2025 lEEE International Conference on Predictive Control of Electrical Drives and Power Electronics (PRECEDE 2025)

Deadbeat predictive control for Permanent magnet synchronous motor uses model predictive control to select the alternative voltage vector that is closest to the ideal voltage vector. The more the number of alternative voltage vectors, the better the model predictive control performance, but the computational burden of the system will increase with the increase of the number of alternative voltages, so the control effect will be limited by the computational capability of the system. In this paper, a generalized n-order subdivision method of voltage vector hexagonal internal tangent circle is proposed based on the differential beat-free control of magnetic chain and torque of surface-mounted permanent magnet synchronous motors, where n-order equal divisions are made on the angle and radius of the internal tangent circle, and 6n²+1 alternative voltage vectors are obtained. In order to address the problem that the number of alternative voltage vectors increases with the increase of subdivision order, a simplified determination method is proposed to find the optimal order, and then determine the optimal voltage vectors by traversing the four points of the sector ring where the ideal voltage vector is located, and the maximum number of traversal times is fixed at four times and is independent of the subdivision order n. In order to further reduce the computational amount, the method of directly determining the optimal alternative voltage vector based on the ideal voltage vector magnitude and angle is proposed without MPC traversal. The simulation and real-time experimental results show that through the n-order subdivision and the simplified determination of the optimal alternative voltage vectors, it is easy to extend the number of voltage vectors arbitrarily to improve the control performance, and at the same time, the number of alternative voltage vectors can be fixed to four or without MPC traversal, which reduces the amount of computation, and achieves the optimization of the control and real-time performances at the same time.

permanent magnet synchronous motor,model predictive control,deadbeat control,calculation burden,real-time performance