Wind Power Forecasting Based On SST And Neural Network

With the further shortage of the globe resource, wind power is becoming more ofconcern as a clean, rich-reserving, and renewable energy. The most urgent problem is topromote the accuracy of wind power prediction during the trend of connection betweenthe large wind farm and traditional power systems. Nowadays, the study of wind powerprediction pays more attention to the good or bad attribute of forecasting model,ignoring the analysis of the characteristic of wind power time series, which results theprediction model cannot learn all the information about the wind power time series fully.According to what is motioned above, this thesis uses SST and Neural Networks tomodify extant wind power prediction model, which is demonstrated as follows.We use the SST scale transform time series of wind power, to achieve itspreliminary signal denoising, sharpening processing, improve precision of the frequencysignal time-frequency curve and effectively identify the structure of the instantaneousfrequency. Using the method of C-C to phase space reconstruction of power signal,Estimating the embedding dimension and time delay of time series, using the method ofsmall data quantity calculate the maximum Lyapunov index time series, determine thechaos characteristics of time series. Construct radial basis function (RBF) neuralnetwork and echo state neural network based on SST, and predict the wind power.Experiments show that: Through two kinds of neural network prediction method basedon SST prediction effect is improved greatly, prediction accuracy is greatly improved.This proved that the prediction method with synchrosqueezing wavelet transformcombined neural network model for wind power has a strong practical, more enrich thetheory of short-term wind power prediction.