基于GCA-MVO-ICA优化BP的负荷预测研究  

Load Prediction Study of Optimized BP Based on GCA-MVO-ICA

作　　者：王忠峰 王智 Wang Zhongfeng;Wang Zhi(Jilin Machinery Special Industry Co.,Ltd.,Jilin,China;Harbin Nuoxin Gongda Measurement and Control Technology Co.,Ltd.,Harbin,China)

机构地区：[1]吉林江机特种工业有限公司,吉林吉林 [2]哈尔滨诺信工大测控技术有限公司,黑龙江哈尔滨

出　　处：《科学技术创新》2024年第10期211-214,共4页Scientific and Technological Innovation

摘　　要：为提高微网负荷预测的精度,提出来一种利用灰色关联度分析优化BP神经网络优化输入层数据,利用新型的多元宇宙算法优化隐含层权重,采用帝国竞争算法优化输出层结果的逐层优化模型,对2组实测数据算例分析。结果表明,所提GCA-MVO-ICA优化BP网络的方法能够提高微网负荷的预测精度,并且具有较好的普适性。To enhance the accuracy of microgrid load forecasting,a method is proposed that utilizes grey correlation analysis to optimize the input layer data of the BP neural network,employs a novel multidimensional universe algorithm to optimize the hidden layer weights,and utilizes an imperial competitive algorithm to optimize the output layer results in a layer-by-layer optimization model.The method is applied to the analysis of two sets of actual measurement data.The results indicate that the proposed GCA-MVO-ICA optimization method for the BP network can improve the prediction accuracy of microgrid load and exhibits good universality.

关 键 词：BP神经网络 灰色关联度 多元宇宙算法 帝国竞争算法 负荷预测 

分 类 号：TM76[电气工程—电力系统及自动化]