel fault tolerant tracking control (FTTC) strategy based on robust fault estimation and compensation of simultaneous actuator and sensor faults. Within the framework of fault tolerant control (FTC) the challenge is to develop an FTTC design strategy for nonlinear systems to tolerate simultaneous actuator and sensor faults that have bounded first time derivatives. The main contribution of this paper is the proposal of a new architecture based on a combination of actuator and sensor Takagi-Sugeno (T-S) proportional state estimators augmented with proportional and integral feedback (PPI) fault estimators together with a T-S dynamic output feedback control (TSDOFC) capable of time-varying reference tracking. Within this architecture the design freedom for each of the T-S estimators and the control system are available separately with an important consequence on robust L 2 norm fault estimation and robust L 2 norm closed-loop tracking performance. The FTTC strategy is illustrated using a nonlinear inverted pendulum example with time-varying tracking of a moving linear position reference.

Active FTC for hydraulic pitch system for an off-shore wind turbine

Abstract: This paper describes the development and testing of the implementation of a linear parameter varying (LPV) fault tolerant control (FTC) approach to hydraulic pitch actuation for an off-shore wind turbine (OWT) system with independent pitch actuators. The study involves two separate actuator fault scenarios. In the first scenario the pitch actuator position sensor signal is considered to be stuck at a fixed value. In this case, a fault hiding approach is used to compensate for this fault in the control system, as a form of sensor FTC scheme with fault estimation. The second fault scenario is caused either by a drop in pressure in the hydraulic supply system or the high air content in the oil. In this case, the fault-free pitch actuator systems are both utilized to synchronize all three pitch systems in order to prevent further damage caused by unbalanced structural loads.

Robust actuator multiplicative fault estimation with unknown input decoupling for a wind turbine system

Abstract: A robust actuator multiplicative fault estimation approach is designed by combining the unknown input-decoupling principle and H∞ optimization. The fault is constructed as an augmented system state and estimated by the proposed observer. In this observer, the unknown input-decoupling principle is used to decouple the unknown input (the wind force) effect from the fault estimation signal. From the practical point of view, the actuator measurement sensors are corrupted by the sensor measurement noise. An H∞ approach is used based on the unknown input-decoupling observer to minimize the effect of this sensor noise on the fault estimation signal. In order to estimate the multiplicative fault, a fault model modification is used to reform the multiplicative fault into an additive fault representation by which the fault can be estimated. Finally, the proposed design is used to estimate the hydraulic leakage fault occurring in a wind turbine pitch actuator system based on a non-linear benchmark model.

Reconfigurable flight control using feedback linearization with online Neural Network adaption

Abstract: This work suggests a design approach building on past results in the area of adaptive Neural Network (NN) based reconfigurable flight control which has been successfully utilized for a variety of aerospace applications, while incorporating recent advances in the areas of output feedback and NN adaptation algorithms. The paradigm is based on feedback linearization and synthesis of a fixed-gain dynamic compensator, whilst incorporating a neural network by using concurrent update information to compensate for model uncertainties, system faults and external disturbances. The concurrent learning networks update law is simplified and restructured for better and easy use during applications. The stability and reconfigurable ability of the improved control system based on the concurrent learning algorithm is validated with simulation performances of the nonlinear Machan UAV.

A model-based predictive control for FTC for wind turbine wind speed sensor fault

Abstract: This paper proposes an approach to fault tolerant control (FTC) of variable-speed wind turbine subject to wind speed sensor faults when the turbine is operating below rated wind speed. Both hardware and analytical redundancy are utilized in this approach. The least-Squares Support Vector Machine (LS-SVM) is proposed as the Fault Detection and Isolation (FDI) unit for the wind speed sensor fault. Single and multiple sensor faults are both considered in this strategy. LS-SVM is also used to estimate the effective wind speed (EWS), which is an unmeasurable signal. Based on the FDI unit, good FTC performance is achieved by identifying the faulty sensors and switching to the healthy sensor or the estimated EWS to provide the required controller reference signal. A robust MPC controller is then designed in order to consider the uncertainty due to error of EWS estimation and respect the physical system constraints.

Design and evaluation of FTFC scheme for sensor loss detection based on RECOVER benchmark

Abstract: The work proposes a study on fault tolerant flight control of the GARTEUR RECOVER benchmark aircraft model. Instead of using control surface failure scenarios, a sensor loss detection scenario is defined in this benchmark for the first time, to demonstrate the proposed fault tolerant flight control capability. Therefore, the model-based fault tolerant flight control is well posed for this failure scenario, which is different from traditional sensor failure schemes relying on sensor redundancy and consolidation detection. The proposed fault tolerant flight control scheme is simulated and evaluated in various testing manoeuvres during a landing process. In this scheme, a fault estimator based upon ℋ-/ℋ∞ robustness index is utilised to estimate the loss detection. A sensor fault compensator is also included to restore the fault-free sensor measurement on-line.

Static output feedback adaptive integral sliding control for interconnected nonlinear systems

Abstract: The problem of static output feedback decentralization of uncertain inter-connected systems is concerned with the goal of decoupling Lipschitz non-linear systems into individual “decentralized” subsystems, satisfying stability and fault-tolerance objectives. This work proposes a strategy for decentralized control in which each subsystem uses a static output structure invoking an approach to separation principle recovery. The approach is based on Adaptive Integral Sliding Model Control (AISMC) with careful consideration of both matched and unmatched uncertainties arising from inter-connections and disturbances. The proposed design strategy for the static output feedback and uncertainty de-coupling designs involves an LMI procedure to solve a non-convex problem. An example of 3 unstable inter-connected non-linear systems is used to illustrate the power of the approach.