Robust and Computationally Efficient Estimation of State of Power and Maximum Energy for Lithium-Ion Batteries in Electric Vehicles

Battery management system (BMS) affects the safety and performance of electric vehicles by providing adequate power ranges (e.g., state of power or power profile of maximum energy) to the vehicle dynamics control. The main task of the BMS is to ensure battery longevity while maximizing the power ranges the battery can give or receive. This paper proposes a procedure for estimating the state of power (SoP) and maximum energy (ME) for a lithium-ion battery cell based on sequential linear programming and using the nonlinear dual polarization equivalent circuit model (ECM) with average power input. The estimation procedure is based on the optimization problem with linear optimization criterion and nonlinear constraints on a fixed prediction horizon. The ECM with power input enables timewise efficient on-line SoP and ME estimation for battery charging and discharging. The SoP and ME estimation is verified through a case study by using the derived ECM with power input.