基于KELM与PSO-LSSVM组合核方法的风电功率区间预测研究

Research on Wind Power Interval Prediction Based on Combination Kernel Method of KELM and PSO-LSSVM

针对风电功率单一方法区间预测性能较差的问题,该文提出一种基于组合核方法的风电功率区间预测模型。首先利用混合核密度估计法,对核极限学习机(kernel extreme learning machine,KELM)和粒子群优化最小二乘支持向量机(particle swarm optimization least square support vector machine,PSO-LSSVM)两种不同的核方法的风电功率点预测误差进行概率密度拟合,并建立区间预测模型;然后使用熵权法确定KELM和PSO-LSSVM的权值并加权组合,得到最终预测区间;最后利用甘肃某风电场的真实风电功率数据验证该方法有效性。实验结果表明,该方法可有效弥补KELM准确性差和PSO-LSSVM可靠性低的缺陷,兼顾风电功率区间预测可靠性和准确性,有效提升区间预测性能。

To save the problem that single method of wind power interval prediction has poor performance,a wind power interval prediction method based on combining kernel method is proposed in this paper. Firstly,the mixed kernel density estimation was used to distribute and fit the wind power prediction errors of two different kinds of kernel methods kernel extreme learning machine(KELM) and particle swarm optimization least square support vector machine(PSO-LSSVM),and the interval prediction probability density was performed. Then the entropy method was used to determine the weights of KELM and PSO-LSSVM,and the weighted combination was used to get the final prediction interval;Finally,the real wind power data of a wind farm in Gansu were used for verification. The experimental results show that this method can compensate the flaw of KELM’s less reliability and PSO-LSSVM’s poor accuracy effectively,and combine the reliability and accuracy of wind power interval prediction,which improve the interval prediction performance effectively.