计及经济因素的混合模型电量预测研究

Research on Electricity Consumption Prediction under Mixed Model Considering Economic Factors

摘要:

精确的电量预测是进行电网规划和建设的重要依据。为解决月度用电量预测方法思路单一、预测模型通用性不强、预测精度低等问题，通过对区域用电量与经济的历史数据研究分析，提出了一种计及经济因素影响的混合模型月度用电量预测方法。该方法将回归分析法和指数平滑法这两种预测方法相结合，建立电量预测模型，取得了较高的预测精度；最后，将本模型运用于浙江嘉兴地区月用电量预测，结果表明：该模型预测平均绝对百分误差为（MAPE）2.31%，从而验证了该模型的有效性与通用性。

Abstract:

Precise electricity consumption prediction provides an important basis for layout and construction of power grid. In order to solve the problems of single method, weak universality and low accuracy of prediction models, a monthly electricity consumption prediction method under mixed model on the basis of economic factors is proposed through analyzing the historical data of electricity consumption and economy. The regression analysis method and exponential smoothing method are combined to set up a mixed model of electricity consumption prediction, which contributes to the high prediction accuracy. At last, the proposed model is applied to the monthly electricity consumption prediction of Jiaxing area. The results show that its mean absolute percentage error (MAPE) is 2.31%, which verifies the effectively and universality of the model.