MIV-PSO-BP神经网络用户热负荷预测  

Prediction of User Heat Load with MIV-PSO-BP Neural Network

作　　者：王新雨 于丹[1] 刘益民 崔治国 岑悦 WANG Xinyu;YU Dan;LIU Yimin;CUI Zhiguo;CEN Yue

机构地区：[1]北京建筑大学,北京100044 [2]中国建筑科学研究院有限公司,北京100013

出　　处：《煤气与热力》2022年第2期V0001-V0003,V0010,共4页Gas & Heat

基　　金：中国建筑科学研究院青年科研基金项目"建筑供热系统仿真及控制策略评价工具的研究与开发"(20200109331030019)。

摘　　要：提出利用MIV-PSO-BP神经网络预测用户热负荷。MIV-PSO-BP神经网络基于BP神经网络,利用PSO算法优化神经网络初始参数,采取MIV算法筛选与输出变量相关程度最大的输入变量。以绝对误差、均方根误差作为指标,评价MIV-PSO-BP神经网络的预测效果。结合箱线图,比较BP神经网络、MIV-PSO-BP神经网络的预测相对误差分散程度与异常点数量。与BP神经网络相比,MIV-PSO-BP神经网络的预测效果更理想。由BP神经网络、MIV-PSO-BP神经网络的预测结果相对误差箱线图可知,MIV-PSO-BP神经网络预测结果相对误差集中,异常点少;BP神经网络预测结果相对误差分散,异常点多。MIV-PSO-BP神经网络预测准确性、稳定性更高。It is proposed to predict user heat load by using MIV-PSO-BP neural network.The MIV-PSO-BP neural network is based on the BP neural net-work,it uses the PSO algorithm to optimize the initial parameters of the neural network,and adopts the MIV algorithm to screen the input variables that have the greatest correlation with the output variables.Taking absolute error and root mean square error as indicators,the prediction effect of MIV-PSO-BP neural network is evaluated.Combined with the box plot,the relative error dispersion degree and the number of abnormal points of BP neural network and MIV-PSO-BP neural network are compared.Compared with the BP neural network,the prediction effect of the MIV-PSO-BP neural network is more ideal.According to the relative error box plot of the prediction results of the BP neural network and the MIV-PSO-BP neural network,it can be seen that the relative errors of prediction results of the MIV-PSO-BP neural network are concentrated and the abnormal points are few.The relative errors of prediction results of the BP neural network are scattered,and there are many abnormal points.MIV-PSO-BP neural network has higher prediction accuracy and stability.

关 键 词：热负荷 预测 BP神经网络 MIV-PSO-BP神经网络 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]