Novel active-disturbance-rejection controller with embedded model

Abstract

The invention discloses a novel active-disturbance-rejection controller with an embedded model. The novel active-disturbance-rejection controller comprises an extended state observer module arranged in the inner ring and a proportional differential feedback module arranged in the outer ring; model information is embedded in the extended state observer module and used for observing the total disturbance of a system outside a nominal model, and then transforming a controlled object into a series integrator standard type by compensating the part of the total disturbance and known dynamics in realtime in the inner ring; the proportional differential feedback module is based on the series integrator standard type and adjusts he output control amount through parameter setting according to a quantitative relationship between an outer ring controller differential parameter kd and a second-order system damping coefficient zeta to achieve the control of the controlled object; the system is thesystem of the controlled object. The novel active-disturbance-rejection controller with the embedded model can reduce the burden of an extended state observer, achieves the purposes of more accurate disturbance estimation and set value tracking and reduction of the influence of noise on the output control amount, and provides convenience for parameter setting and adjustment in an actual control process.

Description

technical field [0001] The invention relates to the field of automatic control, in particular to a model-embedded novel active disturbance rejection controller. Background technique [0002] Active Disturbance Rejection Control (ADRC) is a new type of control technology proposed by Mr. Han Jingqing, which absorbs the essence of PID control's idea of ​​"eliminating errors based on errors", and attributes all uncertain factors acting on the controlled object to "Unknown disturbance", and design the extended state observer Extended State Observer (ESO), according to the input-output data of the object to observe the influence of the unknown part of the model and the external unknown disturbance on the control object, and then give the control amount to compensate these disturbances, The impact of disturbance on the controlled quantity is greatly reduced. [0003] However, because the mathematical model of the controlled object is not included in the proper estimation of the sy...

AI Technical Summary

Problems solved by technology

[0003] However, because the mathematical model of the controlled object is not included in the proper estimation of the system, it poses considerable difficulties to the accuracy of the estimation and the setting of the parameters in the subsequent control action

A large number of control simulations and practices have shown that the control effect

Images