改进的灰色Verhulst模型在中长期负荷预测中的应用

Application of Improved Gray Verhulst Model in Middle and Long Term Load Forecasting

针对中长期电力负荷预测"小样本"、"贫信息"、"不确定"、"非线性"等特点,提出了基于最小二乘支持向量机算法与等维新息技术的改进灰色Verhulst模型,并将该模型用于具有"S型"增长或处于饱和增长状态的中长期电力负荷预测。根据原始数据建立了灰色Verhulst模型,利用LS-SVM算法对模型中的参数进行了估计,基于等维新息递补预测法对负荷数据进行了预测。实例计算结果表明,基于该模型得到的预测结果相对误差在3%以内,与传统预测模型相比,采用文中的模型可获得更高的预测精度。

According to such features in middle and long term load forecasting as small samples, poor information, uncertainty and nonlinearity, an improved Verhnlst model based on least square-support vector machine （LS-SVM） algorithm and equal-dimension and new-information technique is built and applied to the middle and long term load forecasting for load growth in S-type or load growth being saturated. The parameters of the model are evalutated by LSSVM algorithm and the load data is forecasted by equaldimension and new-information addition prediction. Case study results show that the relative errors of forecasting results by the proposed modes are less than 3%, thus in comparison with traditional forecasting models, the proposed model can offer more accurate forecasting results.