The performance of a standard model predictive controller (MPC) is directly related to its predictive model. If there are unmodeled periodic disturbances in the actual system, MPC will be difficult to suppress the disturbances, thus causing fluctuations of system output. To solve this problem, this paper proposes an improved MPC named predictive-integral-resonant control (PIRC). Compared with the standard MPC, the proposed PIRC could enhance the suppression ability for disturbances by embedding the internal model composing of the integral and resonant loop. Furthermore, this paper applies the proposed PIRC to PMSM drives, and proposes the PMSM control strategy based on the cascaded PIRC, which could suppress periodic disturbances caused by the dead time effects, current sampling errors, and so on. The experimental results show that the PIRC can suppress periodic disturbances in the drive system, thus ensuring good current and speed performance. Meanwhile, the PIRC could maintain the excellent dynamic performance as the standard MPC.

Traditional perspective considers that there is a trade-off between the postfault operational performance and the implementation complexity of the control scheme. To approach this problem, this paper proposes a natural full range minimum-loss strategy by analyzing the current relation between torque subspace and harmonic subspace. The harmonic subspace current references satisfying full range minimum-loss are generated by a simple expression according to the real-time subspace current vector features. Compared with existing fault-tolerant control methods, the proposed strategy does not need complicate fault diagnosis, as well as offline or online optimization process. The experimental results demonstrate that the proposed fault-tolerant control strategy can achieve smooth prefault to postfault transition and FRML simultaneously.

Considering the periodic disturbances caused by non-ideal factors like dead time effect, permanent magnet flux harmonics and current sampling errors in the permanent magnet synchronous motor (PMSM) drives, this paper establishes an augmented state space model (SSM) which contains information of periodic disturbances, and then more accurate description for the PMSM system is achieved. Meanwhile, since the three augmented state variables, related to periodic disturbances in the augmented SSM, cannot be measured directly, a novel periodic disturbance observer (PDO) depending on the Kalman filter is designed to estimate these state variables. On the basis of the augmented SSM and PDO, a linear-quadratic-regulator (LQR)-based state feedback controller, which can achieve a successful suppression of periodic disturbances, is proposed to ensure the PMSM acquires a more smooth speed output. Finally, the proposed PDO-LQR is tested in a 6kW PMSM, and the experimental results show that the proposed controller can efficiently improve the steady-state performance of the PMSM system without sacrificing dynamic-state performance.

Finite control-set model predictive control (FCS-MPC) has received extensive attention because of its excellent dynamic performance, however, it is still need to be studied in terms of torque ripples reduction and parameter robustness. To reduce the steady-state torque ripples and maintain the appropriate anti-parameter perturbation ability, this paper proposes a model predictive direct duty-cycle control (MPD2C) strategy based on “variable control-set (VCS)”. In the proposed strategy, the direct mapping relations between the electromagnetic torque, flux and three-phase duty-cycles has been established, by exploring the dual relationship of vector synthesis and duty-cycles. On this basis, the novel predictive model and VCS that using the three-phase duty-cycles as the key variable are constructed. Finally, the proposed strategy determines the three-phase duty-cycles from VCS by employing a two-stage optimization mechanism, which means the proposed strategy can generate virtual vectors with multiple IGBT switching patterns. Then, the high-precision torque and flux adjustment would be achieved. The experimental results show that VCS-MPD2C has excellent static and dynamic control performance, acceptable execution efficiency and relatively good parameter robustness.

The synchronization ability of multi-motor system is important to the textile manufacturing and printing industries. The conventional synchronization control strategy has good steady-state performance, while its dynamic response is limited, constrained by the performance of the main controller. The model predictive controller is recognized as a controller with excellent dynamic performance. The explicit model predictive controller (EMPC) will be employed as the main controller of multi-motor system in this paper. EMPC has potential in real-time control because it can rely on offline optimization methods to avoid iterative optimization. To adapt EMPC to multi-motor systems, the unified predictive model of multi-motor drive is constructed. The multiple voltage and current constraints are simplified and all feasible control laws for multi-motor systems were successfully solved. And, an efficient binary tree polling algorithm is designed to find the optimal control law online. Meanwhile, to suppress the impact of non-ideal factors, such as load disturbance, parameter mismatch, harmonic interference, etc. on the model accuracy, the Kalman observer is designed to compensate the load disturbances and sensor errors. As verified from experimental data, the dynamic response time of EMPC has reduced to 50% of existing methods, while maintaining steady-state performance.