Research And Application Of Self-discharge Detection For Li-ion Battery

In dealing with the increasingly urgent energy shortage around the world,the rechargeable battery on behalf of the newly emerging resources is beingadvanced recent years. Li-ion battery is widely applied among various devicesand electric vehicles with superior characteristics in environment friendly,specific energy and cycle life as power source. However, the self-dischargephenomenon existed within the battery has induced energy loss and the generallife reduction of battery pack composed by those with inconsistency self-discharge as well, which may also induce rapid reduction in battery capacity. Theinconsistency of self-discharge can induce significant error in predicting the stateof charge (SOC) for battery management system (BMS) and invalidation inelectric vehicle control strategy, which may induce over-discharging for battery.Therefore, it is of significance to observe the self-discharge parameters rapidly inactual application.Self-discharge is defined as a phenomenon existed within the battery, wheredirect measurement cannot be conducted, which adding to difficulties forobservation of self-discharge in advance. Common measurement is roughlyconducted by putting the object battery aside for a standard long time andobserving the capacity loss of the battery, which greatly consumes time andperiod and rapidity and real-time performance. In dealing with the problem above,this paper has proposed a precise digital self-discharge detection system withindirect measurement in order to improve rapidity and real-time performance.The system is composed with MSP430F149as micro-controller and DAC1220D/A convertor for high-resolution voltage, which is applied as the referencevoltage for detection circuit. Corresponding comparison between traditionalbattery put-aside experiment and newly proposed detection system is conductedand the results have proved the validity of the system in reliability, precision and rapidity for self-discharge detection.The effects of self-discharge for SOC prediction are also analyzed with thedescription above in advance in terms of the amendment of SOC initial valuewith self-discharge current based on extended-Kalman filter and the establishedMap model of self-discharge current. Corresponding working conditionsimulation experiment is included to prove the validity of the amendment in theprediction precision of SOC, which improve the general precision of BMS inadvance.