Optimal kinetic energy recovery algorithms for electric machines

Abstract

Electric machines are used to accelerate and decelerate mechanical loads. During deceleration events, a significant portion of stored kinetic energy can be converted into electrical form for storage in a battery or capacitor, through the use of regenerative braking. By reducing the net energy flow out of the electric power source, regenerative braking is one of the mechanisms by which the overall efficiency of an application can be improved. This research details the development, analysis, and implementation of optimal kinetic energy recovery algorithms for surface permanentmagnet synchronous machines, interior permanent-magnet synchronous machines and induction machines. Braking events over user-defined time-interval lengths that include both stopping and slowing down to non-zero speeds are considered. Mechanical loads that include viscous friction, Coulomb friction, and aerodynamic drag are considered in the development of the algorithms. The trade-off between braking time and energy recovery is clearly illustrated and closed-form expressions for the optimal length of the braking time-interval are developed wherever possible to help control engineers design optimal braking trajectories. A universal optimal kinetic energy recovery algorithm is developed for all three types of electric machines under constant-flux operation. The theory is extended to include braking with final position constraints and braking under variable-flux methods of operation such as maximum-torque-per-amp and flux-weakening. The optimal control solutions are implemented using standard feedback controllers and are validated through simulations and experiments.