为提高风功率短期预测的准确率,提出一种基于改进灰狼算法优化加权最小二乘支持向量机(Weighted Least Squares Support Vector Machine,WLSSVM)的短期风功率预测方法.采用C-C法对风功率时间序列的嵌入维数进行了计算,根据计算结果确定短期风速预测输入量与输出量的关系.利用Tent映射和参数非线性调整策略对灰狼算法进行改进,得到了优化性能更强的改进灰狼优化(Improved Grey Wolf Optimization,IGWO)算法,并利用测试函数验证了IGWO算法能够加快迭代收敛,提高计算精度.采用IGWO算法对WLSSVM的惩罚系数和核参数进行优化,建立基于IGWO-WLSSVM的短期风功率预测模型.采用某风电场春夏两个不同季节的风功率数据进行算例分析,结果表明,所提短期风功率预测结果的平均相对误差、均方根误差和最大相对误差更小,风功率预测精度和预测结果的稳定性均优于其他方法,验证了所提方法的有效性和实用性.
Short Term Wind Power Prediction Based on WLSSVM Optimized by Improved Grey Wolf Optimization Algorithm
In order to improve the accuracy of short-term wind power prediction,the author proposes a short-term wind power prediction method based on an improved grey wolf algorithm optimized weighted least squares support vector machine.The embedding dimension of wind power time series is calculated using the C-C method,and the relationship between short-term wind speed prediction input and output is determined based on the calculation results.The improved Grey Wolf Optimization algorithm with stronger optimization performance is obtained by utilizing Tent mapping and parameter nonlinear adjustment strategies.The improved IGWO algorithm is validated through test functions to accelerate iterative convergence and improve computational accuracy.Using the IGWO algorithm to optimize the penalty coefficients and kernel parameters of Weighted Least Squares Support Vector Machine,a short-term wind power prediction model based on IGWO-WLSSVM is established.Using wind power data from two different seasons of a wind farm,spring and summer,a numerical analysis is conducted.The results show that the accuracy and stability of the wind power prediction results are better than other methods,verifying the effectiveness and practicality of the proposed method.
wind powerthe improved grey wolf optimization algorithmweighted least squares support vector machineC-C method
国家科技期刊平台
NSTL
万方数据
