Ultra-short-term Prediction Of Wind Power Based On Wavelet-least Squares Support Vector Machine

With the accelerated development of wind power generation technology in the world,the improvement of power quality and the security and stability of power system has become an important issue after grid connection.In order to resolve the volatility and volatility of wind resources and improving the safe and stable operation of wind energy production,wind power prediction technology is of great importance.A precise prediction of wind power is extremely important in order to improve the economic benefits of wind power production,reducing the spare capacity of power systems,and scheduling wind farms to shut down groups.This paper proposes a combined prediction model to predict wind power.The basic idea of the model is to first process the wind speed and power sequence.For the nonlinearity and volatility of the wind speed signal and the power signal,the wavelet transform method is used to decompose the wind speed and power signals into sub-sequences of different frequencies.The wavelet sequence is denoised and reconstructed using the wavelet soft threshold denoising method.The moving average method and the median filtering method are used to denoise the wind speed and power data to compare the denoising effect.The processed wind speed,temperature,and power data are input into a Least Squares Support Vector Machine(LSSVM)for training,and ultra-short-term prediction is performed on a single turbine wind turbine.The most suitable kernel function in LSSVM is determined by comparison experiments,and the parameters of the kernel function in the model are optimized by Particle Swarm Optimization(PSO).Using the data rolling method,the prediction accuracy of the three schemes is predicted by comparing two,six,and twelve data in one time and adding them to the training set.Finally,the method of predicting 6 data at a time is used to predict and reduce the prediction error.In this paper,the combined prediction model is applied to a wind turbine in a wind farm in Fuxin City,Liaoning Province for simulation experiments.The results show that the combined prediction model proposed in this paper can accurately predict the output power of a single wind turbine.The prediction results of the combined prediction model are all within the allowable range of the Wind Power Prediction Function Specification,especially at the power signal mutation point,which is better than the existing single prediction models.This shows that the model is suitable for the power prediction problem of turbine wind turbines.