Xuezhu Mei

Model-Based Predictive Direct Control Strategies for Electrical Drives: An Experimental Evaluation of PTC and PCC Methods

Model-based predictive direct control methods are advanced control strategies in the field of power electronics. To control an induction machine (IM), the predictive torque control (PTC) method evaluates the electromagnetic torque and stator flux in the cost function. The switching vector selected for the use in the insulated gate bipolar transistors (IGBTs) minimizes the error between references and the predicted values. The system constraints can be easily included. The predictive current control (PCC) strategy assesses the stator current in the cost function. The weighting factor is not necessary. Both the PTC and PCC methods are very useful direct control methods that do not require the use of a modulator. In this paper, the PTC and PCC methods are carried out experimentally for an IM on the same test bench. The behaviors and the robustness in steady state and the performances in transient state are evaluated.

Predictive current control of an induction machine by using dichotomy-based method

This paper presents a novel model predictive current control method for induction machine with constant switching frequency. Using model predictive current control, a cost function considering error between the reference current and the predicted current is designed. A numerical method of quasi-continuous reference voltage calculation based on dichotomy is applied to generate the optimized reference voltage vectors which are selected as the input of PWM before switching the inverter. This can significantly improve the current quality, reduce the torque ripples and ensure fixed switching frequency of the inverter. The proposed method is verified by simulation, in which the induction motor operates with high performance.

A circular dichotomy-based method for model predictive control with fixed switching frequency for electric drives

A new model predictive control method of electric drives with fixed switching frequency is presented. It proposes a numerical method based on circular dichotomy to calculate and select reference voltage vectors, which are given to pulse width modulator to generate switching signals. It is verified and compared with the existing dichotomy-based method in a predictive current controlled electric drive system with space vector pulse width modulation through simulation. Compared to existing method, the proposed method can significantly reduce code complexity and calculation effort without deteriorating system performance.

PLL with Piecewise Judgement Function for SMO Beased Sensorless Control of PMSM

Sliding-mode observer (SMO) is widely used in the sensorless control strategy of PMSM due to its robustness. And phase locked loop (PLL) is usually applied to reduce the chattering introduced by SMO. However, the conventional PLL shows a type of intrinsic error during speed reversal or startup. To solve the problem, a novel PLL with modified control function is proposed and it can achieve excellent performance in simulation.

Model predictive control for electrical drive systems-an overview

This paper reviews the classification and application of the model predictive control (MPC) in electrical drive systems. Main attention is drawn to the discrete form of MPC, i.e. finite control set model predictive control (FCS-MPC), which outputs directly the switching states of power converters. To show the diversity and simple realization with various control performances of the strategy, in this paper, several different FCS-MPCs with their working mechanisms are introduced. Comparison of FCS-MPC with conventional control strategies for electric drives is presented. Furthermore, extensive control issues, e.g. encoderless control and disturbance observation are also included in this work. Finally, the trend of research hot topics on MPC is discussed.

Deadbeat Boolean logic predictive current control for induction machine without cost function

A modified model predictive control without cost function through deadbeat control for expected voltage calculation and Boolean logic for applicable voltage vector selection is proposed. Because of the absence of PWM and cost function, this method reduces simultaneously the systems hardware complexity and the calculation efforts as well as model dependency. The proposed method is verified and compared with the conventional predictive current control system of induction machine by simulation.