非线性回归校正模型在突变负荷预测中的研究及应用  

Research and Application of Nonlinear Regression Correct Model in Sudden Change Load Forecasting

作　　者：赵海青[1] 

机构地区：[1]华北电力大学数理学院,河北保定071003

出　　处：《数学的实践与认识》2008年第1期88-91,共4页Mathematics in Practice and Theory

摘　　要：非线性回归模型是中长期负荷预测的一种有效的方法,对常规变化趋势的负荷有很好的拟合性,但对有转折点的突变趋势或增长处于饱和阶段的负荷进行预测误差较大.通过对历史数据的最优分段,提出了非线性回归校正模型,可以很好地解决这个问题.实例表明,此模型在中长期负荷预测中是适用的,尤其对于有转折点的突变趋势,具有很高的预测精度.Nonlinear regression correct model prediction model is an effective forecast method of long-term load, it is a very good fit to the load of Conventional tendency, but for the load of sudden change or growth in the saturation stage, the error is larger. Based on the optimum and segmental of historical data, the nonlinear regression correct model is proposed, it is a good solution to this problem. Example shows that this model is applicable in the long-term load forecast, especially for sudden change load forecasting, it has a high forecast accuracy.

关 键 词：突变负荷预测 非线性回归 最优分段 校正模型 

分 类 号：O22[理学—运筹学与控制论]