Effect of plastic oil addition on performance and emission characteristics of biogas-diesel dual fuel engine using taguchi method and prediction of performance parameter using artificial neural network

Abstract

Ever increasing usage of fossil fuels and dwindling natural resources led researchers to concentrate on investigating other sources which can satisfy our demands and reduce pollution levels. Present research work aims to investigate the performance and emission characteristics of plastic, diesel and biogas as fuel blend operated in a dual-fuel engine with biogas as a primary fuel and plastic oil – diesel blends as secondary fuel and also predict the output variable using artificial neural network. A modified four-stroke single cylinder CI engine was used for experiments conducted at varying load, percentage of plastic oil percentage in diesel and biogas flow rate. Based on the levels and factors a Taguchi L9 orthogonal matrix was designed to find the optimal combination of input indices. The signal to noise ratios in taguchi method were applied based on the desired output characteristics and according to the respective SNR ratios an ANOVA table was created to determine the major contributor effecting output parameters such as brake thermal efficiency, CO, HC NOx and smoke emissions. ANN model helped to predict BTE with same input parameters but with an increased set of sample data. Based on Gradient descent and Levenberg-Marquardt algorithm the ANN architecture was trained, validated and tested to predict the response with least error. The ANOVA calculated indicates load to be the prime factor effecting BTE and NOx emission

Design of a prediction model based on improved BP neural network and particle swarm optimization for more accurate budget of biogas production

“…In [1], the mathematical model and multiple regression equation were used to predict the biogas yield of anaerobic fermentation of mixture. In [2], the introduction of kinetic (SD) theory, the establishment of kinetic model to predict gas production; [3][4][5] modeled by artificial neural network to predict biogas production. These models are different.…”