Development of indoor air quality prediction model and management tools
| Autor: | Jeffrey, Sy-Yuan KANG, 康思源 |
|---|---|
| Rok vydání: | 2016 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 105 This study is developed three self-management systems to solve current management problems of the indoor air quality, including the wireless indoor air quality monitor system, air pollution automatic broom system and the prediction model of air pollutant concentration. In the different monitor systems, NDIR sensing module can be used for continuous monitoring in the workplace, because the estimated value of liner correlation analysis and average error of ZG01 and AZ7722 sensing module are accepted. At high concentration, CTS-M5 sensing module has better accuracy, it can provide the continuous monitoring for caution. Although PPD42NS sensing module can measure all selected concentration, but it has slightly higher detection limit for real field monitoring. Because of handheld indoor air pollution automatic broom system (Air Pollution Automatic Broom System) development, we can connect the each point data that direct reading instrument or wireless monitoring systems measured to become lines and surfaces, and can be established indoor air pollution map and marked the position of serious pollution (Hot spot), and we can use the gradient color change to visualize the data of pollution. The indoor produce source is carbon dioxide, our effect factors of indoor air quality pollutant concentrations prediction model include the number of indoor people, outdoor and indoor ventilation and time, the volume of indoor space and outdoor CO2 concentration. The model estimated result of MAPE is 7.7%, the MAPE value lower than 10%, indicating that carbon dioxide prediction model has excellent accuracy. The effect factors of CO, PM10 and PM2.5 prediction model include indoor and outdoor ventilation and time, indoor space volume and outdoor concentration. MAPE estimated results are 34%, 28% and 25% respectively. The prediction model accuracy of CO, PM10 and PM2.5 is accepted, when MAPE value ranged from 20% to 50%. In this study, we use wireless systems to connect all monitoring equipment in the workplace, the monitoring value can be uploaded to the cloud to analyze with the artificial intelligence. Handheld indoor air pollution automatic broom system (Air Pollution Automatic Broom System) development, we can connect the each point data from direct reading instrument or wireless monitoring systems measured to become lines and surfaces, and can be established indoor air pollution map and marked the position of serious pollution (Hot spot), and we can use the gradient color change to visualize the data of pollution. Using indoor air quality pollutant concentrations prediction model, we can estimate immediate and future concentrations of pollutants in the space and can be provided improvement strategy and management decision-making to improve indoor air quality. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|