Data-driven thermal fault detection of lithium-ion batteries

Abstract

The electric vehicle (EV) industry is beginning to prosper at an unprecedented rate due to laws against the production of internal combustion engine (ICE) vehicles. The laws were implemented as the supply of petrol is running out and due to global warming concerns. Most if not all electric vehicles rely on lithium-ion batteries (LIBs) and rapid technological advancement of LIBs in their power density and energy poses reliability and safety concerns. One of the critical hazards is thermal fault as it can be potentially catastrophic and may cause fire accidents or even explosions.

This report illustrates the algorithm and technique used to implement an anomaly detection for thermal fault. Cell 1 parameters are modelled to simulate thermal fault and the data is used to train a machine learning (ML) model. The model is then used to predict for cell 2 to detect thermal fault. The algorithms were programmed in Python and the LIBs used during experiments were 18650 cells.