Nonlinear Adaptive Controller Design For Large Scale Wind Turbine Systems

Wind energy has drawn more and more attention around the world as a clean, renewable resource. This is partly driven by the increasing concern for the environment pollution caused by the traditional energy sources. Moreover, energy crisis and warning due to depletion of existing fossil fuel accentuate the need of deployment of renewable energy resources. Hence, wind energy conversion, due to its potential to provide an environmental friendly and economically competitive means of electricity, has become a fast-growing energy source in the global market. Also, wind energy is considered to be one of the important ways to solve energy crisis.Control technology is the core technology of wind energy conversion systems especially with the development of large-capacity offshore wind turbines. Due to aerodynamic uncertainties caused by random wind, as well as complexities of power electronics (e.g., nonlinear character-istics of thyristors), the model of wind energy conversion systems is difficult to determine. For example, changes in Reynolds number, paddle sediments, rain, etc. will cause a fluctuation in the power output. Meanwhile, aging, atmospheric conditions and other factors will lead to variation in the wind energy conversion systems. Therefore, the wind energy conversion system is a high-ly nonlinear uncertain system, which makes its control design a very challenging task. Based on the state-of-the-art study, this dissertation researches on nonlinear adaptive methods in large-scale wind power systems. The main work and contributions are summarized as follows1. A brief review of the background, overview, and related works on control of wind power systems is provided.2. Research on power capture control of variable-speed wind energy conversion systems. The control objective is to optimize the capture of wind energy by tracking the desired power output. Arbitrary steady-state performance is achieved in the sense that the tracking error is guaranteed to converge to any predefined small set. In addition, to maximize the wind energy capture, transient performance is enhanced such that the convergence rate can be larger than an arbitrary value, which further limits the maximum overshoot. First, an adaptive controller is designed for the case where known aerodynamic torque is assumed. Then, by utilizing an online approximator to estimate the uncertain aerodynamics, the need for the exact knowl-edge of the aerodynamic torque is waived to imitate the practical experience. With the aid of a novel output error transformation technique, both of the proposed controllers are capable of shaping the system performance arbitrarily on transient and steady-state stages. Meanwhile, it is also proved that all the signals in the closed-loop system are bounded via Lyapunov synthesis. Finally, the feasibility of the proposed controllers is demonstrated on an 1.5-MW three-blade wind turbine using the FAST code developed by the National Renewable Energy Laboratory (NREL).3. Research on power acquisition control of variable-speed wind energy conversion systems un-der inaccurate wind speed measurements. The control goal is to optimize the power capture from wind by tracking the maximum power curve. Firstly, the controller is designed for the case with known aerodynamic torque, which is a common assumption in many literatures. In this controller, the need for the exact knowledge of the system model is waived by using adaptive technologies. The chattering phenomenon in the generator torque, which can result in high mechanical stress, is avoided by adopting a modified robust term. Then, by utilizing an online approximator to learn an auxiliary term induced by the uncertain aerodynamics, the need for the exact knowledge of the aerodynamic torque is waived. Both of the proposed controllers are capable of providing good performance under inaccurate wind speed mea-surements. The control objective is obtained in the sense that the tracking error is guaranteed to converge to an arbitrarily small set. The results are theoretically proved, and verified using simulations.4. Research on a novel power control strategy for variable speed wind turbines equipped with Doubly Fed Induction Generators (DFIG). The control objective is to optimize the extract-ed power from wind while regulating the stator reactive power to meet grid requirements. Firstly, in order to optimize the extracted power, an adaptive control technique is designed to drive the electromagnetic torque to follow its reference generated by maximum power point tracking (MPPT) algorithm. Subsequently, aiming at satisfying reactive power requirements on the grid side, an adaptive reactive power controller is proposed to manipulate the stator reactive power to follow a given desired reactive power determined by the grid. Compared with most existing studies, we are capable of quantifying and further guaranteeing the system performance on both transient and steady state stages. All signals in the closed-loop system are proved to be bounded via standard Lyapunov synthesis. Finally, the effectiveness of the proposed scheme is validated via simulations.5. Research on the pitch control of wind turbine operating in high speed region. In this region, the focused model is a nonaffine nonlinear model, which has no linear variable as control input. Therefore, control of nonaffine nonlinear systems is a very challenging task. In this thesis, we firstly propose a nonlinear pitch controller using an augmented filtered error and neural networks. Thereafter, by using an output transformation and system transformation techniques, another nonlinear pitch controller is presented, which can overcome the diffi-culties caused by nonaffine uncertain properties, and guarantee that the rotor angular speed stays within predefined bounds.The conclusions are drawn with future work at the end of the dissertation.