Industrial Park Oriented Monitoring System Design Of Pollutant Gas Emissions

The rapid development of industry leads to different kinds of pollutions, especially in industrial parks and their surroundings, which should aggravate environmental pollution and even result in the frequent pollution accidents. Industrial park, the key area to exhausting polution gases, causing serious dispersal of pollution, will finally arouse public complaints. Various types of enterprises in the park will produce amount of air pollutants with different complex components. Therefore, it is important to design a smart system for the purpose of monitoring pollution in industrial park.After describing basic concepts of wireless sensor networks, the thesis designed three modules, including perception module, processor module and communication module. To meet requirements of application and performance for monitoring system, we showed hardware system and related communication protocol.For the purpose of monitoring exhaust gas, we exploited Gaussian Plume Model to analyze gas diffusion. According to this, a monitoring sensor distribution scheme was proposed, which could remove redundant sensors by adopting sensor correlation coefficients in order to reduce additional equipment and communication costs. For case of multiple emission sources, we built source strength inverse model, aiming to transform it into an optimization problem to find its optimal solution. Based on PSO, we put forward an improved Adaptive Iteration Times Random Particle Swarm Optimization (AIT-SPSO) algorithm to improve the global search capability as well as reduce computational complexity, which also could quickly and accurately obtain the source strength value of every emission point. In addition, adopting BP neural network training and its historical data gathered from adjacent topology of sensor network, we could be easy to determine the zero drift of sensors by our proposed algorithm.Moreover, we employed SpringMVC+Hiberaate as the underlying development framework to implement the JAVA-based application software for Industrial Park Monitoring System, which was also a lightweight system developed by Eclipse platform. Considering the capability of object data persistence, the on-demand database was designed via Hibernate technology. Finally, we also introduced the development of post-optimization software system testing, so as to improve the system scalability.