A control system is described to improve all dynamic, multi-cell metal air batteries to ensure load requirements are met while optimizing battery performance according to a range of performance criteria. This control system can be augmented with Machine Learning to further improve both the effectiveness and efficiency of the battery system over time. A dynamic multi-cell metal air battery system design is disclosed to achieve continuous or intermittent high power, broadening the applicability of metal air batteries combined with electric motors to applications traditionally reserved for internal combustion engines.