Využití dronů pro monitorování kvality ovzduší a detekci zdrojů znečištění
Loading...
Date
Authors
Aung Latt, Htet
ORCID
Advisor
Referee
Mark
B
Journal Title
Journal ISSN
Volume Title
Publisher
Vysoké učení technické v Brně. Fakulta strojního inženýrství
Abstract
V této studii jsme zhodnotili několik dostupných senzorů na trhu pro levné senzory částic, přičemž jsme se zaměřili především na jejich velikost a hmotnostní faktory. Diskutováno bylo několik metod identifikace zdroje, jako je inverzní modelování, modelování dronů a senzorová síť. Mezi popsanými senzory byl vybrán senzor PM Novafitness SDS011 kvůli jeho přesnosti a snadné implementaci. Spolu s ním byl vybrán mikrokontrolér WeMos D1 mini s rozšířením na kartu MicroSD. S využitím ThingSpeak a SD karty pro ukládání dat byla provedena analýza dat pomocí ThingSpeak a MATLABu. Nejprve je levný senzor PM testován v porovnání s laboratorním čítačem PM, TSI APS 3321, v aerosolové laboratoři s použitím částic okolního vzduchu a vodních částic produkovaných nebulizérem. Levný senzor PM nadhodnocuje hmotnostní koncentraci až o 27 % pro PM2,5 a o 58 % pro PM10 s použitím částic okolního vzduchu. U vyšších koncentrací při použití vodního roztoku nízkonákladový senzor podhodnocuje PM2,5 až o 50 %, ale dochází k znatelnému zvýšení přesnosti koncentrací PM10, a to až o 3,7 %. Dále je funkčnost systému testována venku. Pro detekci zdrojů znečištění je dron vybaven senzorem PM a vznáší se v blízkosti větrání vzduchu. Jako kontrolní nastavení je dron umístěn v blízkosti výstupu větrání vzduchu na zemi. Hmotnostní koncentrace jsou v obou případech nižší než při letu v okolním vzduchu. A z pozemního větracího otvoru je zaznamenána mnohem nižší hmotnostní koncentrace. To souhlasí s hypotézou, že větrání vzduchu je vybaveno filtračními systémy, a proto tam musí být koncentrace nižší. Dále je vliv rotorů dronu na hmotnostní koncentraci testován uvnitř, a to letem dronu ve výšce 1 m a umístěním dronu do výšky 1 m s vypnutými rotory. Zjistili jsme, že senzor podhodnocuje koncentraci, když jsou rotory zapnuté, o -10,8 % a -38,7 % pro PM2,5 a PM10. Nakonec byla diskutována celková omezení a výhody používání dronů v oblasti monitorování kvality ovzduší, spolu s jejich budoucími implementacemi a experimenty.
In this study, we reviewed several available sensors in the market for low-cost particle sensors focusing mainly on their size and weight factors. A few methods of source identification methods such as inverse modelling, drone modelling and sensor network were discussed. Among the described sensors, Novafitness SDS011 PM sensor was chosen due to its accuracy, and ease of implication. Along with it, WeMos D1 mini was chosen for microcontroller with MicroSD card extension. Utilizing ThingSpeak and SD card for storing data, data analysis was done using ThingSpeak and MATLAB. First, the low-cost PM sensor is tested against a laboratory grade PM counter, TSI APS 3321, in an aerosol laboratory using ambient air particles and water particles produced by a nebulizer. The low-cost PM sensor overestimates the mass concentration up to 27% for PM2.5 and 58% for PM10 using ambient air particles. For higher concentrations using the water solution, the low-cost sensor underestimates PM2.5 by up to 50% but a noticeable increase in accuracy of PM10 concentrations, as low as 3.7%. Next, the functionality of the system is tested outdoors. For detection of pollution sources, the drone is equipped with the PM sensor and hovers near the air ventilation. As a control setup, the drone is placed near an air ventilation outlet on the ground. The mass concentrations for both cases are lower than when flying in ambient air. And much lower mass concentration is recorded from the ground air vent. This agrees with the hypothesis that air ventilations are equipped with filtration systems and therefore the concentration there must be lower. Next the effect from the drone’s rotors on the mass concentration is tested indoors by flying the drone at 1m height and placing the drone at 1m height with its rotors turned off. We found that the sensor underestimates the concentration when the rotors are turned on by -10.8% and -38.7% for PM2.5 and PM10 respectively. Finally, the overall limitations and advantages of using drones in area of air quality monitoring were discussed, alongside its future implementations and experiments.
In this study, we reviewed several available sensors in the market for low-cost particle sensors focusing mainly on their size and weight factors. A few methods of source identification methods such as inverse modelling, drone modelling and sensor network were discussed. Among the described sensors, Novafitness SDS011 PM sensor was chosen due to its accuracy, and ease of implication. Along with it, WeMos D1 mini was chosen for microcontroller with MicroSD card extension. Utilizing ThingSpeak and SD card for storing data, data analysis was done using ThingSpeak and MATLAB. First, the low-cost PM sensor is tested against a laboratory grade PM counter, TSI APS 3321, in an aerosol laboratory using ambient air particles and water particles produced by a nebulizer. The low-cost PM sensor overestimates the mass concentration up to 27% for PM2.5 and 58% for PM10 using ambient air particles. For higher concentrations using the water solution, the low-cost sensor underestimates PM2.5 by up to 50% but a noticeable increase in accuracy of PM10 concentrations, as low as 3.7%. Next, the functionality of the system is tested outdoors. For detection of pollution sources, the drone is equipped with the PM sensor and hovers near the air ventilation. As a control setup, the drone is placed near an air ventilation outlet on the ground. The mass concentrations for both cases are lower than when flying in ambient air. And much lower mass concentration is recorded from the ground air vent. This agrees with the hypothesis that air ventilations are equipped with filtration systems and therefore the concentration there must be lower. Next the effect from the drone’s rotors on the mass concentration is tested indoors by flying the drone at 1m height and placing the drone at 1m height with its rotors turned off. We found that the sensor underestimates the concentration when the rotors are turned on by -10.8% and -38.7% for PM2.5 and PM10 respectively. Finally, the overall limitations and advantages of using drones in area of air quality monitoring were discussed, alongside its future implementations and experiments.
Description
Citation
AUNG LATT, H. Využití dronů pro monitorování kvality ovzduší a detekci zdrojů znečištění [online]. Brno: Vysoké učení technické v Brně. Fakulta strojního inženýrství. 2025.
Document type
Document version
Date of access to the full text
Language of document
en
Study field
bez specializace
Comittee
prof. Ing. Pavel Novotný, Ph.D. (předseda)
doc. Ing. František Lízal, Ph.D. (místopředseda)
doc. Ing. Pavel Charvát, Ph.D. (člen)
doc. Ing. Lubomír Klimeš, Ph.D. (člen)
prof. Ing. Josef Štětina, Ph.D. (člen)
Date of acceptance
2025-06-12
Defence
The student informed the examination committee about the methodology, results, and conclusions of his bachelor thesis. He answered all questions posed by the reviewer.
Prof. Novotný inquired about the position of the sensor and whether the concentration of particles would change if the position were altered. The student responded that this aspect was not part of the thesis.
Doc. Charvát commented that the results regarding particle concentration with the rotor on and off seemed counterintuitive. The student explained the observed phenomenon.
Prof. Novotný further asked whether the thrust from the rotor could influence the number or size of particles. The student demonstrated good knowledge of the topic and engaged in a thorough discussion with the committee.
Result of defence
práce byla úspěšně obhájena
Document licence
Standardní licenční smlouva - přístup k plnému textu bez omezení