Jan Richter, Martin Doppelbauer

IEEE Trans. on Industrial Electronics, Vol. 63, No. 6, pp. 3915-3924

High-performance permanent-magnet synchronous machines show nonlinear behavior due to saturation effects. Nonlinear differential equations describe these phenomena which makes feedback control challenging. This paper presents a unified approach based on the current, flux linkage and voltage plane to visualize the physics of machine dynamics. A predictive control method to precisely control arbitrary dynamic trajectories in all planes can then be developed. Four real-time strategies for dynamic trajectory identification are deduced: two straight trajectories, a strategy that yields a fast torque response, and a strategy that reaches the reference currents in minimal time. The performance of the four approaches is analyzed using simulations and test bench measurements. Advantages and disadvantages of specific trajectories are identified. It is thus shown how the dynamics of permanent-magnet synchronous machines with nonlinear magnetics can be controlled both precisely and optimally.