Prediction of Respirable Particulate Matter (PM10) Concentration using Artificial Neural Network in Kota city

Recent years concerns related with ambient air quality is prominent due to increments in the entropy and ozone layer depletion. Green house gas emission from the industries is the key contributing factor in the increase in carbon foot prints. The accurate prediction of hazardous gases in the environment can be beneficial information to initiate the corrective strategies for reduction in carbon foot prints. This paper presents a supervised learning based prediction engine for prediction of Respirable Suspended Particulate Matter (RSPM). “Supervised learning” takes a known set of input data and known responses to the data, and seeks to build a predictor model that generates reasonable predictions for the response of new data. Data of 2012,2013 and 2014 of an industrial area of Kota city is employed to train, test and validate four different topologies of the neural networks namely Feed Forward Neural Network (FFNN), Layer Recurrent Neural Network (LRNN),Nonlinear autoregressive Exogenous (NARX) and Radial Basis Function Neural Network (RBFN). A meaningful comparison between these topologies revealed that RBFN is a suitable topology for prediction engine. PM10 constitutes solid and liquid suspended particles having an aerodynamic diameter up to 10 µm (micro meter). It is a common pollutant among all sectors namely transports domestic, industries and manufacturing so it is the major common pollutant need to be taken under consideration in terms of air pollution. Most of the cities of India exceed PM10 levels according to the National ambient air quality standard (NAAQS). PM10 is responsible for heart and lungs diseases. To decrease the mortality rate due to these pollutant Effective countermeasures need to be taken. The precise prediction of pollutant needed to alert the population.