Wind Power Prediction Based On Improved Particle Swarm Optimization Algorithm For Least Squares Support Vector Machine

In recent years,with the rapid development of the new energy industry,the scale of wind farms is also expanding.Due to the uncertainty of wind power output,the large-scale access of wind power leads to the threat of stability and security of the power grid.If the accuracy of wind power prediction can be improved,the difficulty of power dispatching can be effectively reduced,and the safety and stability of the power grid can be indirectly guaranteed.Aiming at the characteristics of wind power output power with volatility,intermittent and randomness,this paper proposes a wind power prediction model based on improved particle swarm optimization least squares support vector machine,and studies from the following aspects:(1)Improvements in particle swarm optimization.The basic principle and optimization process of the basic particle swarm optimization algorithm are analyzed.With reference to the improved direction of the particle swarm optimization algorithm,the shortcomings of the local optimality due to the reduction of population diversity are proposed.Two methods of adaptive adjustment and variation operation of inertia weight are proposed.The improved improvement strategy is to improve the convergence speed and accuracy of the particle swarm.Six test functions are used to test the particle swarm optimization algorithm before and after the improvement,and the convergence speed and accuracy of the algorithm are analyzed and compared.(2)A wind power prediction model based on particle swarm optimization vector machine is constructed.Aiming at the large-scale and non-linear characteristics of historical wind power data and wind speed data,a wind power prediction method based on support vector machine is proposed.To solve the problem of difficult selection of vector machine parameters,the particle swarm optimization algorithm is used to optimize the parameters of vector machine.Using the optimized parameters to construct a vector machine wind power prediction model based on particle swarm optimization.(3)A wind power prediction model based on particle swarm optimization least squares support vector machine is constructed.In view of the quadratic programming and inequality constraints to be solved when the wind turbine power is predicted by the vector machine,the computational complexity is too high and the training time is too long.The improved strategy based on the least squares vector machine as the wind power prediction model is proposed.It is difficult to obtain the parameters of the least squares vector machine,and it is difficult to obtain the appropriate parameters by the empirical method.The optimization of parameters is proposed by using particle swarm optimization.The wind power prediction model based on particle swarm optimization least squares support vector machine is constructed.The training process and result analysis of the sample verify the superiority of the improved model.(4)A wind power prediction model based on improved particle swarm optimization least squares support vector machine is constructed.Aiming at the problem that the particle swarm optimization of the minimum quadratic support vector machine parameters is easy to fall into the local optimum and can't obtain the optimal parameters,the model parameter strategy of the improved particle swarm optimization minimum quadratic support vector machine is proposed.The optimized parameters are used.The wind power prediction model based on improved particle swarm optimization least squares vector machine is constructed.The training results of historical wind power output and wind speed data are analyzed by each model.The wind power of the minimum quadratic support vector machine based on improved particle swarm optimization is obtained.The predictive model predicts faster,more accurate,and more system performance.