| Building intelligent system is a significant means to realize the functions of perception,decision making and reasoning.While the building operators not only concern about providing high quality service in the process of operation and maintenance,but also focus on the automation,intelligence,modularization,low energy consumption and less environmental impact.Aiming at the problems of building indoor environment monitoring,a system platform with cloud pipe and terminal is developed.Its main works include data collection,data acquisition,software development,the research of prediction model et al.The specific works are organized as follows.(1)The development status of building intelligent system is illustrated in this thesis.It is a key issue that system performance will improve by using cloud platform.In addition,the performance of building intelligent system,such as ubiquitous access,intelligent management,extensible and reproducible,are improved via cloud platform.(2)The requirement analyses of building indoor environment monitoring system are carried out and the system composition is put forward.Besides,the freamwork of the system is proposed and the functions of each layers are explained.The system adopts the mode of cloud,pipe and terminal,which consists of sensing layer,networking layer,function layer,modularized service layer and application layer.(3)The hardware system is built by the teconology of Internet of Things(IoT)for realizing real time data collection and uploading.The system adopts ESPDuino development platform based on Arduino development board and ESP8266 EX wireless chip.Monitoring modules of indoor environment system include temperature,humidity,indoor particulate matter 2.5(PM2.5),formaldehyde,carbon dioxide,personnel status information and energy consumption are designed and developed.Furthermore,the dynamic networking configuration is finished for providing comprehensive perception of the physical layer.(4)The data structure of the system is analyzed logically and the cloud database is built by PHP programing.The Web is designed and developed by LNMP solution(Linux,Nginx,MySQL and PHP)based on B/S(Browser/Server)structure.It contains environmental monitoring,personnel monitoring,energy consumption monitoring,building information management and other modules.PHP programming takes the responsibility of Web developing.Combined with the front-end frame(Bootstrap)and the chart frame(Highcharts),the software visualizes the data,such as temperature,humidity,indoor particulate matter,formaldehyde,carbon dioxide,personnel status information and energy consumption,according to the process of data collection,data analysis and data presentation.An App is also carried out,which enables end-users to query,control and manage from mobile terminals.Based on the frame of MVC(model view controller),the development transplants all functions from Web to App via initializing webview widget.(5)The prediction of indoor air quality and energy consumption plays an important role in system operation and management.This thesis adopts the extreme learning machine(ELM)for indoor PM2.5 concentration data and a public building energy consumption data for prediction research respectively.The prediction model is trained and established.The simulation proves the effectiveness of the algorithm and the research is instructive to the building optimization control.Mentioned contributions are progressive and targeted.Firstly,a feasible freamwork of building indoor environment monitoring based on IoT is presented.Secondly,the function of data collection and data acquisition is realized by using the hardware system and the cloud database.Thirdly,the services-oriented application software,such as Web terminal and App,are set up by LNMP solution based on cloud platform.Besides,environmental monitoring modules,personnel monitoring modules and energy consumption monitoring modules are implemented.Finally,the key issue of the system intelligence is data analysis and application.The prediction model of PM2.5 and building energy consumption was built by using ELM.The simulation proves that the algorithm is valid and the training speed is swift.These researches can be used as reference for the development of building Internet of Things,the application of cloud platform and the intelligent management of building operation. |
