Title: Automatic detection method of atmospheric pollutant concentration based on multi-sensor data fusion
Authors: Mingchuan Meng; Dawei Lu; Jiong Zhao; Wei Zhang; Xuehao Zhou
Addresses: Faculty of Electrical Engineering and Information, Qinggong College, North China University of Science and Technology, Tangshan, 063000, China ' Faculty of Electrical Engineering and Information, Qinggong College, North China University of Science and Technology, Tangshan, 063000, China ' Faculty of Electrical Engineering and Information, Qinggong College, North China University of Science and Technology, Tangshan, 063000, China ' Faculty of Electrical Engineering and Information, Qinggong College, North China University of Science and Technology, Tangshan, 063000, China ' Faculty of Electrical Engineering and Information, Qinggong College, North China University of Science and Technology, Tangshan, 063000, China
Abstract: Aiming at the problem that the traditional air pollutant concentration detection method cannot accurately detect the concentration of multiple pollutants, an automatic detection method of air pollutant concentration based on multi-sensor data fusion is proposed. A multi-sensor environment detection system is constructed by the temperature and humidity sensor DHT22, the NO2 sensor ZE08-CH2O and the dust sensor ZPH01, and the multi-sensor data fusion is realised by the Kalman filter algorithm. The improved BP network is used to classify pollution source data, and the detection data of the same pollution source is back calculated by adaptive simulated annealing algorithm to realise the automatic detection of the concentration of similar air pollutants. The experimental results show that the detection errors of this method for the concentration of atmospheric pollutants NO2, TVOC and PM2.5 inhalable particulate matter are 0.03 μg·m-3, 0.02 μg·m-3, 0.03 μg·m-3 respectively, which proves the effectiveness of this method.
Keywords: multi-sensor; data fusion; air pollutants; concentration detection.
International Journal of Data Science, 2022 Vol.7 No.4, pp.365 - 382
Received: 13 Jun 2022
Accepted: 03 Sep 2022
Published online: 19 Jan 2023 *