Multilayer perceptron artificial neural network for the prediction of heating value of municipal solid waste

Abstract

Energy from municipal solid waste is steadily being integrated into the global energy feedstock, given the huge amount of waste being generated from various sources. This study develops a Multilayer Perceptron Artificial Neural Network for the prediction of High Heating Value of municipal solid waste as a function of moisture content, carbon, hydrogen, oxygen, nitrogen, sulphur, and ash. A total of 123 experimental data were extracted from reliable database for training, testing, and validation of the model. This model was trained, validated and tested with 70%, 20%, and 10% of the municipal solid waste biomass datasets respectively. The predicted High Heating Value was compared with the experimental data for two different training functions: Levenberg Marquardt backpropagation and Resilience backpropagation, and with some correlation from the literature. The accuracy of the model was reported based on some known performance criteria. The values of Root Mean Squared Error (RMSE), Mean Absolute Deviation (MAD), Mean Absolute Percentage Error (MAPE), and Coefficient of Correlation (CC) were 3.587, 2.409, 21.680, 0.970 respectively for RP and 3.095, 0.328, 22.483, 0.986 for LM respectively. Regression analysis was also carried out to determine the level of correlation between the experimental and predicted High Heating Values (HHV). The authors concluded that these models can be a useful tool in the prediction of heating value of MSW in order to facilitate clean energy production from waste.

Keywords

• municipal solid waste
• high heating value
• multilayer perceptron
• clean energy
• levenberg marquardt backpropagation
• resilience backpropagation