Robust Adaptive And Fault-tolerant Control Of Large-scale Wind Turbines

Wind power is regarded as a new energy type, with larger reliability, development-scale and cost efficiency. However, with the rapid growth of unit capacity and total installed capacity, it existes some conflicts between the development and the strict demand of being highly efficient, reliable and grid-friendly. This thesis adopts advanced control algorithms to address this problem from the following aspects:Considering the complicated and strongly nonlinear coupled characteristics of large scale wind turbine systems, the robust adaptive control strategies are developed to ensure the system safety, reliability, and efficiency. First, for the maximum power point tracking (MPPT) control, a novel neuro-adaptive control strategy combining radial basis function (RBF) and neural networks (NNs) with adaptive techniques is developed to accommodate the system uncertain and external disturbance. Second, to deal with the generator faults, a robust adaptive fault-tolerant controller is designed with corresponding torque constraints. Third, to address the pitch control problem, an L\ output feedback controller is proposed without wind speed estimation, ensuring that the generator speed tracks the reference trajectory with robustness to uncertain parameters and time-varying disturbances. It is worth noting that the proposed controllers do not depend on the system parameters and fault information. Meanwhile, compared with most existing methods, they are less demanding for on-line computation, more user-friendly in control design, and easier for real-time implementation.One the other hand, the thesis investigates a novel grid-friendly wind turbine, called "front-end variable speed wind turbine" with an embedded variable ratio gearbox. Corresponding control methods are developed to improve the system reliability and efficiency. First, the dynamic and operation scheme is studied for the novel grid-friendly wind turbine based on the GH Bladed software. Second, a model independent speed control algorithm for the permanent magnet synchronous motor is developed based on the neuro-adaptive backstepping approach, demonstrating higher speed tracking precision in the presence of parameter uncertainties and external load disturbances.This study provides some valuable results for the controller design and load mitigation of large-scale wind turbines, and helps enhance the operation performance, reliability and grid-friendness of existing and future large-scale wind turbine systems, and thus might be useful both in theoretical guidance and technical reference.