通道奇异谱分析的绿色建筑能耗预测方法仿真  

Simulation of Green Building Energy Consumption Prediction Method Based on Channel Singular Spectrum Analysis

作　　者：丁晓红[1] 蒋雪峰 DING Xiao-hong;JIANG Xue-feng(Faculty of Architecture and Civil Engineering,Huaiyin Institute of Technology,Huai'an Jiangsu 223001,China;College of Architect and Urban Planning,Kunming University of Science and Technology,Kunming Yunnan 650500,China)

机构地区：[1]淮阴工学院建筑工程学院,江苏淮安223001 [2]昆明理工大学建筑与城市规划学院,云南昆明650500

出　　处：《计算机仿真》2024年第9期351-355,共5页Computer Simulation

基　　金：国家自然科学基金(51808246);江苏省产学研合作项目(BY2021405);江苏省住房和城乡建设厅科技项目(2018ZD312)。

摘　　要：建筑能耗受到外部环境、用户行为等因素的影响,这些因素难以准确预测或低频率更新,使得能耗预测存在一定的不确定性。为了得到高准确率的绿色建筑能耗预测结果,提出一种面向绿色建筑的能耗贝叶斯预测方法。通过多通道奇异谱分析(MSSA)对历史绿色建筑能耗数据去噪,将去噪后的数据作为训练样本。利用训练样本展开目标定位和搜索,对获取的目标展开特征提取,根据获取的目标特征和目标身份之间的关联关系建立能耗贝叶斯预测模型,通过相关判决条件,输出目标预测结果,完成绿色建筑能耗预测。仿真结果表明,所提方法可以获取高准确率的绿色建筑能耗预测结果,同时具有良好的数据去噪能力。Generally,building energy consumption is influenced by the external environment,user behavior,and other factors.However,these factors are difficult to predict accurately or update at a low frequency,resulting in uncertainty in energy consumption prediction.In order to achieve highly accurate energy consumption prediction for green buildings,this article presented a Bayesian method for predicting the energy consumption of green buildings.Firstly,Multi-channel Singular Spectrum Analysis(MSSA) was employed to denoise historical data about the energy consumption of green buildings.And then,these data were used as the training samples.Moreover,target localization and search were conducted using the training samples.Meanwhile,feature extraction was performed on the obtained targets.Furthermore,a Bayesian prediction model of energy consumption was built based on the correlation between target features and target identities.Finally,the target prediction results were output through relevant decision conditions,thus completing the energy consumption prediction of green buildings.Simulation results show that the proposed method can achieve highly accurate energy consumption prediction and has good data denoising capability.

关 键 词：绿色建筑 能耗预测 贝叶斯 多通道奇异谱分析 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]