基于RBF神经网络分位数回归的电力负荷概率密度预测方法

A Power Load Probability Density Forecasting Method Based on RBF Neural Network Quantile Regression

针对电力系统短期负荷预测问题,在现有的组合预测和概率性区间预测的基础上,提出了基于RBF神经网络分位数回归的概率密度预测方法,得出未来一天中任意时期负荷的概率密度函数,可以得到比点预测和区间预测更多的有用信息,实现了对未来负荷完整概率分布的预测。中国某市实际数据的预测结果表明,提出的概率密度预测方法不仅能得出较为精确的点预测结果,而且能够获得短期负荷完整的概率密度函数预测结果。

According to the problem of short-term load forecasting in the power system, this paper proposed a probability density forecasting method using radical basis function （RBF） neural network quantile regression based on the existed researches on combination forecasting and probability interval prediction. The probability density function of load at any period in a day was evaluated. The proposed method can obtain more useful information than point prediction and interval prediction, and can implement the whole probability distribution forecasting for future load. The practical data of a city in China show that the proposed probability density forecasting method can gain more accurate result of point prediction and obtain the forecasting results of integrated probability density function of short-term load.