Modeling Simulation And Optimal Control For Double-Fed Wind Turbines

As a kind of clean,pollution-free,easy to obtain and huge reserves of renewable energy,wind energy has been vigorously used and developed in recent years.However,wind power is intermittent and random,which leads to a large fluctuation of the output power of wind turbine.With the increasing installed capacity of wind power,large-scale grid connected wind power has an impact on the power quality of the grid.In order to ensure the safe and stable operation of wind turbines,provide highquality power output to the grid,and extend the life of wind turbines,this paper takes doubly fed wind turbines as the research object,with the goal of maximum wind energy tracking and smooth output power,focuses on the modeling of wind turbine,proposing a BP neural network algorithm based on the optimization of improved artificial bee colony algorithm to forecast wind speed,and adopting a multi model predictive control strategy.Firstly,the principle and operation characteristics of each part of the doubly fed wind turbine are studied deeply,and the mathematical model is established,which is built on MATLAB / Simulink platform.Considering the strong nonlinearity of the unit and the randomness of the wind speed,the model is linearized in sections,and the state space model of the unit is further derived,which lays a good foundation for the follow-up control strategy research.Secondly,in view of the shortcomings of the standard artificial bee colony algorithm,such as slow convergence speed and easy to fall into local optimum,an improved artificial bee colony algorithm is proposed,which is combined with cross mutation mechanism of genetic algorithm,and added in Gauss mutation and chaotic perturbation in the later stage of the algorithm.The improved bee colony algorithm is applied to the parameter optimization of BP neural network,and the wind speed prediction is carried out.The comparative experimental analysis shows that the CGCGABC-BP algorithm proposed in this paper has higher prediction accuracy and better prediction effect.Finally,aiming at fixing the problems of large power fluctuation and output overshoot in the traditional PI control strategy,a multi model prediction controller based on the state space model is designed,and the forecasted wind speed value is used for the feedforward compensation of wind speed disturbance.The simulation results show that compared with PI control,the fluctuation of generator speed and output power of MMPC designed in this paper is significantly reduced.At the same time,the effect of wind speed disturbance suppression makes the pitch actuator act in advance,and the variation of pitch angle action is also reduced.