基于廉價傳感器的城市大氣顆粒污染物監(jiān)測系統(tǒng)
發(fā)布時間：2018-09-10 13:08
【摘要】：空氣是我們共享的重要的生存資源,但是近年來某些區(qū)域的空氣質(zhì)量卻呈現(xiàn)出逐漸惡化的趨勢,尤其是對于一些發(fā)展中國家的超大型城市,如北京和新德里,空氣污染已經(jīng)成為這些城市一個很嚴重的問題。在空氣污染的眾多維度中,細粒度顆粒,尤其是PM2.5近來越來越受到人們的關(guān)注,其中很重要的原因在于PM2.5在人們身體健康的直接影響,有研究資料表明PM2.5粒子很容易被人體肺部器官吸收,長時間的PM2.5環(huán)境會導致呼吸系統(tǒng)甚至血液系統(tǒng)疾病,PM2.5相關(guān)的問題也是本文主要研究的內(nèi)容。由于PM2.5的嚴重影響,引發(fā)了民眾想要時刻的知道自身周圍環(huán)境的空氣質(zhì)量指數(shù),由此來做出相應(yīng)的防護措施。目前有很多的手機應(yīng)用,直接使用公共監(jiān)測站的數(shù)值提供給用戶,但是這些數(shù)值并不是用戶更加關(guān)心的當前位置準確的數(shù)值,因為用戶絕大部分時間所處場所的PM2.5的數(shù)值對其健康的影響才是用戶關(guān)注所在。同時,政府也意識到了更密集部署監(jiān)測設(shè)備的數(shù)據(jù)可能會對“科技治霾”起到很大的幫助作用。因此,本項目我們針對如何在控制成本保證精度的情況下進行PM2.5設(shè)備的大密度監(jiān)測方案進行了研究。我們提出了Air Cloud系統(tǒng),Air Cloud采用了clientcloud架構(gòu)。硬件前端,我們設(shè)計開發(fā)了廉價聯(lián)網(wǎng)式的PM2.5監(jiān)測設(shè)備——基站式AQM和移動式mini AQM。我們設(shè)計了其中的風道和結(jié)構(gòu),確�？梢詼蚀_一致的測到PM2.5的數(shù)值。云端我們設(shè)計了云端的數(shù)據(jù)校正處理框架,設(shè)備的數(shù)據(jù)將會得到實時校正。我們使用了5個月的實驗室數(shù)據(jù)和2個月的實際部署數(shù)據(jù),對云端算法和整體系統(tǒng)的性能進行了測試,經(jīng)過試驗,我們的云端校正框架可以將傳感器的準確率提高53.6%。通過這種校正處理,我們在降低前端硬件成本的基礎(chǔ)上,保證了準確率,使得大密度部署并保證數(shù)據(jù)準確性成了可能。我們后續(xù)在北京海淀進行了高密度的部署實驗(200個基站式AQM設(shè)備)。我們對部署數(shù)據(jù)進行了統(tǒng)計分析并給出了一些有意義和初創(chuàng)的發(fā)現(xiàn),這些發(fā)現(xiàn)對后續(xù)PM2.5在時空層面上傳播路徑還原,污染源區(qū)域發(fā)現(xiàn)包括時間維度上的預測都會起到一定的幫助作用,而這些結(jié)論將直接幫助政府作出相應(yīng)的決策,以科學高效的實現(xiàn)降霾目標。
[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.
【學位授予單位】：哈爾濱工業(yè)大學
【學位級別】：碩士
【學位授予年份】：2015
【分類號】：X84
本文編號：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.
【學位授予單位】：哈爾濱工業(yè)大學
【學位級別】：碩士
【學位授予年份】：2015
【分類號】：X84
【參考文獻】
相關(guān)期刊論文 前2條
1 周濤;汝小龍;;北京市霧霾天氣成因及治理措施研究[J];華北電力大學學報(社會科學版);2012年02期
2 尹洧;;大氣顆粒物及其組成研究進展(下)[J];現(xiàn)代儀器;2012年03期
，本文編號：2234535
本文鏈接：http://sikaile.net/kejilunwen/huanjinggongchenglunwen/2234535.html
最近更新
教材專著
