The current control in permanent magnet synchronous machine (PMSM) drives is a challenging problem which has to deal with physical and real-time constraints. The goal of this paper is to provide a control solution that can deal effectively with these challenges. The objective is to adjust the stator voltages in order to obtain fast and monotonic transient current response while satisfying the existing constraints. To this end, using the PMSM model in the rotor coordinates, an explicit one-step ahead predictive control law which uses tools from Lyapunov theory is developed. Firstly, a polytopic approximation of the quadratic constraints is used to reduce the complexity. Then, the optimization problem is reduced to a linear program. Furthermore, a piecewise affine explicit control law is obtained via multi-parametric linear programming. As a consequence, the on-line computation of the control law is reduced to a point location problem yielding a fast control method suitable for real-time implementation. Secondly, the speed dependent terms dependent terms of the mathematical model are considered as disturbance. Then, a state and disturbance observer is designed using a quadratic Lyapunov function to guarantee asymptotic stability of the estimation error. Real-time results obtained in an industrial hardware-in-the-loop test-bench are reported and analyzed.

A relevant challenge in hybrid electric vehicles (HEVs) and full EVs is the torque control of externally excited synchronous machines (EESMs). Effective torque control requires an efficient solution to the state reference generation problem, which is a nonlinear nonconvex optimization problem. The goal of this paper is to develop a state reference generation algorithm based on the gridding of the state and output spaces. First, an approximation defined over a cubic partition of the torque function with a piecewise affine (PWA) function is made. As a result, the state reference generation problem is reduced in each cube to solving a convex optimization problem. Moreover, this approach provides guarantees about the error bound introduced by the state reference generation procedure for the full operational state-space. To illustrate the effectiveness and robustness of the proposed algorithm, several real-time results obtained on an industrial hardware-in-the-loop (HIL) test-bench are presented. The obtained results show significant improvement compared with existing state-of-the-art reference generation methods.

No abstract

No abstract

No abstract