Optimising Weighting Factor Design for Model Predictive Torque Control in IPMSM Drives: A Grey Relational Analysis Approach

AC electric drives have widely incorporated model predictive control (MPC) over the past decade. Even with a variety of proposed solutions, there are still challenges related to designing influential weighting factors (WF), reducing parameter dependence, mitigating current/torque harmonics, managing variable switching frequencies and minimising computational complexity. The performance of controllers can be significantly reduced by incorrectly selecting WFs. The cost function of conventional finite-control set model predictive torque control (MPTC) includes multiple control objectives with different units of measurement. WFs are therefore introduced so that these objectives can be prioritized. An extensive comparison of WF design techniques for MPTC strategies in interior permanent magnet synchronous motors (IPMSM) drives is presented in this paper. The first step in the MPTC process is to present a description of conventional MPTC techniques with manual WF tuning, VIKOR and the entropy method for WF optimisation. To address the issue of WF design, the grey relational analysis (GRA) approach is proposed. This method separates the flux and torque objective functions using a single cost function and the grey relational grade is employed for each sampling to determine the appropriate action. Experimental validation of the proposed approach is performed by means of real-time hardware-in-the-loop (HIL) simulations on a prototype IPMSM drive with a TMS320F28335 floating-point digital signal processor. Finally, MPTC is compared with conventional MPTC, VIKOR and entropy-based WF optimisation methods to compare its performance.