基于季节ARIMA模型的月度用电量置信区间预报模型

Seasonal ARIMA Models Based Confidence-intervals Forecast Model for Monthly Electricity Consumption

结合广东某地区用电量预测系统的开发工作,建立了基于季节ARIMA模型的月度用电量置信区间预报模型。该模型可方便求解出含一定置信度的预报结果范围,能体现出月用电量的真实值;时间序列方法的应用,避免了预测其他非用电量的困难。此外,还分析了数据统计中存在着的数据缺失问题及其处理方法。对地区电网月度用电量进行实际预测,取得了理想的结果。

Based on the electricity consumption forecast system of a certain area in Guangdong,a seasonal ARIMA models based confidence-intervals forecast model for monthly electricity consumption is established.This model can obtain forecast interval with certain degree of confidence,it can manifest the reality of the monthly electricity consumption.The sequential model successfully avoids the difficulties in forecasting with other non-electricity consumption factors.Finally,missing values processing methods due to statistical work problems is studied.Ideal results are obtained in practical prediction about the monthly electricity consumption in a certain area in Guangdong.