Minimum variance-based nonlinear model prediction controller design method for permanent magnet servo system

Abstract

The invention discloses a minimum variance-based nonlinear model prediction controller (NMPC) design for a permanent magnet servo system, and belongs to the technical field of high-performance servo control systems. The method comprises the steps as follows: firstly, a nonlinear mathematic model of a PMSM is built under a dq coordinate system; secondly, a cost function of the prediction control rate of the PMSM is selected on the basis of the model according to the minimum variance prediction control theory; and the selected cost function is minimized to search for the optimal control rate in a receding horizon, so that output of the system within prediction time can track a given reference value to reach the prediction control target. Through a receding horizon control strategy of model prediction control (MPC), adverse effects caused by various system model parameter changes and uncertain random disturbance can be processed online by the control system; and the disturbance resistance of the control system is strengthened. Through an experiment, high-rotating speed tracking performance and good robustness of the method in an actual control system are verified.

Description

technical field [0001] The invention relates to a design method of a nonlinear model predictive controller of a permanent magnet servo system based on minimum variance, and belongs to the technical field of high-performance servo control systems. Background technique [0002] At present, the AC permanent magnet servo system is widely used in wind power generation, industrial robots, new energy vehicles and other fields. High-performance control occasions require PMSM to have a strong speed tracking ability, and to ensure the strong performance of the system under system parameter changes and load disturbances. robustness. However, the strong coupling and nonlinearity of the PMSM system, the uncertainty of model parameters, external disturbances and end effects make it difficult for traditional control methods and conventional linear control methods to effectively control it. [0003] In recent years, in order to improve the speed tracking and anti-disturbance characteristic...

AI Technical Summary

Problems solved by technology

Model predictive control was originally proposed for linear systems. However, since PMSM is a very typical nonlinear, multivariable, and strongly coupled model system, when the model parameters are inaccurate, the output prediction of the linear model is more accurate than the actual deviation. Large, the effect of optimal control cannot be achieved, and it is difficult to achieve high-performance control of the motor, so it can no longer be simply processed by the linear model predictive control algorithm

Images