Improved Observer Based Active Disturbance Rejection Control And Its Application

Real-time estimation and suppression problems for disturbances are popular in many network-based location and control applications.For examples,in high-precision vehicle location and control based on ad-hoc network,the wireless communication of the high-speed mobile vehicle has some characteristics,such as fast change of dynamic networks,fast fading of channels,Doppler effect,and all kinds of coexisting disturbances,which seriously affect the stable operation of systems.Therefore,the real-time estimation and suppression of disturbances are the key to ensure the stable operation of the system.In past decades,several methods for real-time estimation and suppression of disturbances have been developed.For instance,in 1980s,Ohnishi,a Japan professor,investigat-ed a disturbance observer based control(DOBC);In 1990s,Prof.Han Jingqing of the Chinese Academy of Sciences presented an active disturbance rejection control(ADRC)method based on the extended state observer(ESO);An extended high-gain state observ-er(EHGSO)was proposed by the professor Khalil at Michigan university in 2008,et.al.The basic idea of these methods is that the system disturbance is augmented into the system state to estimate the disturbance.These methods have the advantage that both model uncertainties and unmodeled parts can be estimated by regarding them as distur-bances of system,which extend the application scope of the linear model to some extent and simplify the control problems of some complex systems.However,because the band-width of mobile network and the computation capacity of embedded devices are limit-ed,the implementation of various ADRC observers on mobile chips is always a difficult problem.Since the ADRC has been put forward,it has received considerable attention in industrial fields,which apply the idea of“observation+compensation”to realize the controller design;that is to say,the disturbance,unmodeled dynamics and uncertainty in the control system is firstly estimated,and the effects of them are compensated by a non-linear feedback.The ADRC method has the following advantages:the control problem of certain systems and uncertain systems can be dealt with in a unified way;it does not require that the mathematical model of the external disturbance is known or the external disturbance is observable;the linear and nonlinear systems can be dealt with in a unified way.On the other hand,the drawback of this method is that the design method of the ob-server and controller parameters to some extent depends on experience knowledge instead of strict theoretical basis.The main contributions and innovations of this paper are as follows:Firstly,the ADRC method is introduced to study the stabilization problem for Euler-Bernoulli beam equation with constant boundary output disturbance.To this end,we present a tractable tuning method for the case with a single parameter.Then the well posedness and uniqueness of the solution based on the state space are proved.This method requires the displacement signal and not the speed signal,which simplifies the design of the control system and provides a new solution the control system for the same kind.Secondly,three new ADRC methods are presented for the ADRC control problem of lumped parameter systems.The one is the ADRC control based on the H_?ESO which can provide an optimal observer parameters with a prescribed performance,has smaller estimated relative errors and a smaller convergence rate,and is suitable for the case of the low frequency of disturbances.The second is the ADRC method based on an impulsive ESO,which can achieve fast convergence rates by reseting the observer state,and can shorten the settling time of estimation errors,but the relative error of the estimation is slightly larger,especially for the case of high frequency of disturbances.In addition,in view that the PID method is popular in industrial fields,this thesis also give the ADR-C method based on a simple PID observer and the design method of the PID observer parametes,which is convenience to an embedded ADRC systems.Finally,the application of ADRC technology in intelligent vehicle control is studied.Two ADRC algorithm are respectively designed for the problems of distance keeping control and suspension control.And a useful attempt is made to the application for the new algorithm in the industrial field.