特殊事件对月售电量及其预测的影响分析

特殊事件会使月售电量发生很大变化，导致实际售电量曲线明显偏离典型售电量曲线。然而由不考虑特殊事件的传统预测模型得到的月售电量预测曲线却更接近于典型售电量曲线，这将不可避免地导致月售电量预测精度降低。为解决该问题，本文以异常高温，政治事件和超强台风为例，分析研究了特殊事件对月售电量及其预测的影响。首先，介绍了“互联网+”背景下基于大数据的月售电量预测模型，并对其预测精度进行了评估；其次，针对异常高温、政治事件及超强台风三种特殊事件，描述了各事件的特殊情况，以实际月售电量数据说明了特殊事件对月售电量的影响，然后利用月售电量预测模型研究了特殊事件对月售电量预测的影响，并详细分析了产生这种影响的原因，在此基础上，针对不同的特殊事件提出了相应的初步改善对策。

Influence Analysis of Special Events on Monthly Electricity Sales and Its Forecast

Due to the special events, the monthly electricity sales change greatly, which will lead the real electricity sales curve to the obvious deviation from the typical electricity sales curve. However the predicted electricity sales curve obtained by traditional forecast model without the special events considered is more like the typical electricity sales curve. The forecast accuracy of electricity sales decreased inevitably. In order to solve the problem, this paper took the abnormal high temperature, political events and super typhoon as an example to analyze the influence of special events on the monthly electricity sales and its forecast. Firstly, the forecast model of electricity sales based on big data in the background of internet + was introduced and its forecast accuracy was test. Then, the special events are described in three special events, such as abnormal high temperature, political events and super typhoon. The effect of special events on electricity sales was verified with real monthly electricity sales. The impact of special events on electricity sales forecast was researched by the forecast model and the reasons for this impact were analyzed. On this basis, the corresponding initial improvement methods were put forward for different special events.