Abstract:

In order to diagnose the fault of the input current sensor of the lithium-ion （Li-ion） battery， a linear first-order Thevenin equivalent circuit model of the Li-ion battery， coupled with lumped thermodynamics model， is adopted. Based on this model， a sliding mode observer is designed for fault diagnosis strategy research. First， the nonlinear varying parameters Li-ion battery model is linearized to establish a linear varying parameters Li-ion battery model that satisfies the operating accuracy of the entire operating condition. Then， the linearized Li-ion battery model with varying parameter is recombined and transformed， so that the current in the original system as the input is transformed into the output of the new system. Third， based on the new system of transformation， a sliding mode observer based on temperature estimation error correction is designed to estimate the output current in the new system. Finally， a residual evaluation function based on norm is designed. Simulation results show that the linearized model is feasible； the fault diagnosis strategy designed based on the model can quickly detect the occurrence of input current sensor faults， thus verifying the effectiveness of the proposed fault diagnosis method.

Key words: power engineering, lithium-ion battery, model linearization, current sensor, sliding mode observer, fault diagnosis