基于K-means聚类与支持向量机相结合的短期负荷预测

Short-term Load Forecasting of Power System Based on K-meas Clustering and Support Vector Machine

精确的负荷预测对提高电力系统运行的可靠性和经济型起着重要作用。针对当存在大量数据样本时,支持向量机在进行训练时收敛时间太长的问题,本文首先利用K-means聚类算法按照影响负荷的因素挑选相似日,然后用利用支持向量机对聚类后的数据进行负荷预测,并将预测结果与神经网络算法预测的结果进行对比。结果表明,基于K-means聚类与支持向量机相结合的负荷预测方法可以提高收敛速度和荷预测的精确度。

Accurate load forecasting plays an important role in improving the reliability and economy of power system operation.Aiming at the problem that the support vector machine has too long convergence time when training,there is a problem that the K-means clustering algorithm selects the similar day according to the factors affecting the load,and then uses the support vector machine to cluster the cluster.The data is used for load forecasting and the predicted results are compared with those predicted by neural network algorithms.The results show that the load forecasting method based on K-means clustering and support vector machine can improve the convergence speed and the accuracy of the load prediction.