改进的时变非线性负荷预测组合算法  

An Improved Time-varying Nonlinear Power Load Combined Forecasting Algorithm

作　　者：陈锦攀[1] 罗滇生[1] 周勇[2] 贺辉[2] 肖时勇[3] 

机构地区：[1]湖南大学电气与信息工程学院,长沙410082 [2]湖南省电力公司调度通信局,长沙410007 [3]湖南省电力公司娄底电业局,娄底417000

出　　处：《电力系统及其自动化学报》2012年第1期115-119,共5页Proceedings of the CSU-EPSA

摘　　要：为了提高超短期负荷预测精度,提出了一种改进的基于学习的时变非线性组合预测算法,该算法在基预测器中增加了基于最大Lyapunov指数的混沌时间序列预测模型,其中最大Lyapunov指数为序列特征属性,在进行组合预测时将序列的特征属性和基预测器预测的结果形成元知识,作为元预测器的输入,从而发现并且纠正基预测器的系统偏差。在元预测器中,通过门控网络确定各基预测器的权重,保证了权重的时变性和非负性。预测结果表明,该算法的预测精度较高,具有实际应用价值。An improved combined algorithm of time-varying nonlinear forecasting is presented in order to improve the accuracy of ultra short-term power load forecasting.The algorithm increases chaotic time series prediction model based on the largest Lyapunov exponent in base predictors.The largest Lyapunov exponent is a kind of feature attributes of series.Meta knowledge formed by the results of base predictors and feature attributes of series is used as inputs of meta predictor when combined forecasting is applied.System bias can be founded and rectified.The weights of base predictors are calculated using gating network in metal predictor.Weights of base predictors are time-varying and non negative.The new algorithm is applied in ultra short-term load forecasting.Results show that forecasting precision of the proposed method is higher and it has practical application value.

关 键 词：组合预测 元学习 门控网络 最大李雅普诺夫指数 超短期负荷预测 

分 类 号：TM715[电气工程—电力系统及自动化]