ANN-Based Prediction of PM_2.5 for Delhi

Air pollution is one of the prime factors responsible for poor health of mankind. With the advent of technology, industrialization and urbanization, there has been an increase in the pollutants which are hazardous to the mankind. As per the WHO statistics 2016, Delhi, the capital of India is fifth most polluted city. In the present work, an attempt has been made to develop an artificial intelligence-based prediction model. Meteorological parameters such as temperature, vertical wind speed, wind direction, solar radiation, relative humidity and wind speed have been incorporated as input parameters in order to predict PM_2.5 concentration present in the air. The authors have been successful to develop an algorithm which is able to forecast PM_2.5 up to one day advance. The efficacy of the algorithm is determined by coefficient of correlation, mean square error and root mean square error, respectively. The results obtained are very promising after an extensive simulation of the neural network for both 70/15/15 and 80/10/10, respectively. The maximum R value obtained is 0.923.