Mobile system shutdown prevention via energy storage-aware predictive control

This paper presents an energy storage-aware model predictive control approach for online mobile system shutdown prevention that exploits the rate-capacity and recovery dynamic effects of batteries. System shutdown, for these systems such as ground robotics, commonly occurs as a result of transient loads that result in the battery voltage crossing a shutdown voltage threshold in the protective circuitry. The proposed control methodology optimizes the vehicle drive command by incorporating battery shutdown constraints and battery dynamic effects into a model predictive control quadratic program optimization. The proposed model predictive control scheme is shown to extend mobile system run-time and total distance through Monte Carlo simulation and experimental studies of a small unmanned ground vehicle.