Comparative study of artificial neural network and mathematical model on marine diesel engine performance prediction
The investigation of marine diesel engines is still limited and considered new in both: physical testing and prediction. Therefore, this study deals with an artificial neural network (ANN) modeling for a marine diesel engine performance prediction such as the brake power (BP), brake specific fuel co...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
ICIC International
2018
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/21172/1/Publication%204_IJICIC.pdf |
| _version_ | 1825812082710282240 |
|---|---|
| author | C. W., Mohd Noor R., Mamat Ahmed, Ali Najah |
| author_facet | C. W., Mohd Noor R., Mamat Ahmed, Ali Najah |
| author_sort | C. W., Mohd Noor |
| collection | UMP |
| description | The investigation of marine diesel engines is still limited and considered new in both: physical testing and prediction. Therefore, this study deals with an artificial neural network (ANN) modeling for a marine diesel engine performance prediction such as the brake power (BP), brake specific fuel consumption (BSFC), brake thermal efficiency (BTE), volumetric efficiency (VE), exhaust gas temperature (EGT) and nitrogen oxide (NOX) emissions. Input data for network training was gathered from laboratory engine testing operated at various speed, load and fuel blends. ANN prediction model
was developed based on standard back-propagation with Levenberg-Marquardt training algorithm. The performance of the model was validated by comparing the prediction data sets with the experimental data and the output from the mathematical model. Results showed that the ANN model provided a good agreement to the experimental data with the coefficient of determinations (R2) of 0.99. Mean absolute prediction error (MAPE) of ANN and the mathematical model is between 1.57-9.32% and 4.06-28.35% respectively.
These values indicate that the developed ANN model is more reliable and accurate than the mathematical model. The present study reveals that the ANN approach can be used to predict the performance of marine diesel engine with high accuracy. |
| first_indexed | 2024-03-06T12:23:52Z |
| format | Article |
| id | UMPir21172 |
| institution | Universiti Malaysia Pahang |
| language | English |
| last_indexed | 2024-03-06T12:23:52Z |
| publishDate | 2018 |
| publisher | ICIC International |
| record_format | dspace |
| spelling | UMPir211722018-05-21T02:33:17Z http://umpir.ump.edu.my/id/eprint/21172/ Comparative study of artificial neural network and mathematical model on marine diesel engine performance prediction C. W., Mohd Noor R., Mamat Ahmed, Ali Najah TJ Mechanical engineering and machinery The investigation of marine diesel engines is still limited and considered new in both: physical testing and prediction. Therefore, this study deals with an artificial neural network (ANN) modeling for a marine diesel engine performance prediction such as the brake power (BP), brake specific fuel consumption (BSFC), brake thermal efficiency (BTE), volumetric efficiency (VE), exhaust gas temperature (EGT) and nitrogen oxide (NOX) emissions. Input data for network training was gathered from laboratory engine testing operated at various speed, load and fuel blends. ANN prediction model was developed based on standard back-propagation with Levenberg-Marquardt training algorithm. The performance of the model was validated by comparing the prediction data sets with the experimental data and the output from the mathematical model. Results showed that the ANN model provided a good agreement to the experimental data with the coefficient of determinations (R2) of 0.99. Mean absolute prediction error (MAPE) of ANN and the mathematical model is between 1.57-9.32% and 4.06-28.35% respectively. These values indicate that the developed ANN model is more reliable and accurate than the mathematical model. The present study reveals that the ANN approach can be used to predict the performance of marine diesel engine with high accuracy. ICIC International 2018 Article PeerReviewed application/pdf en http://umpir.ump.edu.my/id/eprint/21172/1/Publication%204_IJICIC.pdf C. W., Mohd Noor and R., Mamat and Ahmed, Ali Najah (2018) Comparative study of artificial neural network and mathematical model on marine diesel engine performance prediction. International Journal of Innovative Computing, Information and Control, 14 (3). pp. 959-969. ISSN 1349-4198. (Published) http://www.ijicic.org/ijicic-140313.pdf DOI: 10.24507/ijicic.14.03.959 |
| spellingShingle | TJ Mechanical engineering and machinery C. W., Mohd Noor R., Mamat Ahmed, Ali Najah Comparative study of artificial neural network and mathematical model on marine diesel engine performance prediction |
| title | Comparative study of artificial neural network and mathematical model on marine diesel engine performance prediction |
| title_full | Comparative study of artificial neural network and mathematical model on marine diesel engine performance prediction |
| title_fullStr | Comparative study of artificial neural network and mathematical model on marine diesel engine performance prediction |
| title_full_unstemmed | Comparative study of artificial neural network and mathematical model on marine diesel engine performance prediction |
| title_short | Comparative study of artificial neural network and mathematical model on marine diesel engine performance prediction |
| title_sort | comparative study of artificial neural network and mathematical model on marine diesel engine performance prediction |
| topic | TJ Mechanical engineering and machinery |
| url | http://umpir.ump.edu.my/id/eprint/21172/1/Publication%204_IJICIC.pdf |
| work_keys_str_mv | AT cwmohdnoor comparativestudyofartificialneuralnetworkandmathematicalmodelonmarinedieselengineperformanceprediction AT rmamat comparativestudyofartificialneuralnetworkandmathematicalmodelonmarinedieselengineperformanceprediction AT ahmedalinajah comparativestudyofartificialneuralnetworkandmathematicalmodelonmarinedieselengineperformanceprediction |