Prediction of spark ignition engine performance responses fueled with fusel oil/gasoline blends by artificial neural network

Authors : Samet USLU, Süleyman ŞİMŞEK

Pages : 100-110

Doi:10.18245/ijaet.807339

View : 16 | Download : 5

Publication Date : 2021-10-14

Article Type : Research Paper

Abstract :In the present study, the performance parameters of a single-cylinder, air-cooled spark ignition insert ignore into journalissuearticles values(SI); engine using fusel oil-gasoline fuel blends were predicted by artificial neural network insert ignore into journalissuearticles values(ANN);. The SI engine was operated with gasoline/fusel oil insert ignore into journalissuearticles values(10% and 20%); blends at different engine load insert ignore into journalissuearticles values(1000, 2000, 3000, 4000, 5000, 6000, 7000 and 8000 Watt); and compression ratios insert ignore into journalissuearticles values(8.00, 9.12 and 10.07); to obtain data essential to create the ANN model. In the constructed ANN model, brake thermal efficiency insert ignore into journalissuearticles values(BTE); and brake specific fuel consumption insert ignore into journalissuearticles values(BSFC); are chosen as output  parameters, while engine load, compression ratio insert ignore into journalissuearticles values(CR); and fusel oil ratio are chosen as input factors. 75% of the test results were employed to train the ANN. The performance of ANN model was determined by comparing it with the data produced from the part not used for training. According to the found data, ANN model estimated engine performance parameters such as BTE and BSFC by an overall regression coefficient insert ignore into journalissuearticles values(R); at 0.99384. Simultaneously, mean absolute percentage error insert ignore into journalissuearticles values(MAPE); were found as 5.027% and 7.847% for BTE and BSFC, respectively. When ANN results and experimental results are compared for BTE and BSFC responses, it is determined that ANN results are close to experimental results with an error rate of less than 5%.
Keywords : Fusel oil, artificial neural network, performance, gasoline engine