Current prediction control method and apparatus for permanent magnet synchronous motor (PMSM)

Abstract

The invention discloses a current prediction control method and apparatus for a permanent magnet synchronous motor (PMSM). The method comprises the steps: acquiring the motor stator current and stator voltage under two-phase rotating coordinates; constructing a current and perturbation sliding-mode observer, and predicting and obtaining the motor stator current and the parameter perturbation estimated value of the next moment; according to the principle that the practical current value of the next moment is equal to a current command value, obtaining a PMSM current prediction control expression; taking the current value, predicted by the observer, of the next beat, as current feedback, and obtaining the command voltage of a motor driver through calculation; and further adding the command voltage and the parameter perturbation estimated by the observer together, taking the sum as the final motor driver command voltage, and finally generating a pulse signal for controlling motion of a switch tube through an SVPWM device. The current prediction control method for a permanent magnet synchronous motor can predict the current of the next beat and estimate parameter perturbation at the same time, and can effectively restrain the current tracking static error and reduce the motor harmonic wave current and optimize the current control performance when a parameter mismatching problem exists.

Description

technical field [0001] The invention relates to the technical field of motor control, in particular to a current predictive control method and device for a permanent magnet synchronous motor. Background technique [0002] On the basis of the discrete mathematical model of the motor, the deadbeat current predictive control technology can accurately predict the voltage vector of the next control cycle, so that the motor current can accurately track the command current (or current command ), with excellent dynamic and steady-state characteristics. At present, deadbeat current predictive control technology has been widely used. However, this control technology is essentially a model-based current control method, which is highly dependent on the motor model. When the parameters of the motor model do not match, it will cause static errors and oscillations in the motor current, which will affect the control system. performance. [0003] For this reason, scholars at home and abro...

AI Technical Summary

Problems solved by technology

In existing methods, such as: patent document CN201510455057.5 and document "An adaptive robust predictive current control for PMSM with online inductance identification", the influence of inductance parameter mismatch on current tracking performance is suppressed through inductance identification, but the influence of inductance parameter mismatch on current tracking performance is ignored, but the motor model is ignored. Influence of Mismatch Between Resistor and Permanent Magnetic Flux Linkage Parameters

Although there are other control methods that can compensate the model parameter mismatch, they need to introduce PI adjustment or integral adjustment, which makes the control system more complicated. For example, the literature "Analysis and implementation of a real-time predictive current controller for permanent-magnet synchronous servo drives》

[0004] In order to improve the current dynamic response performance, in order to reduce the impact of one-beat control delay on the system control performance, various control methods have been proposed, but none of these methods can eliminate the current static difference caused by parameter mismatch, for example: patent document CN201210379496 .9

Therefore, there is currently no current control method that can satisfy the following conditions at the same time: 1) It can estimate the parameter disturbance while predicting the next beat current value; 2) Compensate the control delay and current static error at the same time when the parameters are mismatched; 3 ) can be easily integrated with existing current predictive control methods

Images