Abstract

Two problems can cause control performance degradation on permanent magnet synchronous motor (PMSM) systems, namely, fluctuation of PMSM parameters and the time delay between current sampling and command value update. In order to reduce the influence of these problems, a new current-predictive control strategy is proposed in this article for medium- and high-speed PMSM. This strategy is based on the discrete mathematical model of PMSM. This new control strategy consists of two main steps: First, an integrator is applied to calculate current compensation value; second, the predictive current value is obtained through deadbeat-current predictive method. The stability of predictive control system is also proved in the article. With this deadbeat-current predictive control scheme, the real current can reach the desired value within one control-step. Based on this new current control method, Luenberger observer and phase-locked loop position tracker is applied in this article. Experimental results for 0.4 kW surface-mounted PMSM confirm the validity and excellent performance for parameters fluctuation of new current predictive control.

Highlights

  • Nowadays, permanent magnet synchronous motors (PMSMs) have been widely used in industrial applications due to their high power and torque density, small size and wide-speed-range

  • The experimental drive setup consists of a 0.4 kW PMSM connected to a dynamometer machine and a load motor used for loading

  • Both of load motor and tested motor are supplied by a inverter which is connected with a common direct current (DC) link and controlled by Infineon TC1782, the sampling-time and PWM frequency is fs =

Read more

Summary

Introduction

Permanent magnet synchronous motors (PMSMs) have been widely used in industrial applications due to their high power and torque density, small size and wide-speed-range. Compared with aforementioned sensorless control methods, the method in this paper can decrease signal phase-lag and few fluctuations on model parameters will not produce the large position error, because a current compensation method is proposed to reduce current ripple. Key PWM time-sequence logic relationship between current sampling and voltage command update is established Based on this relationship, a deadbeat-current predictive method is proposed to calculate real-time voltage command value.

PMSM Mathematical Model
Second-Order PLL Rotor Position Tracker
Proposed and Conventional Current
Conventional Current Predictive Model
Deadbeat Linearized Current Predictive Model
Current Compensation
System
Experimental Results
Experimental Results of Deadbeat-Current Predictive Control Method
Experimental
10. Current
Experimental thecurrent
12. Current responsecomparison comparison based based on in in controller
Experimental Results of Rotor Position and Speed Sensorless Observer
Conclusions
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call