Load prediction method and load prediction system based on general distribution

The invention discloses a load prediction method and prediction system based on general distribution, and aims to realize short-term load power prediction of a power system. According to the method, historical meteorological information data and electrical data are combined, and corresponding statistical analysis models are constructed for different temperature increment levels. Firstly, a histogram model is adopted to approximately represent actual distribution, then a proper model is selected to fit the actual distribution, a universal distribution function model is adopted to fit, and a confidence interval of predicted power under a certain confidence level is obtained according to a closed analytical expression of an inverse function of a universal distribution CDF function. 一种基于通用分布的负荷预测方法及预测系统，以实现电力系统短期负荷功率预测。该方法将历史气象信息数据和电气数据相结合，针对不同的温度增量水平分别构建相应的统计分析模型。首先采用直方图模型来近似表征实际分布，然后选取适当的模型对实际分布进行拟合，本文采用通用分布函数模型进行拟合，并且依据通用分布CDF函数的逆函数的闭合解析表达式，得出一定置信水平下预测功率的置信区间。