Blog
Hybrid deep learning model for efficient state of charge estimation of Li-ion batteries in electric vehicles
Exciting News! Our novel research on Li-ion battery technology in electric vehicles (EVs) has been accepted for publication in the renowned journal of Energy, with an impact factor of 9.0.
Level 2 Journal: Our work has been acknowledged as a Level 2 journal publication according to the Norwegian Scientific Index. This distinction further reinforces the academic value and impact of our study.
Battery Coast Strategy: This paper holds immense relevance to the Battery Coast Strategy at the University of Agder. The Battery Coast serves as a bridge, connecting interdisciplinary battery development and research streams from academia and industry across the entire battery value chain. By fostering collaboration with major battery players in southern Norway and internationally, along with establishing an application-oriented Battery Engineering education, we actively co-create battery competence.
Title: "Hybrid deep learning model for efficient state of charge estimation of Li-ion batteries in electric vehicles"
Authors: Muhammad Hamza Zafar, Majad Mansoor, Mohamad Abou Houran, Noman Mujeeb Khan, Kamran Khan, Syed Kumayl Raza Moosavi, and Filippo Sanfilippo.
Abstract: State of charge (SoC) estimation is critical for the safe and efficient operation of electric vehicles (EVs). This work proposes a hybrid multi-layer deep neural network (HMDNN)-based approach for SoC estimation in EVs. This HMDNN uses Mountain Gazelle Optimizer (MGO) as a training algorithm for the deep neural network. Our method leverages the intrinsic relationship between the SoC and the voltage/current measurements of the EV battery to accurately estimate the SoC in real time. We evaluate our approach on a large dataset of real-world EV charging data and demonstrate its effectiveness in comparison to traditional SoC estimation methods. Four diverse Li-ion battery datasets of electric vehicles are employed which are the dynamic stress test (DST), Beijing dynamic stress test (BJDST), federal urban driving schedule (FUDS), and highway driving schedule (US06) with different temperatures of 0oC,25oC,45oC. The comparison is made with Mayfly Optimization Algorithm based DNN, Particle Swarm Optimization based DNN and Back-Propagation based DNN. The evaluation indices used are normalized mean square error (NMSE), root mean square error (RMSE), mean absolute error (MAE), and relative error (RE). The proposed algorithm achieves 0.1% NMSE and 0.3% RMSE on average on all datasets, which validates the effective performance of the proposed model. The results show that the proposed neural network-based approach can achieve higher accuracy and faster convergence than existing methods. This can enable more efficient EV operation and improved battery life.
Article Link: Hybrid deep learning model for efficient state of charge estimation of Li-ion batteries in electric vehicles
Possible collaborations: Morrow Batteries
AI- and Robotics-enabled Systems, a Forward Leap Into Real Life Applications
I had the privilege of being the International Guest Speaker at the School of Computing Technologies on Tuesday, 13 June 2023. It was an honour to present a seminar as a fellow OpenInnoTrain Secondee.
I would like to express my gratitude to Professor Fabio Zambetta, the Associate Dean of Artificial Intelligence in the School of Computing Technologies, for the warm welcome and introduction. I also would like to thank the entire RMIT team that welcomed me. The event provided a fantastic platform to discuss cutting-edge research and innovative ideas in the field of engineering and science.
International Guest Speaker: Professor Filippo Sanfilippo, Faculty of Engineering and Science, University of Agder, Norway | OpenInnoTrain Secondee
Welcome and introduction: Professor Fabio Zambetta, Associate Dean, Artificial Intelligence in the School of Computing Technologies, RMIT University, Melbourne
Reflection:
In this seminar, Professor Filippo Sanfilippo shared some of his work in Industry 4.0 and 5.0, specifically concerning human-robot collaboration and teaming. Filippo says the fifth revolution of industrial automation is occurring with the convergence of artificial intelligence and the presence of sensors everywhere - on and around the machine and on humans through wearables, providing the possibility for machines and humans to share workspaces and to work hand in hand.
Obstacles to this revolution include that robotics producers currently utilise proprietary software with limitations in usability however, Filippo says that opensource middleware used as a bridge between a proprietary server and a user program provides a way forward. The concept of 'digital twins' where real labs and digital labs are connected is an another trend that will enable the evolution of Industry 5.0.
During the seminar Filippo spoke about the challenges for human-robot collaborations including the trade-off between control and mechanical/software design. He also shared case studies of human-robot interaction and collaboration with regards to applications for intelligent health, along with the work his team has progressed around potential applications for search-and-research collaborations.
Following the seminar, Professor Zambetta shared his thoughts on the seminar and Professor Sanfilippo’s OpenInnoTrain secondment to RMIT Melbourne:
We are very happy to be hosting Professor Sanfilippo at RMIT and we would like to thank OpenInnoTrain for providing us with a new collaboration opportunity.
Fililppo's expertise in robotics, mechatronics and control is complementary to the Ai Discipline expertise in machine learning (reinforcement learning), automated planning, evolutionary computing and AR/VR simulation.
More opportunities for collaboration in the School also include potential opportunities with staff whose expertise lies in data science as well as IoT and smart sensing.
Seminar participant, Jenny Hedley, also shared:
Seeing how agile and responsive the robot snake is compared with robot hands/grabbers is an eye opener! As an avid hiker I imagine how useful this would be in wilderness rescue situations when someone triggers an emergency transponder.
A recording of the seminar is available to view here.
Acknowledgments:
OpenInnoTrain Project, is a global network of researchers and industry practitioners across Europe and Australia for promoting the translation of research between university-industry through cooperation and Open Innovation in the sectors of: FinTech, Industry 4.0, CleanTech, FoodTech. This project has received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 823971.
RMIT University, Melbourne, Australia - Global Business Innovation Enabling Impact Platform (Director, Professor Anne-Laure Mention).
University of Agder
 Algorithm for Engineering Design Problems-323_L.png "Improved Barnacles Movement Optimizer (IBMO) Algorithm for Engineering Design Problems")
Improved Barnacles Movement Optimizer (IBMO) Algorithm for Engineering Design Problems
Exciting News!
I am thrilled to announce the publication of our paper titled "Improved Barnacles Movement Optimizer (IBMO) Algorithm for Engineering Design Problems" in the Proceedings of the 22nd International Conference on Artificial Intelligence and Soft Computing (ICAISC 2023) held in Zakopane, Poland.
I am grateful to all my co-authors for their incredible contributions! Syed Kumayl Raza Moosavi presented the paper at the conference and received very good and constructive feedback!
- Syed Kumayl Raza Moosavi, Muhammad Hamza Zafar, Seyedali Mirjalili and Filippo Sanfilippo. Improved Barnacles Movement Optimizer (IBMO) Algorithm for Engineering Design Problems. In Proceeding of 22nd International Conference on Artificial Intelligence and Soft Computing (ICAISC 2023), Zakopane, Poland. 2023, –. PDF
Filippo Sanfilippo