基于分行业负荷曲线的地区电网分级负荷预测  

Multi-level Load Forecasting Based on Load Curve for Different Industries

作　　者：薛季良[1] 顾洁[1] 杜渐[1] 储琳琳[2] 张宇俊[2] 陆慧丰[2] 

机构地区：[1]电力传输与功率变换控制教育部重点实验室,上海交通大学电气工程系,上海200240 [2]上海市电力公司市南供电公司,上海201100

出　　处：《电力系统及其自动化学报》2014年第1期18-21,共4页Proceedings of the CSU-EPSA

基　　金：国家高技术研究发展计划("863"计划)项目(2011AA05A108)

摘　　要：由于测控设备未能完全安装或测量设备非正常投运,使得地区电网中地块负荷数据获取较为困难,从而影响了配电网规划的效果。从上下级的负荷动态平衡以及用户日用电量平衡的双向平衡出发,结合不同行业或功能块的负荷特性,将日负荷曲线应用于协调预测的最优化建模,对地区电网的分级负荷预测方法进行挖掘,提出了分级负荷预测的新算法。结果表明,该方法在解决未知用户问题的同时拥有较高的预测精度,是进行地区电网分级负荷预测工作中一个有益的尝试。The problems in monitoring apparatus easily lead to information loss in regional power grid.This deficiency phenomenon in devices becomes a key barrier to the existing multi-level load forecasting methods.This paper proposes a new multi-level load forecasting algorithm which establishes an optimization model through the bilateral balance between dynamic power load balance in neighboring voltage network and daily electricity consumption balance of devices.Daily load curve for different industries or functional blocks is applied to the method based on typical load characteristics.Result shows that the method can cope with information loss issue with a high prediction accuracy which indicates it is a meaningful attempt to the regional load forecasting.

关 键 词：分级负荷预测 日负荷曲线 地区电网 负荷特性 双向平衡 

分 类 号：TM715[电气工程—电力系统及自动化]