基于数据挖掘技术的短期电支负荷预测

负荷预测是电力市场技术支持系统的一个重要组成模块，对电网的安全、经济运行具有重要的意义。负荷预测主要综合考虑系统的运行特性、社会影响、自然条件以及增容决策等因素，在历史负荷数据的基础上，进行一系列数学计算，在满足一定精度要求的情况下，得出未来某特定时刻的负荷值。传统短期电力负荷预测方法易受随机因素的干扰，尤其在小水电分布众多的地区预测精度不高。文章针对短期负荷预测的特点，将数据挖掘技术引入短期负荷预测中，并给出了系统的解决方案，可应用于小水电众多的电网环境或类似环境。实例运算表明该系统可有效地提高预测精度。

A Short-term Power Load Forecasting System Using the Data Mining Technology

As an important part of the electric power market technical support system, load forecasting has an important significance for safe and economic grid operations. The traditional short-term power load forecasting methods mainly take several factors including system operation characteristics, social influence, natural conditions and decisions into account, and predict the future load at a given time point by a series of calculations with a required precision. However, these methods are vulnerable under the case of the interference of random factors, especially the prediction accuracy will be greatly reduced in the environment of many small hydropower distribution. According to the characteristics of the short-tem~ load forecasting, this paper introduces the data mining technology and presents a solution of the short-term load tbrecasting system applicable to the environment of many small hydropower distribution, or similar environments. Example computation shows that the system can effectively improve the prediction accuracy.