Consensus-Based Stochastic Control for Model-Free Cell Balancing

Balancing the state of charge (SOC) of the cells improves both the capacity and the health of the battery. State of the art cell balancing algorithms assume highly accurate SOC estimation which can only be achieved using model-based control solutions. Unfortunately, electrochemical models of batteries are complicated and are still not reliable enough. We propose a simple consensus-based cell balancing algorithm that does not assume any knowledge of the electrochemical model of the battery. Instead, our algorithm only uses SOC estimates from online measurements to make sequential balancing decisions. Using martingale techniques, we prove that our algorithm equalizes the SOC of all cells with probability 1 even with very noisy SOC estimates under a general non-i.i.d. noise model. We then analyze the convergence time of our algorithm as a function of the problem parameters. We also prove performance guarantees for the more challenging case where balancing is done while the battery is charging or discharging. Finally, we present simulations that support our theoretical claims and show the reliability of our algorithm. Our model-free cell balancing is easy to implement and therefore is a major step towards making active cell balancing practical.

Automated summary

This paper introduces a consensus-based algorithm for cell balancing that does not rely on the electrochemical model of the battery, but instead utilizes online SOC estimates for sequential balancing decisions. The algorithm is proven to equalize the SOC of all cells with high probability, even under noisy SOC estimates, and is shown to be reliable and effective through simulations.