The Research Of Data-based Control For Wind Energy Conversation System

With the sustained growth of economy and the fast development of society, the progress of mankind is gradually restricted by shortage of traditional energy and environmental pollution. Wind energy is an inexhaustible, clean and renewable new energy, to develop and use renewable energy resources such as wind power, which became the strategic choice for human solving the problems of existence. So the research of Wind power generation system obtained the extensive attention of scholars, especially the control technology research. Wind energy conversion system is a high order, nonlinear, strong coupling, time varying and complex system. It is very difficult to establish the accurate mathematical model. Meanwhile, it may consume large amount of time and there are inevitably unmodeled dynamics, intrinsic error and uncertainty. Then the data-driven control theory is introduced in wind energy conversion system (WECS).In order to ensure maximum energy capture below rated wind speed, a dynamic time-varying linear system is used to reasonably replace the wind energy conversion system, and the gradient-like vector of it can be obtained through the projection estimation algorithm or the least square estimation algorithm by using input and output data of the wind energy conversion system, which is adjusted on-line to design the adaptive controller. Simulation results indicate that below the rated wind, the data-based adaptive control can effectively implement maximum energy capture. Then a data-based adaptive quasi-sliding mode controller is got by combining sliding mode variable structure control with it. Simulation results verify the effectiveness of the data-based adaptive control and the the priority of the data-based adaptive quasi-sliding mode control.When wind speed is under the rated speed, the maximum wind energy is obtained by controlling the generator rotor speed to track the best power coefficient curve. When the wind speed is above the rated value, pitch angle and electromagnetic torque of the wind energy conversion system are controlled simultaneously to provide optimal regulation of the generated power and the generator speed to their rated values. With the multivariable control strategy above-rated, the Markov parameters can be obtained and a data-based controller state observer can be constructed by using input and output data of the wind energy conversion system, then an optimal feedback controller is designed through the closed form of difference Riccati equation. Simulation results indicate the data-based optimal control can effectively implement maximum energy capture when wind speed is below-rated and mantain the rated power output when wind speed is above-rated.