An IOT based Novel approach to predict Air Quality Index (AQI) using Optimized Bayesian Networks


Krishna Chaitanya Atmakuri,Y Venkata Raghava Rao,



Bayesian Classification Algorithm,IOT,Air Quality Index,Data Pre-processing,


As the size of the air quality data increases, it is difficult toforecastthe air quality metrics due to the non-stationary and randomization form of data distribution. Air quality prediction refers to the problem of finding the air quality by using statistical inference measures. However, traditional air prediction models are based on static fixed parameters for quality prediction. Also, it is difficult to classify and predict the air quality index for both rural and urban areas due to change in data drift and distribution. PM2.5 is one of the major factor to predict the air quality index (AQI) and its severity level. Due to high noisy and outliers in the PM2.5 data, it is difficult to classify and predict the air quality by using the traditional quality prediction models. In order to overcome these issues, an optimized Bayesian networks based probabilistic inference model is designed and implemented on the air quality data. An IOT enabled Air pollution monitoring system includes a DSM501A Dust sensor which detects PM2.5, PM1.0, MQ series sensor interfaced to a Node MCU equipped with ESP32 WLAN adaptor to send the sensor reading to Thing Speak cloud. In the proposed model, the data is initially gathered from the ICAO records of Safdarjung weather station and pre-processed.An improved discrete and continuous parameter estimation and bayes score optimization are implemented on the air quality prediction process. Experimental results show that the present optimized Bayesian network classify and predicts the air quality data with high less computational error rate and high accuracy. Further the proposed optimized model is applied on the real data which is gathered using IOT enabled gas sensors and the model is giving best results in predicting the air quality Index.


I.Ayaskanta Mishra, Air Pollution Monitoring System based on IoT: Forecasting and Predictive Modeling using Machine Learning”, International Conference on Applied Electromagnetics, Signal Processing and Communication (AESPC), 22nd -24th October-2018, Bhubaneswar, Odisha, India, IEEE, Paper ID# 9.

II.C. Li and Z. Zhu, “Research and application of a novel hybrid air quality early-warning system: A case study in China”, Science of The Total Environment, vol. 626, pp. 1421-1438, 2018. Available: 10.1016/j.scitotenv.2018.01.195 [Accessed 20February 2019].

III.Hybrid improved differential evolution and wavelet neural network with load forecasting problem of air conditioning Int. J. Electr. Power Energy Syst. 61, 673–682IV.H. Li, J. Wang, R. Li and H. Lu, “Novel analysis–forecast system based on multi-objective optimization for air quality index”, Journal of Cleaner Production, vol. 208, pp. 1365-1383, 2019. Available: 10.1016/j.jclepro.2018.10.129 [Accessed 20 February 2019.V.

VI.K. Gan, S. Sun, S. Wang and Y. Wei, “A secondary-decomposition-ensemble learning paradigm for forecasting PM2.5 concentration”, Atmospheric Pollution Research, vol. 9, no. 6, pp. 989-999, 2018. Available: 10.1016/j.apr.2018.03.008 [Accessed 20 February 2019.

VII.S. Feng, F. Jiang, Z. Jiang, H. Wang, Z. Cai and L. Zhang, “Impact of 3DVAR assimilation of surface PM 2.5 observations on PM 2.5 forecasts over China during wintertime”, Atmospheric Environment, vol. 187, pp. 34-49, 2018. Available: 10.1016/j.atmosenv.2018.05.049 [Accessed 20 February 2019.
VIII.T. Fontes, P. Li, N. Barros and P. Zhao, “A proposed methodology for impact assessment of air quality traffic-related measures: The case of PM2.5 in Beijing”, Environmental Pollution, vol. 239, pp. 818-828, 2018. Available: 10.1016/j.envpol.2018.04.061 [Accessed 20 February 2019.
IX.Wang, J., Hu, J., 2015. A robust combination approach for short-term wind speed forecasting and analysis -Combination of the ARIMA (Autoregressive Integrated Moving Average), ELM (Extreme Learning Machine), SVM (Support Vector Machine) and LSSVM (Least Square SVM) forecasts using a GPR (Gaussian ProcessRegression) model. Energy 93, 41–56.
X.World Health Organization, “Monitoring ambient air quality for health impact assessment,” WHO Regional Office Eur., Copenhagen, Denmark, Tech. Rep. 85, 1999.
XI.World Health Organization. Occupational and Environmental Health Team. (2006). WHO Air quality guidelines for particulate matter, ozone, nitrogen dioxide and sulfur dioxide: global update 2005: summary of risk assessment. Geneva: World Health Organization. /10665/69477.
XII.Yuan, X., Tan, Q., Lei, X., Yuan, Y., Wu, X., 2017. Wind power prediction using hybrid autoregressive fractionally integrated moving average and least square support vector machine.
XIII.Y. Cheng, H. Zhang, Z. Liu,L. Chen and P. Wang, “Hybrid algorithm for short-term forecasting of PM2.5 in China”, Atmospheric Environment, vol. 200, pp. 264-279, 2019. Available: 10.1016/j.atmosenv.2018.12.025 [Accessed 20 February 2019]
Krishna Chaitanya Atmakuri, Y Venkata Raghava Rao View Download