Machine Learning-Based Control for Fuel Cell Hybrid Buses: From Average Load Power Prediction to Energy Management
In this work, a machine learning-based energy management system is developed using a long short-term memory (LSTM) network for fuel cell hybrid buses. The neural network implicitly learns the complex relationship between various factors and the optimal power control from massive data. The selection...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-12-01
|
| Series: | Vehicles |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2624-8921/4/4/72 |
| _version_ | 1797454941691838464 |
|---|---|
| author | Hujun Peng Jianxiang Li Kai Deng Kay Hameyer |
| author_facet | Hujun Peng Jianxiang Li Kai Deng Kay Hameyer |
| author_sort | Hujun Peng |
| collection | DOAJ |
| description | In this work, a machine learning-based energy management system is developed using a long short-term memory (LSTM) network for fuel cell hybrid buses. The neural network implicitly learns the complex relationship between various factors and the optimal power control from massive data. The selection of the neural network inputs is inspired by the adaptive Pontryagin’s minimum principle (APMP) strategy. Since an estimated value of the global average fuel cell power is required in the machine learning-based energy management strategy (EMS), some global features of driving cycles are extracted and then applied in a feedforward neural network to predict the average fuel cell power appropriately. The effectiveness of the machine learning-based energy management, with the integration of the mechanism of estimating the average fuel cell power based on the forward neural network, is tested under two different driving cycles from the training environment, with comparisons to a commercially used rule-based strategy. Based on the simulation results, the learning-based strategy outperforms the rule-based strategy regarding the charge-sustaining mode conditions and fuel economy. Moreover, compared to the best offline hydrogen consumption, the machine learning-based strategy consumed 0.58% and 0.36% more than the best offline results for both driving cycles. In contrast, the rule-based strategy consumed 1.80% and 0.96% more than optimal offline results for the two driving cycles, respectively. Finally, simulations under battery and fuel cell aging conditions show that the fuel economy of the machine learning-based strategy experiences no performance degradation under components aging compared to offline strategies. |
| first_indexed | 2024-03-09T15:45:21Z |
| format | Article |
| id | doaj.art-06ec5d2f481149bcb8f31f4ef24d53f4 |
| institution | Directory Open Access Journal |
| issn | 2624-8921 |
| language | English |
| last_indexed | 2024-03-09T15:45:21Z |
| publishDate | 2022-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Vehicles |
| spelling | doaj.art-06ec5d2f481149bcb8f31f4ef24d53f42023-11-24T18:34:41ZengMDPI AGVehicles2624-89212022-12-01441365139010.3390/vehicles4040072Machine Learning-Based Control for Fuel Cell Hybrid Buses: From Average Load Power Prediction to Energy ManagementHujun Peng0Jianxiang Li1Kai Deng2Kay Hameyer3Institute of Electrical Machines (IEM), RWTH Aachen University, 52062 Aachen, GermanyInstitute of Electrical Machines (IEM), RWTH Aachen University, 52062 Aachen, GermanyInstitute of Electrical Machines (IEM), RWTH Aachen University, 52062 Aachen, GermanyInstitute of Electrical Machines (IEM), RWTH Aachen University, 52062 Aachen, GermanyIn this work, a machine learning-based energy management system is developed using a long short-term memory (LSTM) network for fuel cell hybrid buses. The neural network implicitly learns the complex relationship between various factors and the optimal power control from massive data. The selection of the neural network inputs is inspired by the adaptive Pontryagin’s minimum principle (APMP) strategy. Since an estimated value of the global average fuel cell power is required in the machine learning-based energy management strategy (EMS), some global features of driving cycles are extracted and then applied in a feedforward neural network to predict the average fuel cell power appropriately. The effectiveness of the machine learning-based energy management, with the integration of the mechanism of estimating the average fuel cell power based on the forward neural network, is tested under two different driving cycles from the training environment, with comparisons to a commercially used rule-based strategy. Based on the simulation results, the learning-based strategy outperforms the rule-based strategy regarding the charge-sustaining mode conditions and fuel economy. Moreover, compared to the best offline hydrogen consumption, the machine learning-based strategy consumed 0.58% and 0.36% more than the best offline results for both driving cycles. In contrast, the rule-based strategy consumed 1.80% and 0.96% more than optimal offline results for the two driving cycles, respectively. Finally, simulations under battery and fuel cell aging conditions show that the fuel economy of the machine learning-based strategy experiences no performance degradation under components aging compared to offline strategies.https://www.mdpi.com/2624-8921/4/4/72fuel cell hybrid vehiclesmachine learningfeedforward neural networkLSTM networkenergy managementcomponent aging |
| spellingShingle | Hujun Peng Jianxiang Li Kai Deng Kay Hameyer Machine Learning-Based Control for Fuel Cell Hybrid Buses: From Average Load Power Prediction to Energy Management Vehicles fuel cell hybrid vehicles machine learning feedforward neural network LSTM network energy management component aging |
| title | Machine Learning-Based Control for Fuel Cell Hybrid Buses: From Average Load Power Prediction to Energy Management |
| title_full | Machine Learning-Based Control for Fuel Cell Hybrid Buses: From Average Load Power Prediction to Energy Management |
| title_fullStr | Machine Learning-Based Control for Fuel Cell Hybrid Buses: From Average Load Power Prediction to Energy Management |
| title_full_unstemmed | Machine Learning-Based Control for Fuel Cell Hybrid Buses: From Average Load Power Prediction to Energy Management |
| title_short | Machine Learning-Based Control for Fuel Cell Hybrid Buses: From Average Load Power Prediction to Energy Management |
| title_sort | machine learning based control for fuel cell hybrid buses from average load power prediction to energy management |
| topic | fuel cell hybrid vehicles machine learning feedforward neural network LSTM network energy management component aging |
| url | https://www.mdpi.com/2624-8921/4/4/72 |
| work_keys_str_mv | AT hujunpeng machinelearningbasedcontrolforfuelcellhybridbusesfromaverageloadpowerpredictiontoenergymanagement AT jianxiangli machinelearningbasedcontrolforfuelcellhybridbusesfromaverageloadpowerpredictiontoenergymanagement AT kaideng machinelearningbasedcontrolforfuelcellhybridbusesfromaverageloadpowerpredictiontoenergymanagement AT kayhameyer machinelearningbasedcontrolforfuelcellhybridbusesfromaverageloadpowerpredictiontoenergymanagement |