Nonlinear model predictive control for cell balancing in Li-ion battery packs

Abstract

A major issue with Li-ion batteries hindering their widespread application in transportation industry is safety related concerns that should be effectively addressed within the battery management system design. Overcharge/discharge of the cells within a battery pack is one undesirable outcome that can affect the battery's health. Differences between the cells within a battery pack can lead to cell over-charge/discharge that is not detectable from battery pack's voltage. Thus, a cell balancing circuit is usually employed in battery packs in order to keep all the cells in balance. The control of cell-balancing circuits is mostly addressed by logical-based algorithms where the dynamical model of the system is not taken into account. This work considers the control problem of a cell-balancing circuit in a model-based framework by proposing a nonlinear model predictive control (NMPC). The NMPC problem is solved using a genetic algorithm and simulation studies show the effectiveness of the proposed algorithm.

Review of Battery Cell Balancing Methodologies for Optimizing Battery Pack Performance in Electric Vehicles

“…It is therefore advised to take cells that have common chemical characteristics, so that same age cells are assembled together in a battery pack. Because of these, the BMS employs cell balancing as a control measure to avoid the overcharge or over-discharge of battery cells [27], and this balancing must take place while the cells are charging or discharging. Cell balancing is done to ensure that the delivered energy to the cells during the charging process is maximized, and the energy released from the cells during the discharging process is also maximized [28].…”

Integrated Systems: Innovations and Applications

“…Danielson et al [68] opens an optimal control view into the problem where however the battery and balancing circuit dynamics are not considered in the problem. Samadi et al [69] has developed a non-linear model predictive control to the cell balancing problem where the balancing problem is converted into an optimization problem. The optimality of the cell-balancing control scheme such that the constraints of the system are met, the life-time of the batteries are maximized and energy is optimized while a complete model of the battery is taken into account and SoC and SoH are considered in the process defines an open field for the research.…”

Robust control of a forward‐converter active battery cell balancing

“…In the decentralized ACB, cell balancing is performed locally and autonomously on each individual cell. Several techniques based on convex optimization [14], model predictive control [15][16][17], adaptive control [18], fuzzy-logic and intelligent techniques [19] have also been proposed for the optimization of energy reallocation, through total energy loss minimization [14], battery range maximization [20] etc. Interested readers are directed to [5] for a comprehensive survey on the topologies and computational algorithms for the energy-reallocation-based ACB.…”