Battery Cell Identification and SOC Estimation Using String Terminal Voltage Measurements

A battery system consists of many battery cells or modules that may have different characteristics. To achieve reliable, efficient, and extended utilization of battery systems, the battery management system (BMS) must keep track of individual cell-level dynamics, state of charge (SOC), state of health (SOH), failure status, and life-expectancy prediction. Current battery technology employs cell-or module-level voltage sensors, with high costs for sensors and packaging, and substantial reliability issues. This paper introduces new methods that utilize existing cell-balancing circuits to estimate an individual cell's voltage and current from battery string terminal voltage/current measurements. This is achieved by actively controlling balancing circuits to create partial observability for battery cell subsystems. Control strategies, estimation algorithms, and their key properties are developed. Some typical battery model structures are used to illustrate the usage of the methods.