A safety-reinforced mutual pulse heating strategy based on microscopic-state estimation for power-redistributable lithium-ion battery pack

Advanced heating is essential for mitigating temporary losses in battery capacity and peak power at low temperatures, thereby enhancing utilization efficiency and reducing safety risks. Expanding upon our prior research, this paper proposes a safety-reinforced mutual pulse heating (MPH) strategy based on microscopic-state estimation, in which multiplexing converter-based drivers are reused as on-board heating topologies. An adaptive super-twisting sliding mode observer is utilized to capture battery solid-liquid phase Li+ concentration, electrode and electrolyte potentials, and side-reaction overpotentials. Then, a microscopic-state constraints-based MPH current prediction method is formulated to delineate the safe operational area of battery micro-states. The MPH current amplitude is updated by an online negative feedback-based optimization controller, while the heating duty ratio is still adjusted using fuzzy feedback mechanisms. Experimental validation of the proposed method is conducted across five critical metrics: heating duration, temperature-rise rate, safety risk, temperature uniformity, and degradation. The method demonstrated its efficacy by elevating the battery temperature from −20 °C to 0 °C within 178.3 s, while keeping the microscopic states within safe thresholds, with a peak temperature gradient of <2 °C. During 800 heating cycles, battery integrity remains intact as evidenced by stable capacity, coulomb efficiency, internal resistance, and differential voltage curves.