基于MLR-GA-WNN公共建筑能耗预测研究  

Building Energy Consumption Forecast Based on MLR-GA-WNN

作　　者：叶永雪 马鸿雁 李晟延 YE Yong-xue;MA Hong-yan;LI Sheng-yan(School of Electrical and Information Engineering,Beijing University of Civil Engineering and Architecture,Beijing 100044,China;Beijing Key Laboratory of Intelligent Processing for Building Big Data,Beijing 100044,China;National Virtual Simulation Experimental Center for Smart City Education,Beijing 100044,China)

机构地区：[1]北京建筑大学电气与信息工程学院,北京100044 [2]建筑大数据智能处理方法研究北京市重点实验室,北京100044 [3]智慧城市国家级虚拟仿真实验教学中心,北京100044

出　　处：《计算机仿真》2023年第3期514-519,共6页Computer Simulation

基　　金：北京建筑大学博士基金项目(ZF15054)。

摘　　要：为解决现有公共建筑能耗预测中因数据量少、样本维度高导致的精度低、误差大问题,提出一种基于多元线性回归(Multiple linear regression, MLR)与遗传算法(Genetic algorithm, GA)优化小波神经网络(Wavelet neural network, WNN)的建筑能耗预测模型。利用Pearson相关系数分析方法与多元线性回归对历史数据进行预处理,选取相关性强的因素用于GA-WNN模型的训练与测试,构成MLR-GA-WNN建筑能耗预测模型,精度达到了96.4%。仿真结果表明,文中提出的方法不但预测精度优于WNN、GA-WNN、GA-BP与GA-SVM模型,而且仿真运行时长、误差也均小于上述四种模型,验证了提出模型对于公共建筑能耗预测的可行性。In order to solve the problems of low accuracy and large error caused by the small amount of data and high sample dimension in existing public building energy consumption prediction,this paper proposes a building energy consumption prediction model based on multiple linear regression(MLR)and genetic algorithm(GA)to optimize wavelet neural network(WNN)for predicting building energy consumption.The MLR-GA-WNN building energy consumption prediction model was formed by using Pearson correlation coefficient analysis method and multiple linear regression to pre-process historical data and selecting factors with strong correlation for the training and testing of the GA-WNN model,and the accuracy of the model reached 96.4%.The simulation results show that the proposed method outperforms the WNN,GA-WNN,GA-BP and GA-SVM models in terms of prediction accuracy,and the simulation run time and error are smaller than those of the above four models,which verifies the feasibility of the proposed model for predicting energy consumption in public buildings.

关 键 词：数据处理 多元线性回归 遗传算法 小波神经网络 能耗预测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]