Prediction of municipal solid waste generation for optimum planning and management with artificial neural network—case study: Faridabad City in Haryana State (India)

Abstract

Accurate prediction of municipal solid waste generation has an important role in future planning and waste management system. The characteristics of the generated solid waste are different at different places (municipality to municipality or country to country). The accurate prediction of municipal solid waste (MSW) generation becomes a crucial task in modern era. Its prediction requires accurate MSW data. The aim of the present study is to design the time series model for predicting monthly based municipal solid waste generation in Faridabad city of Haryana State (India) using artificial neural network (ANN) time series autoregressive approach. The collected municipal solid waste observations have been arranged monthly from 2010 to 2014. The 60 months data set is divided into 42 training data sets, 9 testing data sets and 9 validating data sets. Various structures of ANN have been investigated by changing the number of hidden layer neurons. Finally best optimized structure of neural network is found. The proposed model is validated by the minimum value of performance parameters such as mean square error 0.0003714, root mean square error 0.01927 and the high value of the coefficient of regression 0.8385. On the bases of these performance parameters it is concluded that the proposed ANN model gives accurate predictive results.