两种短期电力负荷预测模型的比较

The comparison between two kinds of short-term power load forecast models

负荷预测是电力系统规划、计划、用电、调度等部门的基础工作,电力负荷是影响电网寿命和可靠度的一个重要因素。结合某省电网近五年来总电力负荷数据,对电力负荷数据进行分析与预处理,分别运用ARIMA模型与灰色GM(1,1)模型针对该省的电力总负荷分别做未来3天和10天两种情况下的短期预测,重点比较了所用两种方法的优缺点和精准度,得到3天的预测精度上GM(1,1)模型效果高于ARIMA模型,10天的预测精度上ARIMA模型相对较好。

The load forecasting for power system in planning, supply, scheduling and other departments is fundamental. Power load influences the life and reliability of its grid. According to the total power data of a provincial grid over the past five years, the load data analysis and pretreatment are conducted. By using ARIMA model and grey GM （ 1, 1 ） model the total electricity load of the grid is divided in two cases ： 3 days and 10 days, to make short-term prediction. Compared with the accuracy, advan- tages and disadvantages of the two methods. The precision of 3-day forecast in GM （ 1, 1 ） model is more effective than that in ARIMA model. The precision of 10-day forecast is relatively higher in ARIMA model.