论文部分内容阅读
为解决变电站综合负荷模型参数的随机时变性问题,提出一种基于支持向量机(support vector machines,SVM)的变电站用电行业负荷构成的预测方法。运用SVM算法预测变电站日负荷曲线,提取变电站日负荷特征量。在此基础上,利用负荷控制系统的用户日负荷曲线,通过模糊C均值聚类获得各行业的典型特征量,将其分别投影到变电站日负荷特征量上;然后进一步计算权值,得到各行业负荷比例。根据某地区的用电特点,对该地区某变电站的夏季最大负荷日的行业构成比例进行预测,结果表明该方法符合电网实际运行情况。
In order to solve the stochastic time-varying problem of substation integrated load model parameters, a load forecasting method based on support vector machines (SVM) is proposed. The daily load curve of substation is predicted by SVM algorithm and the daily load characteristics of substations are extracted. On this basis, using the daily load curve of the load control system, the typical characteristic quantities of each industry are obtained by fuzzy C-means clustering and projected onto the daily load characteristics of substations respectively. Then, Load ratio. According to the characteristics of electricity consumption in a certain area, the ratio of the industry composition of the maximum load day of a certain substation in the region is predicted. The results show that this method is in line with the actual operation of the power grid.