基于数据分析和改进Chebyshev神经网络的风速时间序列预测

为提高风速时间序列预测精度,基于风速时间序列的随机性和波动性,提出互补集合经验模态分解(Complete Ensemble Empirical Mode Decomposition,CEEMD)和正交粒子群算法(Orthogonal Particle Swarm Optimization,OPSO)优化Chebyshev基函数神经网络的混合风速时间序列预测模型(CEEMD-OPSO-Chebyshev)。利用CEEMD将原始风速时间序列分解成有限个固有模态分量,避免了传统的分解信号重建中冗余噪声残留问题。同时引入排列熵分析各分量内在特性进行聚类,提出基于OPSO优化算法的Chebyshev神经网络风速预测模型,利用OPSO优化预测网络权值,进一步提高预测精度,通过对实际采样的风电场风速时间序列进行预测分析,结果可得所提出的混合预测模型与传统预测模型相比能得到更高的预测精度。

Time Series Prediction of Wind Speed Based on Data Analysis and Improved Chebyshev Neural Network Zhang Xu1 , Zhang Hongli1 , Fan Wenhui2,Wang Cong1

In order to improve the prediction accuracy of wind speed time series, based on the randomness and fluctuation of wind speed time series, a hybrid wind speed time series prediction model (CEEMD-OPSO-Chebyshev) based on Chebyshev basis function neural network is proposed by using complementary set empirical mode decomposition and orthogonal particle swarm optimization. The original wind speed time series is decomposed into finite intrinsic modal components by CEEMD, which avoids the residual problem of redundant noise in traditional decomposition signal reconstruction. At the same time, permutation entropy is introduced to analyze the intrinsic characteristics of each component for clustering, and a Chebyshev neural network wind speed prediction model based on orthogonal particle swarm optimization algorithm is proposed, which uses orthogonal particle swarm optimization to predict the weight of the network to further improve the prediction accuracy. Through forecasting and analyzing the actual wind speed time series of wind farms, the results show that the proposed hybrid forecasting model can obtain higher forecasting accuracy than the traditional forecasting model.