Fortschritte im Automobilbau
Offener Zugang

Abstrakt

Fault diagnosis of lithium-ion batteries based on electrochemical models

Md. Ashiqur Rahman, Sohel Anwar and Afshin Izadian

In this work, a Multiple Model Adaptive Estimation (MMAE) based approach for fault diagnosis of lithium-ion batteries is illustrated. The electrochemical modeling approach is integrated into MMAE for fault diagnosis. This model of a lithium-ion battery (with Li-Co-O2 cathode chemistry) based on real physical laws with nominal model parameters is considered as a model of a healthy battery. Battery fault conditions such as aging, overcharging and deep discharging lead to significant deviations of the parameters from the nominal values ​​and can be considered as separate models. Output fault injection based partial algebraic differential equation (PDAE) observers are used to generate the residual voltage signals. These residuals are then used in the MMAE algorithm to detect persistent battery fault conditions. Simulation results show that the fault conditions can be accurately detected and identified, which supports the effectiveness of the proposed battery fault detection method.