摘要
采用RBF网络模型对电力市场中的清算电价进行预测,聚类算法采用改进的模糊C均值聚类,减小了野值对输出结果的影响,隐层的输出采用聚类结果的隶属度函数,省掉了对径向基函数宽度的计算。通过美国加州电力市场公布的历史数据对该模型进行验证,结果表明该模型应用于电价预测具有较高的预测精度,并且具有训练速度快、不存在局部极小和过拟合等优点。
It presents the method of Power Market Clearing Price(MCP) forecasting based on Radial Basis Function (RBF) network with improved fuzzy C means clustering. The advantages of this method over BP network method are: higher precision and faste r training speed, free of local optimality and over fitting. A example on Califo rnia PX history data shows that the method is applicable.
