基于改进AHP法和改进熵权法结合的组合电量预测模型

Combined Electric Quantity Forecasting Model Based on the Combination of Improved AHP Method and Improved Entropy Weight Method

为了更加精确地预测电量,提出了一种基于改进层次分析法(AHP法)和改进熵权法模型的权重分配方法。相对于传统AHP法和熵权法,所提方法可以构造必定具有一致性的判断矩阵,无需一致性校验和调整,同时解决了多熵值分布都趋于1时熵权分配紊乱问题,并基于该方法对多种单项预测模型进行权重分配,从而得到组合电量预测模型。最后以某地区一个月用电量为例,用上述模型进行电量预测,结果得到预测电量曲线和预测误差。结果表明,使用改进方法后的模型可有效提高预测的精确性,大大降低预测误差。

In order to predict electricity quantity more accurately,a weight distribution method based on improved analytic hierarchy process(AHP method)and improved entropy weight method model was proposed.Compared with the traditional AHP method and entropy weight method,this method can construct a judgment matrix that must have consistency without consistency check and adjustment.At the same time,it solved the problem of entropy weight distribution disorder when the multi-entropy value distribution tends to 1.This method assigned weights to multiple single-term forecasting models to obtain a combined electricity forecasting model.Finally,taking one month’s electricity consumption in a certain area as an example,the above model was used to predict the electricity quantity,and the forecasting electricity curve and forecasting error were more accurately obtained.The results show that the model after using the improved method can effectively improve the accuracy of prediction and greatly reduce the prediction error.