A Multi-Input Multi-Output Model Predictive Direct Torque Control for Dual Mechanical Port Machine

Abstract

In this paper, a multi-input multi-output (MIMO) model predictive direct torque control (MPDTC) strategy for a dual mechanical port machine (DMPM) is proposed. A MIMO state space model is developed, incorporating the electromagnetic interaction between the stator, inner rotor, and outer rotor of the DMPM. The prediction model is then derived to adjust the currents of stator and inner rotor windings, enabling independent control of the electromagnetic torques of the mechanical ports. Compared to single-input single-output based MPC and PI-based field-oriented control strategies, the proposed approach effectively mitigates the undesired electromagnetic interactions, guaranteeing smooth speed control of both mechanical ports. The simulation results carried out by MATLAB/Simulink verify the precise performance of the DMPM in different operating modes. The Hardware-in-the-Loop results further confirm that the proposed direct torque control not only regulates the inner and outer rotor independently, but can also be executed in real-time without posing any computational burden on the processor. A sensitivity analysis for evaluating the performance of the proposed strategy in the presence of variations in the electrical parameters of the DMPM is also provided.INDEX TERMS Dual mechanical port machine, model predictive control, direct torque control, multi-input multi-output system, hardware in the loop.