Research on Mid-term Prediction of Regional Electricity Consumption Considering the Impact of Major Social Activities

In this paper, the method of combining discrete wavelet transform and artificial neural network under the grey prediction model is used to construct a quantitative hierarchical model of the advancing wave of regional electricity consumption trends. And inserting sensitive quantity correction is used to obtain an improved regional power prediction model. This method extracts relevant feature quantities according to the influence of many types of large-scale social activity factors on electricity. At the same time, it also solves the situation that the trend of electricity consumption is ambiguous and difficult to quantify under the background of major social activities. This paper uses the above improved power prediction model to verify the electricity consumption during the G20 meeting in Hangzhou. The results have good fitting degree and accuracy, which meet the engineering needs.