基于廉价传感器的城市大气颗粒污染物监测系统
发布时间：2018-09-10 13:08
【摘要】：空气是我们共享的重要的生存资源,但是近年来某些区域的空气质量却呈现出逐渐恶化的趋势,尤其是对于一些发展中国家的超大型城市,如北京和新德里,空气污染已经成为这些城市一个很严重的问题。在空气污染的众多维度中,细粒度颗粒,尤其是PM2.5近来越来越受到人们的关注,其中很重要的原因在于PM2.5在人们身体健康的直接影响,有研究资料表明PM2.5粒子很容易被人体肺部器官吸收,长时间的PM2.5环境会导致呼吸系统甚至血液系统疾病,PM2.5相关的问题也是本文主要研究的内容。由于PM2.5的严重影响,引发了民众想要时刻的知道自身周围环境的空气质量指数,由此来做出相应的防护措施。目前有很多的手机应用,直接使用公共监测站的数值提供给用户,但是这些数值并不是用户更加关心的当前位置准确的数值,因为用户绝大部分时间所处场所的PM2.5的数值对其健康的影响才是用户关注所在。同时,政府也意识到了更密集部署监测设备的数据可能会对“科技治霾”起到很大的帮助作用。因此,本项目我们针对如何在控制成本保证精度的情况下进行PM2.5设备的大密度监测方案进行了研究。我们提出了Air Cloud系统,Air Cloud采用了clientcloud架构。硬件前端,我们设计开发了廉价联网式的PM2.5监测设备——基站式AQM和移动式mini AQM。我们设计了其中的风道和结构,确保可以准确一致的测到PM2.5的数值。云端我们设计了云端的数据校正处理框架,设备的数据将会得到实时校正。我们使用了5个月的实验室数据和2个月的实际部署数据,对云端算法和整体系统的性能进行了测试,经过试验,我们的云端校正框架可以将传感器的准确率提高53.6%。通过这种校正处理,我们在降低前端硬件成本的基础上,保证了准确率,使得大密度部署并保证数据准确性成了可能。我们后续在北京海淀进行了高密度的部署实验(200个基站式AQM设备)。我们对部署数据进行了统计分析并给出了一些有意义和初创的发现,这些发现对后续PM2.5在时空层面上传播路径还原,污染源区域发现包括时间维度上的预测都会起到一定的帮助作用,而这些结论将直接帮助政府作出相应的决策,以科学高效的实现降霾目标。
[Abstract]:Air is an important living resource we share, but in recent years there has been a gradual deterioration in air quality in some regions, especially in mega-cities in developing countries, such as Beijing and New Delhi. Air pollution has become a serious problem in these cities. Among the many dimensions of air pollution, fine-grained particles, especially PM2.5, have recently attracted more and more attention because of the direct effects of PM2.5 on people's health. Some research data show that PM2.5 particles are easily absorbed by human lung organs. The problems related to respiratory system and even blood system diseases caused by PM2.5 environment for a long time are also the main contents of this paper. Because of the serious influence of PM2.5, people want to know the air quality index of their surroundings all the time. At present, there are many mobile phone applications that directly use the values of public monitoring stations to provide users, but these values are not the exact values of the current location that the user is more concerned about. Because most of the time the user in the place of the value of PM2.5 on their health is the focus of user attention. At the same time, the government is aware that data from more intensive deployment of monitoring equipment could be of great help to "technology and haze control." Therefore, in this project, we study how to monitor the large density of PM2.5 equipment under the condition of controlling cost and guaranteeing precision. We propose that Air Cloud system adopts clientcloud architecture for Air Cloud. In the front end of hardware, we have designed and developed a cheap networked PM2.5 monitoring equipment, namely, base station AQM and mobile mini AQM.. We have designed the duct and structure to ensure that the PM2.5 values can be measured accurately and consistently. Cloud we designed cloud data correction processing framework, equipment data will be real-time correction. We have used five months of laboratory data and two months of actual deployment data to test the performance of the cloud algorithm and the overall system. After the experiment, our cloud correction framework can improve the sensor accuracy 53.6%. Through this correction process, we can reduce the cost of front-end hardware, ensure the accuracy, make it possible to deploy large density and ensure the accuracy of data. We conducted a high-density deployment experiment (200 base station AQM devices) in Haidian, Beijing. We make a statistical analysis of the deployment data and present some meaningful and initial findings that restore the propagation path of subsequent PM2.5 at the temporal and spatial level. The regional discovery of pollution sources, including the prediction of time dimension, will help the government directly to make corresponding decisions to achieve the goal of reducing haze scientifically and efficiently.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：X84
本文编号：2234535
[Abstract]:Air is an important living resource we share, but in recent years there has been a gradual deterioration in air quality in some regions, especially in mega-cities in developing countries, such as Beijing and New Delhi. Air pollution has become a serious problem in these cities. Among the many dimensions of air pollution, fine-grained particles, especially PM2.5, have recently attracted more and more attention because of the direct effects of PM2.5 on people's health. Some research data show that PM2.5 particles are easily absorbed by human lung organs. The problems related to respiratory system and even blood system diseases caused by PM2.5 environment for a long time are also the main contents of this paper. Because of the serious influence of PM2.5, people want to know the air quality index of their surroundings all the time. At present, there are many mobile phone applications that directly use the values of public monitoring stations to provide users, but these values are not the exact values of the current location that the user is more concerned about. Because most of the time the user in the place of the value of PM2.5 on their health is the focus of user attention. At the same time, the government is aware that data from more intensive deployment of monitoring equipment could be of great help to "technology and haze control." Therefore, in this project, we study how to monitor the large density of PM2.5 equipment under the condition of controlling cost and guaranteeing precision. We propose that Air Cloud system adopts clientcloud architecture for Air Cloud. In the front end of hardware, we have designed and developed a cheap networked PM2.5 monitoring equipment, namely, base station AQM and mobile mini AQM.. We have designed the duct and structure to ensure that the PM2.5 values can be measured accurately and consistently. Cloud we designed cloud data correction processing framework, equipment data will be real-time correction. We have used five months of laboratory data and two months of actual deployment data to test the performance of the cloud algorithm and the overall system. After the experiment, our cloud correction framework can improve the sensor accuracy 53.6%. Through this correction process, we can reduce the cost of front-end hardware, ensure the accuracy, make it possible to deploy large density and ensure the accuracy of data. We conducted a high-density deployment experiment (200 base station AQM devices) in Haidian, Beijing. We make a statistical analysis of the deployment data and present some meaningful and initial findings that restore the propagation path of subsequent PM2.5 at the temporal and spatial level. The regional discovery of pollution sources, including the prediction of time dimension, will help the government directly to make corresponding decisions to achieve the goal of reducing haze scientifically and efficiently.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：X84
【参考文献】
相关期刊论文 前2条
1 周涛;汝小龙;;北京市雾霾天气成因及治理措施研究[J];华北电力大学学报(社会科学版);2012年02期
2 尹洧;;大气颗粒物及其组成研究进展(下)[J];现代仪器;2012年03期
，本文编号：2234535
本文链接：https://www.wllwen.com/kejilunwen/huanjinggongchenglunwen/2234535.html
最近更新
教材专著
