| Popis: |
Lithium ion batteries degrades with time. Battery diagnostics links observable changes in cell parameters to changes that have occurred inside the cell. More accurate diagnosis means a better insight into these internal changes which can aid in predicting failure modes, remaining life, aide in optimising future use. In the case of electric vehicles, an on-board battery management system (BMS) could implement these diagnostic techniques to give real time information on the internal state of the pack. It can then dynamically adjust parameters such as state of charge (SOC) window, maximum power and maximum allowable temperature to optimise performance and pack longevity. Models capable of predicting Li-ion battery health with higher accuracy are highly advanced, and computationally exhaustive requiring complex physics-based models or equivalent circuit models (ECM) that require exhaustive parameter fitting. Other techniques require measurements or hardware which are not available in typical applications and so are only suited to lab-based diagnostics. Current diagnostic techniques are not suitable for in-situ applications; either the BMS does not have the computational capacity to run the models, or the measurements required are not available outside a laboratory environment. In this work, we propose an Open Circuit Voltage (OCV) based model that predicts the degradation of a cell accurately without needing extensive computational capacity. This model has been embedded in a microcontroller with limited computing power which could form part of the BMS system that better predicts the state of health of the cell. The model uses the open circuit potentials of each electrode and fits them to the full cell voltage in order to identify the degradation modes. Typical modes include Loss of Lithium (LLI) and Loss of Active Material from either electrode (LAM). The model does not require the solving of concurrent complex equations or the fitting of extensive parameters to an equivalent circuit whilst still being able to provide accurate insight into the cell’s internal changes. Results from this work demonstrate the usefulness of such a system in electric vehicle and paves the way for implementation of smarter BMS in electric vehicles by manufacturers. This unlocks the potential to predict and increase the longevity of a battery pack whilst still providing optimal performance to the vehicle. |
