基于多维度与主成分分析的年最大负荷分解预测  

Decomposed Forecasting of Annual Peak Load Based on Multi-Dimensional and Principal Component Analysis

作　　者：王宝 马燕如 杨敏 胡光景 伍飞 王重 WANG Bao;MA Yanru;YANG Min;HU Guangjing;WU Fei;WANG Zhong(Institute of Economy and Technology,State Grid Anhui Electric Power Co.,Ltd.,Hefei 230022,China;Anhui Mingsheng Power Investment Group Co.,Ltd.,Hefei 230022,China;State Grid Anhui Electric Power Co.,Ltd.,Hefei 23022,China)

机构地区：[1]国网安徽省电力有限公司经济技术研究院,安徽合肥230022 [2]安徽明生电力投资集团有限公司,安徽合肥230022 [3]国网安徽省电力有限公司,安徽合肥230022

出　　处：《安徽电气工程职业技术学院学报》2021年第1期16-23,共8页Journal of Anhui Electrical Engineering Professional Technique College

摘　　要：为提高年度最大负荷预测准确度,更好地服务电网公司电网规划与运行科学精准管理,结合最大负荷可分解特性,文章将年度最大负荷分解为基础负荷和空调负荷;并从经济增长、城镇化进程、居民与服务业需求、气象和电网供电能力5个维度系统总结了基础负荷和空调负荷共计11个影响因素指标,构建了基于多维度与主成分分析的年最大负荷分解预测模型。安徽省某地市的实例表明,文章提出的年最大负荷分解预测模型由于综合考虑了分类负荷特性及其多维度影响因素,预测准确率高,可作为电力市场分析预测和电网规划人员开展年度最大负荷预测的一种有效方法。In order to improve the accuracy of annual peak load forecasting and better serve power grid companies in scientific and precise management of power grid planning and operation,this paper decomposes annual peak load into base load and air conditioning load in accordance with its decomposability.From such five dimensions as economic growth,urbanization process,household and service demand,meteorology and the grid’s power supply capacity,this paper systematically summarizes eleven influencing factors and establishes a decomposed forecasting model of annual peak load on the basis of multi-dimensional and principal component analysis.An instance of a city in Anhui Province demonstrates that the decomposed forecasting model is of high accuracy with characteristics of different loads and the multi-dimensional influencing factors comprehensively considered.This model can provide an effective method for electricity market analysts and power grid planners in forecasting annual peak load.

关 键 词：最大负荷 分解 多维度 主成分 预测 

分 类 号：TM715.1[电气工程—电力系统及自动化] F407.61[经济管理—产业经济]