应对海量数据的超短期负荷预测在实时电力市场的应用研究  

Research on the Application of ultra-short term load forecasting adapting to actual-time electricity market considering the mass data

作　　者：李家璐 何剑军 张坤[1] 刘敬诚 吕勃翰 LI Jialu;HE Jianjun;ZHANG Kun;LIU Jingcheng;LV Bohan(China Southern Power Grid Co.,Ltd.,Guangzhou 510663 Guangdong,China)

机构地区：[1]中国南方电网有限责任公司

出　　处：《电力大数据》2019年第12期21-27,共7页Power Systems and Big Data

摘　　要：在开展实时现货市场和辅助服务市场的过程中,负荷预测的精度和速度成为影响各主体报价结果的瓶颈。负荷预测越准确,越有利于保障各市场主体报价的公平性和经济性。本文为解决该问题,选择南方电网某区域的历史负荷作为研究对象,通过对其日负荷曲线进行分析,考虑将工作日和非工作日的海量负荷数据进行了筛选和预处理,并针对各自的负荷特性进行了分析,确定了分别预测建模的预测路线,同时本文将当前常用的几种预测算法进行了比较,通过对比优缺点,针对超短期负荷预测的预测时间短、预测速度高的要求,最终选择负荷求导法作为超短期负荷预测的数学模型。最后通过对南网某省的实际负荷进行了算例验证,结果表明该方法具有预测速度快,预测精度高,适应度高,技术系统占用率低的特点。In the process of developing real-time spot market and ancillary service market,the accuracy and speed of load forecasting have become the bottleneck of influencing the bidding results of each main body.The more accurate the load forecasting is,the more conducive to ensuring the fairness and economy of the quotation of each market participant.In order to solve this problem,this paper chooses the historical load of a certain area of the south China power grid as the research object,through the analysis of its daily load curve,considers that the massive load data of working and non-working days are screened and preprocessed,and according to their respective load characteristics,the prediction models are determined respectively.At the same time,this paper compares several commonly used forecasting algorithms,compares their advantages and disadvantages,and finally chooses load derivation method as the mathematical model of ultra-short-term load forecasting to meet the requirements of short forecasting time and high forecasting speed in ultra-short-term load forecasting.Finally,the actual load of a province in south China power grid is verified by an example.The results show that the method has the characteristics of fast prediction speed,high prediction accuracy,high adaptability and low occupancy rate of technical system.

关 键 词：超短期负荷预测 海量数据 负荷求导法 实时现货市场 快速预测 

分 类 号：TM71[电气工程—电力系统及自动化]