基于相空间重构和LSSVM的短期风速预测

Short Term Wind Speed Prediction Based on Phase Space Reconstruction and LSSVM

为了提高短期风速预测精度,提出了一种基于最小二乘支持向量机的短期风速预测方法。首先采用复自相关法和互信息法计算延迟时间,采用伪最邻近点法和Cao式法计算嵌入维数,使延迟时间和嵌入维数取值更合理。其次运用小数据量法计算混沌时间序列的最大Lyapunov指数,G-P算法计算时间序列的关联维数,用以证明风速序列为混沌时间序列并确定支持向量。然后采用扩展记忆粒子群对最小二乘支持向量机的惩罚参数和核函数参数进行优化,建立PSOEM-LSSVM的短期风速预测模型。最后与其他几种风速预测模型对比,仿真结果表明PSOEM-LSSVM预测模型可加快收敛速度,提高计算精度,验证了提出的短期风速预测方法的正确性和实用性。

In order to improve the accuracy of short-term wind speed prediction,a short-term wind speed prediction method based on least squares support vector machine is proposed.Firstly,the complex autocorrelation method and mutual information method are used to calculate the delay time,and the pseudo nearest neighbor method and Cao formula method are used to calculate the embedding dimension and delay time,making the delay time and embedding dimension more reasonable.Secondly,the maximum Lyapunov exponent of chaotic time series is calculated by small data method.The correlation dimension of time series is calculated by G-P algorithm to prove that wind speed series is chaotic time series and support vector is determined.Then,the penalty parameters and kernel function parameters of LSSVM are optimized by extended memory particle swarm optimization,and the short-term wind speed prediction model of PSOEM-LSSVM is established.Finally,compared with other wind speed prediction models,the simulation results show that PSOEM-LSSVM prediction model speeds up the convergence speed and improves the calculation accuracy,which verifies the correctness and practicability of the short-term wind speed prediction method proposed in this paper.