Wireless Sensing Technology Application for Developing Indoor Air Quality Monitoring System

• Main Authors: Tsnag Chu Yu, 游蒼柜
• Other Authors: C. C. Lin
• Format: Others
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/87453740719599709339

碩士 === 長庚大學 === 資訊工程學系 === 99 === Individuals are becoming progressively more aware of health and the environment issues. In response to community concerns, government policy is increasingly focused on air quality monitoring as a means of measuring environmental health. This paper describes the construction of an air quality monitoring system that employs wireless sensor networks (WSN) in conjunction with cloud computing technology. The advantage of the system is the multifunctional capability that supports predicting for future trends in carbon dioxide using an ARIMA model, while integrating the forecast results into a fuzzy controller, which enables decision making and analysis. The flexibility of the system can be enhanced using a pyramid function update mechanism comprised of three components that include parameters adjustment, functional trimming, and program reconstruction. The benefit associated with employing a CMMBCR algorithm to calculate the transmission path of the sensors is a mechanism for determining the shortest path, which not only increases efficiency but also decreases the overall energy requirements, effectively extending the period of time the sensors can operate. The transmission period and energy use of the system is optimized by employing the SPIN method to transmit data, ensuring that retention of packets and broadcasting data is achieved. Results from the study have demonstrated a number of advantages from the use of the multifunctional air quality monitoring system. The advantages are associated with the software update mechanism for WSNs, a reduction in code image size, optimization of parameter adjustment, the accuracy of the ARIMA prediction model and reduced operational energy requirements. The software update mechanism employed in the study is comparable to Deluge when sending 12 pages in a 5 x 5 network topology with a varying rate of packet loss, while reducing transmission time by 19% and overall energy use by 21%. Achieved reductions in code image size from the system method were approximately 83.3%, with the transmission time and energy savings made in a 4 x 1 network topology recorded at 80%. The amount of transmission data was reduced from 26,496 bytes to 22 bytes via adjustment of the parameters used in the system. Furthermore, by calculating the average carbon dioxide concentration over 50 minutes and then extrapolating the data to predict the next 10 minutes using the ARIMA prediction model, accuracies of 96% were achieved. Investigations into the energy saving potential of the system using simulation boxes to replicated daily work situations revealed energy savings of 55% for an optimal working environment.