Robust gain scheduling control method based on regional pole assignment

Abstract

The invention provides a robust gain scheduling control method based on regional pole assignment. The realization principle is as follows: the controlled object is divided into a linear time invariant part and a parameter module by the method, wherein the parameter module is used for reflecting the nonlinear characteristic, time variant characteristic and certain uncertainty of the object; the nominal part is linear time invariant and the parametric variation part varies along with variation of the parameter module of the controlled object, thus ensuring the closed-loop system to satisfy the robust stability and robustness and have specified dynamic characteristics; the performance indexes are written into integral quadratic constraints under a unified framework according to the index requirements; in particular, the dynamic characteristics can be directly reflected by closed-loop poles on the position of the complex plane; therefore the index requirements of the dynamic characteristics are realized through regional pole constraints; and the constraints given by the performance index requirements are transformed into linear matrix inequality (LMI) systems and a solver in an LMI toolbox provided by a matrix laboratory (MATLAB) is utilized for solving, thus finally obtaining the controller satisfying the requirements.

Description

technical field [0001] The invention relates to a robust gain scheduling control method, in particular to a design method of a robust gain scheduling controller using a pole configuration scheme to realize specified dynamic performance, and can be widely used in time-varying nonlinear multiple-input multiple-output with high requirements on dynamic performance The controller design domain of the system. Background technique [0002] Designing controllers for time-varying nonlinear MIMO systems using classical control theory methods must perform Jacobian linearization for each input and output channel at a series of operating points. The resulting controller loses global properties and is not guaranteed to be stable due to possible unstable zero-pole cancellation. However, the controllers designed based on modern control theory, such as linear quadratic regulators, have poor robustness and require extremely high accuracy of the model. Although the controller designed based ...

AI Technical Summary

Problems solved by technology

However, due to the complexity of the algorithm and its inherent conservatism, it is difficult for the robust gain scheduling control method to ensure that the system has specified dynamic characteristics

Images