Real-Time Implementation for Improvement of Weighting Coefficient Selection using Weighted Sum Method for Predictive Torque Control of PMSM Drive

In this article, selection of weighting coefficients is improved using the weighted sum method (WSM)-based objective function optimization for predictive torque control in PMSM drive under steady and dynamic conditions. Essentially, the empirical approaches used for weighting coefficient (WC) selection are intuitive, and the computation time is higher. In WSM approach, the error terms of stator flux and torque are obtained for each available state of the VSI. The attained dataset is normalized to 0-1 scaling, and then normalized error terms of both control parameters are combined into a single performance index for each switching state with suitable WCs. By maximizing this performance index, the optimal choice of switching state will be selected, with respect to priority given to control variables in objective function. Further, it results in reduced average switching frequency, %ripples (torque and flux), and stator current THD. The dSPACE-DS1104 R&D controller board is used for real-time implementation of conventional and proposed control algorithms under low speed, high speed, speed reversal, speed dynamics, and torque dynamics and also simulated in MATLAB/simulink environment. In order to verify the efficacy of PTC based on WSM, the hardware and simulation results were compared with conventional DTFC and CPTC strategies.