Design And Implementation Of Air Quality Monitoring Systerm Based On High Density Deployment Sensors

In recent years, air pollution is becoming more and more serious especially in Beijing and Tianjin. Frequently large-scale smog has been a great concern of the general public and the government. Environmental monitoring is the premise and foundation of environmental protection. The automated air monitoring network constructed in China in the 1990 s has formed a certain scale, whereas it has disadvantages in the sparse deployment and special need of personnel for the maintenance which may be another significant expense. The existing network monitoring data which is mainly used for environmental assessment cannot provide a valid basis for the scientific analysis of the atmospheric environment and governance. With the emergence of new monitoring technology, there are many company monitoring air pollutions on the basis of high density deployment sensors.This paper is based on thousands of small monitoring equipment deployed in Beijing. The air quality monitoring system designed and implemented in the paper is on the foundation of high-density deployment sensors, bringing the new ideas of the Internet of things to solve the traditional problems of environmental monitoring. Monitoring devices use GPRS network to upload the collection data. With the crawled meteorological data, the paper built a high-performance and highly reliable basic data services platform based on the two open source frameworks Tornado and Celery. On this platform, we made use of the SSM framework to design and implement a visual system based on the specific needs of environmental protection.A method for reducing pollution propagation path using the high density data collected by sensors has also been proposed in this paper, which analyzed the suspected pollution location during this period by the generating pollution spread directed graph. At the same time, an area prediction method based on the way of large density sensor deployment in the conventional single-point PM2.5 predictions is provided. Extensive experiments based on the real datasets have compared the deffirent method for the accuracy of prediction, and rendered dynamic prediction heat map by Gaussian process, which gives a detailed analysis of the area forecast.The system is stable after a series of functional and performance tests, and it can provide real and effective theoretical support for the relevant department to improve the atmospheric environment, and achieves the desired results.