EPE Association: EPE 2022 - DS2o: Batteries: Management Systems (BMS), Monitoring and Life-Time Prediction

EPE 2022 - DS2o: Batteries: Management Systems (BMS), Monitoring and Life-Time Prediction
You are here: EPE Documents > 01 - EPE & EPE ECCE Conference Proceedings > EPE 2022 ECCE Europe - Conference > EPE 2022 - Topic 08: Electric Vehicle Propulsion Systems and their Energy Storage > EPE 2022 - DS2o: Batteries: Management Systems (BMS), Monitoring and Life-Time Prediction

   [return to parent folder]

   Battery cycler to generate open li-ion cell aging data and models
By Matthias LUH [View]
[Download]

Abstract: Battery degradation is relevant for the lifetime, cost, and life cycle analysis of electric vehicles and stationary storages. Publicly available, reusable battery aging data is scarce and aging experiments are time-consuming and expensive. This paper starts with an overview of existing battery aging data and models. We then present our battery cycler hardware, which we intend to use to generate open battery degradation data and models. After analyzing the switching behavior, efficiency, and control behavior of the hardware, this paper gives an overview of the generated data and the user interface of the battery cycler. Finally, we provide an outlook on the further development of the project.

   Power Hardware-In-the-Loop test of low-voltage battery for a plug-in hybrid electric vehicle
By Alain BOUSCAYROL [View]
[Download]

Abstract: A methodology developed for a quicker validation of various sub-systems is applied in this paper. It is based on two steps, i.e. the simulation validation and the power Hardware-in-the-Loop (HiL) testing. A new plug-in hybrid electric vehicle with low-voltage battery has been validated on a demo car. A new battery with higher power density is proposed to minimize the energy losses. Before its integration in car, simulation and hardware-in-the-loop tests are achieved. For the HiL part, a real-time simulation of the powertrain is coupled to the battery to test various driving conditions. A dedicated emulation interface is developed. The tests in this paper demonstrate the ability of the battery to operate in a safe condition for the vehicle using a WLTC driving cycle.

   State of Charge Prediction of Lithium-Ion Batteries Based on Artificial Neural Networks and Reduced Data
By Sebastian POHLMANN [View]
[Download]

Abstract: Lithium-ion batteries (LIBs) are the key technology for the electrification of the transport sector. SinceLIBs have a complex, electrochemical structure, it is a challenge to accurately determine the condition,which is crucial for safety and efficiency during operation. This paper presents the forecasting of thestate-of-charge (SOC) of a LIB based on machine learning (ML) algorithms. Data from battery simulation and augmented data are additionally used to train the models. To reduce the dimension of the feature matrix, a singular value decomposition is performed. A multi-layer perceptron (MLP) and a convolutional neural network (CNN) are compared to a linear regression. The impact of the augmented data on the prediction accuracies and the reliabilities of the models is analyzed. The lowest test error is achieved using the CNN with augmented data with a root mean square error (RMSE) of 1.78 \%. The results show the applicability of data-driven models for the SOC prediction and the optimization potential using data augmentation techniques.

EPE 2022 - DS2o: Batteries: Management Systems (BMS), Monitoring and Life-Time Prediction
You are here: EPE Documents > 01 - EPE & EPE ECCE Conference Proceedings > EPE 2022 ECCE Europe - Conference > EPE 2022 - Topic 08: Electric Vehicle Propulsion Systems and their Energy Storage > EPE 2022 - DS2o: Batteries: Management Systems (BMS), Monitoring and Life-Time Prediction
	[return to parent folder]

	Battery cycler to generate open li-ion cell aging data and models By Matthias LUH	[View] [Download]
Abstract: Battery degradation is relevant for the lifetime, cost, and life cycle analysis of electric vehicles and stationary storages. Publicly available, reusable battery aging data is scarce and aging experiments are time-consuming and expensive. This paper starts with an overview of existing battery aging data and models. We then present our battery cycler hardware, which we intend to use to generate open battery degradation data and models. After analyzing the switching behavior, efficiency, and control behavior of the hardware, this paper gives an overview of the generated data and the user interface of the battery cycler. Finally, we provide an outlook on the further development of the project.

	Power Hardware-In-the-Loop test of low-voltage battery for a plug-in hybrid electric vehicle By Alain BOUSCAYROL	[View] [Download]
Abstract: A methodology developed for a quicker validation of various sub-systems is applied in this paper. It is based on two steps, i.e. the simulation validation and the power Hardware-in-the-Loop (HiL) testing. A new plug-in hybrid electric vehicle with low-voltage battery has been validated on a demo car. A new battery with higher power density is proposed to minimize the energy losses. Before its integration in car, simulation and hardware-in-the-loop tests are achieved. For the HiL part, a real-time simulation of the powertrain is coupled to the battery to test various driving conditions. A dedicated emulation interface is developed. The tests in this paper demonstrate the ability of the battery to operate in a safe condition for the vehicle using a WLTC driving cycle.

	State of Charge Prediction of Lithium-Ion Batteries Based on Artificial Neural Networks and Reduced Data By Sebastian POHLMANN	[View] [Download]
Abstract: Lithium-ion batteries (LIBs) are the key technology for the electrification of the transport sector. SinceLIBs have a complex, electrochemical structure, it is a challenge to accurately determine the condition,which is crucial for safety and efficiency during operation. This paper presents the forecasting of thestate-of-charge (SOC) of a LIB based on machine learning (ML) algorithms. Data from battery simulation and augmented data are additionally used to train the models. To reduce the dimension of the feature matrix, a singular value decomposition is performed. A multi-layer perceptron (MLP) and a convolutional neural network (CNN) are compared to a linear regression. The impact of the augmented data on the prediction accuracies and the reliabilities of the models is analyzed. The lowest test error is achieved using the CNN with augmented data with a root mean square error (RMSE) of 1.78 \%. The results show the applicability of data-driven models for the SOC prediction and the optimization potential using data augmentation techniques.