Optimal Load Sharing in Reconfigurable Battery Systems using an Improved Model Predictive Control Method

Optimal load sharing through balancing has the potential of increasing the lifetime and capacity of the lithiumion battery. This paper presents a model predictive control (MPC) approach for optimal load sharing for lithium-ion multi-battery systems in electric vehicles. The primary objective is the satisfaction of the motoring or regenerative load power demand. Secondary objectives are state of charge (SoC) and temperature balancing amongst battery units. An advanced electro-thermal model which includes a two-state thermal model is used, which leads to better performance in terms of temperature balancing and control. The proposed MPC controller, which reduces battery degradation, is validated through simulations under the urban dynamometer driving schedule. The results showed satisfactory power tracking, and SoC and temperature balancing.