Fault Diagnosis Of Overcharge And Overdischarge Of Lithium-ion Battery

Due to the pressure of environment and energy, the development of electric vehicles has become an inevitable choice for the world automotive industry. However, as the power source of the electric vehicle, the battery is facing the challenge of safety and durability, which have been restricting the electric vehicles’ promotion and industrialization development. Overcharge and overdischarge are typical problems in the use of power battery, which have a great influence on the safety and durability of the battery. Therefore, the rapid and accurate detection of overcharge and overdischarge is of great significance for improving the safety performance of power battery and prolonging its service life.In this paper, the lithium-ion battery that was widely used in electric vehicles was selected as the research object. And a study about fault diagnosis such as overcharge and overdischarge which are easy to appear during operation was carried out. First, starting from the charging and discharging characteristics as well as inconsistency of lithium-ion battery, the causes and consequences of overcharge and overdischarge were discussed. Analysis showed that: lithium-ion battery may undergo thermal runaway so as to cause serious unsafe accidents under overcharge condition; overdischarge would cause irreversible damage and serious capacity attenuation to the battery, resulting in decreased cycle life even premature failure.Next, the current common battery models were analyzed, and a second-order RC network equivalent circuit model of lithium-ion battery was established combining the engineering practical and advantages of the existing battery models. On the basis of that, the state space equation and observation equation were deduced. The relationship between open circuit voltage and state of charge was calibrated through constant current discharge experiment. Subsequently, the dynamic simulation model of the battery was built in Simulink, and the accuracy of the model was verified under the conditions of constant current discharge and pulse discharge. The results show that the model has good applicability and can be used in the simulation of lithium-ion batteries with different parameters.Finally, considering that the parameters variation of lithium-ion battery during the process of charging and discharging, a multi-model estimation and diagnosis method based on kalman filter was proposed, and then the simulation analysis was carried out. Parameters of lithium-ion battery under the condition of overcharge and overdischarge were used, along with the second-order resistance capacitance equivalent circuit methodology, to construct corresponding fault model of the battery. Kalman filters were utilized to estimate the output voltage of each model and to generate residual signals. The filtered residuals were used in the hypothesis test algorithms to generate the probability of each model. The simulation results show that the battery fault can be accurately detected, which indicate the multi-model estimation method is feasible and effective for the fault diagnosis of lithium-ion battery.