内蒙古师范大学学报(自然科学汉文版)

电力负荷预测是电力系统规划与调度的重要依据，气象因子则是影响短期电力负荷变化的重要因素。以运城市2016—2019年日最大电力负荷数据为基础，结合同期气象观测数据，分析日最大电力负荷的变化特征。对比分析3种夏季气象电力负荷提取方法，建立其与气象因子及综合气象指数的相关关系，从中选择最优方法，通过多元线性回归、逐步回归和BP神经网络算法构建夏季气象电力负荷预测模型。结果表明：运城市日最大电力负荷呈逐年增长趋势，具有明显的季节性变化，存在明显的节假日（春节、国庆）效应及一定的周末效应；综合气象指数相较于大多数单一气象因子，能更全面地表征夏季气象电力负荷的变化规律；引入气象因子及综合气象指数的BP神经网络算法对气象负荷的拟合精度与预测效果最优。

Power load forecasting is an important basis for the planning and dispatching of power systems, while meteorological factors are significant influences on short-term variations in power load. Based on daily maximum power load data from Yuncheng City from 2016 to 2019, combined with meteorological observation data from the same period, this paper analyzed the variation characteristics of daily maximum power load. By comparing and analyzing three methods for extracting summer meteorological power load, the paper established the correlation between these methods and meteorological factors as well as a comprehensive meteorological index. The optimal method was selected, and a summer meteorological power load forecasting model was constructed using multiple linear regression, stepwise regression, and BP neural network algorithms. The results indicate that the daily maximum power load in Yuncheng City shows a year-onyear increasing trend, with significant seasonal variations. There are notable holiday effects(such as during the Spring Festival and National Day) and some weekend effects. The comprehensive meteorological index represents the variation patterns of summer meteorological power load more comprehensively compared to most individual meteorological factors. The BP neural network algorithm, which incorporates meteorological factors and the comprehensive meteorological index, demonstrates the highest fitting accuracy and forecasting performance for meteorological load.