【摘要】 本文介绍一个正在实际使用的短期电力负荷智能预报系统。针对电力负荷变化的非平稳性,该系统能消除异常数据对预报值的影响,能对预报结果进行检验及修正,能根据预报误差自动校正预报模型参数,能根据“近大远小”的原则进行预报,并能将多种不同的数学模型加权组合预报。该系统对节日负荷预报也有自己的特色。实际使用结果表明,该系统具有一定的智能性,对非平稳的电力负荷序列有很好的自适应性。更多还原

【Abstract】 This paper presents an intelligent system for forecasting the short term load being used. Directed at the non-stationarity of load, this system can eliminate the influence of abnormal data on forcasting, check and amend the forecasting result, automatically modify the model parameters based on the forecasting errors, and give out better fore-casting result by grouping different models. This system has its own character for the fore-casting during the holidays. The results of load forecasting show that this system has intelligence in a certain degree and good adaptability for the nonstationary load series.更多还原