Recent research progress has witnessed the emergence of intelligent batteries integrating cell-level sensors and controllers for distributed and enhanced management. However, the highly integrated structure rises up multifold concerns over the rigorous thermal condition, and the trade-off between sensing quality and cost. Motivated by this, this paper pursuits the possibility of designing an intelligent battery with the function of state of charge (SOC) self-monitoring without using the current sensor. The essence is to propose a method converting the co-estimation of SOC and input current into a state-space model-based optimization problem, which is virtually solved in a moving horizon framework. Results suggest that the proposed method can realize accurate estimation of the SOC even if freeing the need of integrating a current sensor. The encouraging results are insightful for reducing the structural complexity and lowering down the cost of intelligence batteries.
Keywords intelligent battery, state estimation, moving horizon estimation, current sensor