标题:
基于目标规划法的灰色Verhulst负荷预测模型Grey Verhulst Load Forecasting Model Based on Objective Programming
作者:
周德强, 向锋
关键字:
负荷预测, 目标规划法, 最小一乘法, 最小二乘法, 灰色Verhulst模型Load Forecasting; Objective Programming; Least Absolute Deviation; Least Square Procedure; Grey Verhulst Model
期刊名称:
《Hans Journal of Data Mining》, Vol.2 No.3, 2012-07-25
摘要:
为克服传统灰色Verhulst模型中利用最小二乘法估计参数存在的不足，改善灰色Verhulst模型在具有“S型”增长或处于饱和增长状态的中长期电力负荷预测中的精度，提出在最小一乘法准则下，利用目标规划法估计灰色Verhulst模型参数的方法。对某中长期负荷进行预测，并与传统的灰色Verhulst模型进行对比分析。结果表明，该方法发挥了最小一乘法受奇异值影响小，稳健性好的优点，避免了利用最小二乘法估计灰色Verhulst模型参数存在的不足，预测精度更高。
In order to overcome the defects of parameters estimation in traditional grey Verhulst model by means of least square procedure, and enhance the forecasting accuracy of grey Verhulst model in medium and long-term load forecasting for load growth in S-type or load growth being saturated, an estimation method based on least absolute de- viation, which use objective programming to estimate the parameters of grey Verhulst is presented. Then, this model is applied to long-term load forecasting, and is compared with the traditional grey Verhulst model. The results show that the method takes advantages of the benefits of least absolute deviation, which is small influenced by singular value, and robustness is good. This model avoids the defects of parameters estimation in traditional grey Verhulst model by means of least square procedure, and forecasting precision is higher.