Development of ANN model for Prediction of AQI at Sanathnagar, Hyderabad, India.

Abstract

Air pollution is not only environmental problem, but also a major health issues in many countries. The effective methods in controlling the air pollution is by a proper air quality management. One of effective methods is by predicting the air quality index. One such method is based on prediction of data is by using Artificial neural networking (ANN). In this study 21 models were developed for prediction of pollutants. The required data for development of ANN models were collected from the TSPCB (Telangana State Pollution control board), Hyderabad, India. The data comprises 2007-2017 years of hourly data. Best suited models were selected on the basis of statistical tools such as R and MSE. This study was conducted seasonally- Summer, Post-monsoon, Winter. The predicted data is for 2017 – 2023. The AQI (predicted) is Computed by using IND-AQI CPCB method and the quality rating is giving by color coding. After the prediction a comparison was drawn between AQI (observed) and AQI (predicted), then the conclusions were drawn from this table.