Premium
Developing a smart fuel using artificial neural network for compression ignition engine fueled with Calophyllum inophyllum diesel blend at various compression ratio
Author(s) -
Venugopal Paramaguru,
Kasimani Ramesh,
Chinnasamy Suresh
Publication year - 2020
Publication title -
environmental progress and sustainable energy
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.495
H-Index - 66
eISSN - 1944-7450
pISSN - 1944-7442
DOI - 10.1002/ep.13356
Subject(s) - compression ratio , artificial neural network , ignition system , compression (physics) , cylinder , diesel fuel , automotive engineering , computer science , diesel engine , engineering , artificial intelligence , materials science , mechanical engineering , internal combustion engine , aerospace engineering , composite material
In machinery, it is evident that the computing system for self‐automated machinery derives nonlinear and complex equations by comparing the machinery's different input parameters with their corresponding performance output parameters. In order to operate the machinery with good performance and better efficiency, the computing system needs a machine‐learning algorithm. Most recent researchers have concentrated more on self‐driving vehicle, which seems to be lack of developing a strong algorithm for compression ignition (CI) engines to predict the performance and emission output parameter. Thus, this article deals with the prediction of performance and emission characteristics of CI engine fueled with 25% Calophyllum inophyllum and 75% diesel blend (CIB25) at various compression ratios using artificial neural network (ANN). Performance and emission tests were conducted in a single‐cylinder four‐stroke variable‐compression‐ratio CI engine fueled with CIB25 with varying loads and at a constant speed of operation. Experimental investigation indicates that 18:1 compression ratio gives better performance results when CIB25 is used as the fuel. Emission test results show better emission characteristics at 17:1 compression ratio. These results show that some input factors affect the output factors under some set of operating conditions, while some input factors improve them. ANN developed for the CI engine learns how the input factors affect and improve the output factors. Also, developed neural network is found to be satisfactory, and it predicts the output at a regression value of 0.998 with an average error of 1.77% in the case of CIB25.