计及气象因素影响的短期电力负荷预测方法  

A Short-term Load Forecasting Method Considering the Influence of Meteorological Factors

作　　者：周宇 ZHOU Yu(NARI Group Co.,Ltd(State Grid Electric Power Research Institute),Nanjing 210000 China)

机构地区：[1]南瑞集团(国网电力科学研究院),江苏南京210000

出　　处：《自动化技术与应用》2020年第6期107-113,共7页Techniques of Automation and Applications

摘　　要：随着电力系统的智能化、现代化发展,电力负荷的种类越来越多,气象因素对负荷的影响愈显突出。文中提出了针对气象因素的电力系统短期负荷预测方法,共分为回归分析和负荷预测两个部分。在回归分析部分,首先对选取的五个气象因素进行了主成分分析,然后分析自变量与负荷的关系,剔除对负荷影响不大的气象因素。在负荷预测部分,在模型建立过程中,对不同量纲的数据进行了归一化处理,并考虑工作日和双休日对负荷的影响,通过MATLAB建立神经网络,通过算法分析得到历史负荷数据与预测数据之间的非线性映射关系,分别绘制出了两个地区考虑气象因素和不考虑气象因素两种情况下的电力负荷预测图像,误差分析表明该预测算法具有较高的预测精度。With the intelligent and modern development of power systems,the types of power loads are becoming more and more complex,and the impact of meteorological factors on the load becomes more prominent.A short-term load forecasting method for power systems based on meteorological factors is proposed in this paper which is divided into two parts:regression analysis and load forecasting.In the regression analysis part,the principal component analysis of the selected five meteorological factors is first carried out,then the relationship between the independent variables and the load is analyzed,and the meteorological factors that have little effect on the load are excluded.In the load forecasting part,during the model building process,the data of different dimensions are normalized,and the impact of workdays and weekends on the load is considered.The neural network is established by MATLAB,and the nonlinear mapping relationship between historical load data and predicted data is obtained by algorithm analysis.The electric load forecasting images of two regions considering meteorological factors and no meteorological factors are drawn.Error analysis shows that the prediction algorithm has higher prediction accuracy.

关 键 词：电力系统 负荷预测 气象因素 回归分析 神经网络法 

分 类 号：TM715[电气工程—电力系统及自动化]