Real-time Implementation of Lithium Cell Thermal Diagnostic Algorithms in dSpace
Open Access
- Author:
- Frain, Theodore
- Area of Honors:
- Mechanical Engineering
- Degree:
- Bachelor of Science
- Document Type:
- Thesis
- Thesis Supervisors:
- Satadru Dey, Thesis Supervisor
Margaret Louise Byron, Thesis Honors Advisor - Keywords:
- Lithium-ion Battery
dSPACE
Fault Detection - Abstract:
- For battery management systems, fast recognition and mitigation of cell thermal faults is vital for the preservation of battery health and the prevention of dangerous thermal events such as thermal runaway. Current battery management systems (BMSs) found in industry rely on model-based approaches to predict how the temperature of the cell will change based on the current, voltage, and state of the charge of the cell, all three of which can vary greatly throughout an average drive cycle. This thesis presents a thermal fault detection and mitigation system capable of quickly and effectively detecting and mitigating lithium-ion battery thermal faults. The system consists of a dSPACE MicroLabBox communicating via controller area network with an Arbin cycler, which transmits real-time cell data to the MicroLabBox. The MicroLabBox contains a simple cell thermal model which predicts the expected cell surface temperature based on the current being run through the cell. The predicted temperature is compared to the actual surface temperature of the cell, and an error term is generated. When a thermal fault occurs, the actual surface temperature becomes much larger than the predicted temperature, leading to a spike in the error term. This spike triggers a liquid cooling loop to activate and quickly remove excess heat from the cell. The system was tested by first running it through routine charge and discharge cycles to demonstrate that it does not produce false positives, then injecting an artificial thermal fault via a heat gun, which triggers the cooling system to activate.