Pitch Control Of Large Wind Turbine Based On Predictive Feedforward

Fossil energy is the main energy material in the world,including coal,natural gas,oil and so on.It provides energy support for human life and production in all aspects.However,the non-renewability of the fossil energy and its pollution to the environment in the combustion process have become a common concern of the mankind.It is becoming increasingly important and urgent to develop sustainable and clean new energy sources.At present,the new sustainable green energy mainly includes solar energy,tidal energy,wind energy and so on.Wind energy has attracted much attention because of its rich resources,low cost of power generation and mature technology.The pitch control is the core technology of wind power generation.The performance of power generation is directly affected by the control algorithm.Excellent control algorithm can not only improve the accuracy of output power,but also reduce power jitter,hysteresis and response overshoot.At the same time,the wind turbine system is a large and complex system,and its large inertia,time-varying parameters,and the possible turbulence,wind shear effect and other effects make pitch control more complex and difficult to achieve effective control.In order to solve these problems,a large number of scholars in wind power research have developed various excellent control algorithms,including classical sliding mode control algorithm,neural network algorithm,fuzzy control theory,predictive control and so on.These algorithms have solved various control problems in some aspects.Aiming at the control accuracy,delay,robustness and overshoot problems,the thesis proposes a feed-forward predictive pitch control strategy.It is used to compensate the feedback control and optimize the control performance.The main work of this thesis includes:1.Analysis of the principle of pitch control and establishment of the mathematical model of the systemThe system structure and principle of wind power generation are analyzed.The wind energy calculation formula is established based on aerodynamics.The characteristic parameters of wind turbine are discussed,and the pitch control mechanism in different operation areas is obtained.Then,according to the physical model of wind turbine,a specific mathematical model is established and linearized.2.The control algorithm combining linear sliding mode control with variable universe fuzzy PIDA compound control strategy is proposed.The feedforward controller uses sliding mode variable structure control to predict and compensate the feedback link.In order to reduce the complexity of the model,the feedback linearization is applied to the system.In the feedback control,the PID control algorithm is used to track the rated power.In order to improve the control accuracy,the fuzzy intelligent algorithm is used to set the PID parameters.Meanwhile,in order to improve the adaptability,the basic universe is processed with variable universe,which makes the control of large wind speed fluctuation more accurate and the control of small wind speed fluctuation more sensitive.3.The feed-forward pitch control algorithm for Kalman filter to predict the optimal wind speedIn order to overcome the system hysteresis caused by the large inertia of wind turbines,a predictive feedforward control algorithm is studied.Firstly,the state equation of the system is discretized.Then Kalman filter algorithm is used to predict the next moment aerodynamic torque according to the discrete system model.Next Newton Raphson algorithm is used to calculate the predicted wind speed,and the least square method is used to fit the wind speed and pitch angle.The output compensation of pitch angle is obtained and the feedback output is compensated.To ensure the constant output power,the particle swarm optimization algorithm is used to adjust the characteristic parameters of PID.4.The control strategy using CMAC with variable confidence interval and feedback linearization sliding modeWind turbine is a strong non-linear,large time-delay system,conventional PID controller is difficult to meet the control requirements.In order to achieve effective control of wind turbines,a compound control method based on feedforward and feedback is adopted.In the feedback control link,the feedback linearization sliding mode control algorithm is adopted to ensure the stability of output power and improve the robustness of the system.Feedforward control uses a CMAC with variable confidence interval to change the learning performance of storage units,so that learning saturated units can reduce or stop changing,less learning units can strengthen learning,which can save memory overhead,speed up the calculation,and make the control faster and more accurate.This algorithm can approximate the nonlinear model well and impose a control quantity on the system ahead of time,which can compensate the delay caused by the large inertia of the pitch system.