A systematic review of multi-output prediction model for indoor environment and heating, ventilation, and air conditioning energy consumption in buildings
IF 3.2 3区 工程技术Q2 CONSTRUCTION & BUILDING TECHNOLOGY
Kaiyun Jiang, Tianyu Shi, Haowei Yu, N. Mahyuddin, Shifeng Lu
{"title":"A systematic review of multi-output prediction model for indoor environment and heating, ventilation, and air conditioning energy consumption in buildings","authors":"Kaiyun Jiang, Tianyu Shi, Haowei Yu, N. Mahyuddin, Shifeng Lu","doi":"10.1177/1420326x241258678","DOIUrl":null,"url":null,"abstract":"Heating, ventilation and air conditioning (HVAC) systems could significantly impact indoor environmental quality, particularly in terms of thermal comfort and indoor air quality. Achieving a high-quality indoor environment poses challenges to the energy consumption of HVAC systems. Thus, balancing thermal comfort, indoor air quality (IAQ) and energy consumption becomes a challenging task. Currently, indoor environment prediction methods are considered effective solutions to address this issue. However, the published literature usually concentrates on single aspects like thermal comfort, air quality or energy consumption, with multi-aspect prediction methods being rare. The present work reviews research spanning the last decade that employs machine learning methods for predicting indoor environments and HVAC energy consumption through separate and multi-output predictive models. Separate predictive models focus on HVAC systems’ impact on the indoor environment, while multi-output models consider the interplay of various outputs. This article gives a thorough insight into machine learning prediction models’ workflow, detailing data collection, feature selection and model optimization for each research goal. A systematic assessment of methods for data collection of diverse prediction targets, machine learning algorithms and validation approaches for different prediction models is presented. This review highlights the complexities of data management, model development and validation, enriching the knowledge base in indoor environmental quality optimization.","PeriodicalId":13578,"journal":{"name":"Indoor and Built Environment","volume":null,"pages":null},"PeriodicalIF":3.2000,"publicationDate":"2024-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Indoor and Built Environment","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.1177/1420326x241258678","RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"CONSTRUCTION & BUILDING TECHNOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
Heating, ventilation and air conditioning (HVAC) systems could significantly impact indoor environmental quality, particularly in terms of thermal comfort and indoor air quality. Achieving a high-quality indoor environment poses challenges to the energy consumption of HVAC systems. Thus, balancing thermal comfort, indoor air quality (IAQ) and energy consumption becomes a challenging task. Currently, indoor environment prediction methods are considered effective solutions to address this issue. However, the published literature usually concentrates on single aspects like thermal comfort, air quality or energy consumption, with multi-aspect prediction methods being rare. The present work reviews research spanning the last decade that employs machine learning methods for predicting indoor environments and HVAC energy consumption through separate and multi-output predictive models. Separate predictive models focus on HVAC systems’ impact on the indoor environment, while multi-output models consider the interplay of various outputs. This article gives a thorough insight into machine learning prediction models’ workflow, detailing data collection, feature selection and model optimization for each research goal. A systematic assessment of methods for data collection of diverse prediction targets, machine learning algorithms and validation approaches for different prediction models is presented. This review highlights the complexities of data management, model development and validation, enriching the knowledge base in indoor environmental quality optimization.
期刊介绍:
Indoor and Built Environment publishes reports on any topic pertaining to the quality of the indoor and built environment, and how these might effect the health, performance, efficiency and comfort of persons living or working there. Topics range from urban infrastructure, design of buildings, and materials used to laboratory studies including building airflow simulations and health effects. This journal is a member of the Committee on Publication Ethics (COPE).