Research On Adaptive Pitch Control Of Megawatt-level Wind Turbine

With the intensification of global energy consumption and dwindling fossil fuels,environmental pollution is increasing.Facing with this situation,people’s sense of energy crisis and environmental protection awareness have been further strengthened,and countries all over the world have begun to attach importance to the development and utilization of new energy.Wind energy has become the focus of development in various countries because of its renewable,clean and pollution-free advantages.Wind power generation is an important form of wind energy utilization.With the enlargement of wind turbines,the internal structure of wind turbines has become more complex,and the requirements for control systems have gradually increased.Aiming at the problems of uncertain parameters and unknown interference in the wind turbine system,this study adopts an adaptive pitch control strategy to maintain the wind turbine speed and output power near the rated value.The main work of the thesis includes:1.Modeling and characteristic analysis of wind turbine systemFirstly,explain the working principle and internal structure of the wind turbine.Then based on the aerodynamics,a theoretical analysis of the characteristics of the wind turbine system and the force of the blades is carried out.Perform overall modeling of the wind turbine system and analyze the principle of variable pitch of the wind turbine system.Finally,the linearization test of the wind turbine system model is used as a reference for the subsequent design of adaptive variable pitch control strategy.2.Design of differential variable pitch control system based on integral sliding mode adaptive neural networkAiming at the problems of steady-state error and chattering in the output of wind turbines,a differential pitch controller with integral sliding mode adaptive neural network is designed.Firstly,in order to reduce steady-state error,using integral sliding surface design system switching function.For the uncertain parameters in the wind turbine system model,the RBF neural network is used to approximate the uncertain parameters.At the same time,considering the time-varying nature of the uncertain parameters,it is difficult for the system controller to adapt to the changes of the system,and it will be further combined with the adaptive law.Then,the self-adaptive law derived from the Lyapunov function adjusts the weights of the neural network online,so that when the wind turbine system encounters external changes,the system parameters can be adjusted in time to quickly restore the system to a stable state.At the same time,taking into account the system chattering problem caused by the system parameter adjustment stage,differential control is used to increase the nonlinear damping of the system to further reduce the system buffeting.Finally,simulation experiments show that,compared to the fuzzy integral sliding mode variable pitch control algorithm and the feedback linearization variable pitch control algorithm,the variable pitch control algorithm designed in this study has fast response speed,small chattering and high convergence.And the system output has good stability.3.Design of adaptive sliding mode variable pitch control system based on control input fuzzificationAiming at the problem of unknown parameters and internal and external disturbances in the wind turbine system model,a control strategy combining fuzzy logic and adaptive sliding mode is adopted.Firstly,in order to reduce the steady-state error of the system,this study use integral sliding mode control to design the system switching function.Then design the input and output of the fuzzy system.In order to further improve the adaptability of the fuzzy controller,this study adopts the adaptive law derived from the Lyapunov function to adjust the membership degree of the output fuzzy set in the fuzzy system online.At the same time,in order to improve the control accuracy of the system,the error between the fuzzy controller and the ideal controller is compensated by switching the controller.Among them,the switching gain in the switching controller is also adjusted online by an adaptive law.The simulation experiment shows that,compared with the other two variable pitch control algorithms,the variable pitch control algorithm designed in this study has less chattering and faster response speed when the wind speed and output power change during a sudden change in wind speed.It has high control accuracy and stability.4.Design of adaptive dynamic surface variable pitch control system based on extended observerConsidering that the wind turbine system has strong nonlinear characteristics such as uncertainty and hysteresis.Firstly,the expanded observer is used to estimate and observe the uncertain parameters and unknown interference in the wind turbine system model.Among them,the high gain error feedback used in the expanded observer makes the observer’s dynamics much higher than that of the original wind turbine system,and can provide feedback information to the system controller in time.Considering that there are certain limitations in the estimation of unknown parameters and interference in the extended observer,an adaptive dynamic surface controller is used for compensation.Among them,the neural network is used to approximate the estimated parameters in the expanded observer,and at the same time,the adaptive law derived from the Lyapunov function is used to adjust the weights of the neural network.When designing dynamic surface control,a first-order low-pass filter is used to eliminate the "computational expansion" problem of virtual control derivation.The simulation experiment shows that,compared with the other two variable pitch control algorithms,the variable pitch control algorithm designed in this study has better control performance.