Battery State of Charge Estimation With Event-Triggered Parameter Identification

The safe and efficient battery system operation depends on accurate state of charge (SOC) estimation. Parameter identification is vital to SOC estimation. To strike the balance between the computational burden and adaptiveness, this article proposes a state estimation scheme with event-triggered parameter identification for both continuous and sampling-based nonlinear systems. It means that the parameters are identified only when an event is triggered, the parameter changes can be well-tracked, and less computational burden is needed. Then, the system stability can be ensured and the unexpected Zeno behavior can be excluded by the proper design of event-triggered condition. Finally, we apply the theoretical analysis to two different kinds of SOC estimators with parameter identification. The experiment and simulation results demonstrate that the designed event-triggered framework is effective in achieving the accuracy estimation and reducing computational burden.