Observer-Based Diagnostic Scheme for Lithium-Ion Batteries

This paper presents an observer-based fault diagnosis approach for Lithium-ion batteries. This method detects and isolates five fault types, which include sensor faults in current, voltage and temperature sensors, failure in fan actuator and a fault in battery State-of-Charge (SOC) dynamics. Current, voltage and temperature of the battery are taken as the only available measurements and a Kalman filter and a sliding mode observer are constructed. Three residuals derived from a combination of these observers generate fault signatures that are used to detect and isolate the sensor, actuator and SOC faults in the system. Simulation results show the effectiveness of the approach.