基于云计算和极限学习机的分布式电力负荷预测算法.pdfVIP

第 38 卷 第 2 期 电 网 技 术 Vol. 38 No. 2 2014 年 2 月 Power System Technology Feb. 2014 1000-3673 （2014 ）02-0526-06 TM 76 A 470·4054 071003 A Distributed Load Forecasting Algorithm Based on Cloud Computing and Extreme Learning Machine WANG Baoyi, ZHAO Shuo, ZHANG Shaomin (School of Control and Computer Engineering, North China Electric Power University, Baoding 071003, Hebei Province, China) ABSTRACT: To improve the accuracy of load forecasting and 模型的负荷预测精度均优于传统支持向量回归预测算法和 cope with the challenge of single computer’s insufficient 泛化神经网络预测算法，且提出的算法具有优异的并行性能。 computing resource under massive and high-dimension data 云计算；负荷预测；极限学习机；在线序列优化 due to power grid intellectualization, a short-term distributed DOI ：10.13335/j.1000-3673.pst.2014.02.039 load forecasting model based on cloud computing and extreme learning machine is proposed. According to the features of load 0 data the online sequential optimization of load forecasting 负荷预测一直以来都是电力系统的一个关键 algorithm based on extreme learning machine is performed; 操作和规划途径，它影响着电力系统的诸多决策， leading in the distributed and multi-agent thinking the accuracy 如经济调度、自动发电控制、安全评估、维护调度 of load forecasting algorithm is improved; adopting the 和能源商业化。精准的电力负荷预测可以经济合理 MapReduce programming framework of cloud computing the parallelization of the model of the proposed algorithm is 地安排电力系统发电机组启停，对于保持电网运行 carried out to enhance its ability of dealing with massive and 的