Extended State Observer Based Disturbance Rejection Control For Several Classes Of Uncertain Nonlinear Systems

Disturbances and uncertainties widely exist in the applications of control system design,such as measurement bias,parametric variation of the system,unknown external disturbances,etc.Disturbance rejection techniques are very mature in linear control system and it is also not a difficult work for nonlinear control system when system model is accurate.Many strategies have been proposed and successfully applied in engineering.However,for nonlinear system with complicated model,due to the influence of system parameter perturbation,unknown system structure and unavailable system states,the design of observer and controller are very complex,leaving quite large gaps in research.Especially uncertainties and disturbances meanwhile exist in system,disturbance rejection problem needs to be further explored.It is proved that there are not universal design methods for nonlinear systems.For instance,the mobile robot system with nonholonomic constrains and underactuated multi-joint manipulator system,which have complex kinematics and dynamics,together with the unavoidable disturbance in process of movement,it gives rise to difficulty in control system design such that the research is quite difficult.This kind of problem is important and meaningful in both theories and applications.In view of the above facts,this dissertation focuses on the disturbance rejection of several classes of uncertain nonlinear systems.The global robust regulation problem is studied for a class of cascade nonlinear systems subject to the non-vanishing external disturbance.The considered system is integral input-to-state stable(iISS)with unknown control directions.The proposed algorithm is used to analysis the tracking problem for fan speed control system.The output feedback stabilization problems are studied for a class of uncertain chained nonholonomic systems with non-vanishing disturbance and the parking problem of mobile robot systems with parametric uncertainties and deviation in the angle's measurement.The rotation speed regulation of permanent magnet synchronous motor(PMSM)system with uncertain load torque and stator resistance is solved via designing two extended state interconnected observers.The main techniques in controller design include adaptive control,backstepping method and extended state observer,etc.The main results and contributions of this dissertation can be summarized as follows:(1)The problem of global robust regulation is addressed for a class of cascade nonlinear systems.The considered system represents more general classes of nonlinear uncertain systems such as the much weaker iISS cascaded subsystem.The control coefficients are unknown,the unmeasured states appearing in the nonlinear uncertainties and the external disturbance additively in the input channel.Combined the ideas of the Nussbaum-type gain and the disturbance as a generalized state,a dynamic extended state observer(ESO)based on a Riccati differential equation is constructed to estimate system states and disturbance.It is shown that the global robust regulation problem is well addressed by the proposed method.In the simulation part,the fan speed control system is used as a practical example to demonstrate its efficacy(2)The problem of global asymptotic stabilization is considered for a class of nonlinear systems with iISS inverse dynamics and external disturbance.The investigated systems represent more general classes of nonlinear uncertain systems.Compare with the existing results,the type of the known function in model is further considered.An ESO is designed to estimate both the system states and the external disturbance by use of the disturbance as a generalised state.Based on the given reduced-order ESO,a new error dynamic system is constructed.The global asymptotic regulation is guaranteed under the presented anti-disturbance controller by the combined approach of backstepping and linear matrix inequality.The simulation result demonstrates the effectiveness of the given algorithm.(3)The problem of global output feedback stabilization is studied for a class of nonholonomic systems with non-vanishing external disturbances and nonlinear uncertainties.The effects of non-vanishing external disturbances were not considered in the previous researches for uncertain nonholonomic systems which implies that the existing control methods are ineffective to drive systems to the equilibrium point when disturbances exist.With the help of previous research,the non-vanishing disturbance is regarded as the generalized system state and an ESO is designed for the uncertain external disturbance in order to ensure the global asymptotical stability of the closed loop system.A novel observer/estimator is constructed for auxiliary system states and unknown parameter estimates.The integrator backstepping recursive technique based on a construction rule is applied to the design of the output feedback adaptive controller.The simulation example for a mobile robot with velocity and angle velocity measured deviation and disturbance is given to validate the effectiveness of the proposed method.(4)The problem of designing a robust output feedback controller is investigated for a class of mobile robot systems with parametric uncertainties,deviation in the measure angles,and non-vanishing external disturbances.An ESO is constructed with the gain from an offline time-varying Riccati matrix differential equation to estimate the unmeasurable states and external disturbances.This strategy yields a controller that applies time-varying output feedback.The proposed robust controller drives the system states to the origin asymptotically.Examples are provided to demonstrate the validness of the proposed approach(5)The problem of speed regulation is studied for a class of sensorless PMSM system.Few work investigates the robustness of the observer-controller scheme when the observability is lost.Meanwhile the load torque and stator resistance are as the constants.At here they are as the unknown and time varying generalized unmeasurable states.An ESO is designed to estimate both the unmeasurable system speed and uncertain parameters.A new interconnected observer scheme is constructed by the Kalman filter only using the measurements of motor current and voltage.A nonlinear backstepping controller based on the observer signals is developed and the stability of the closed-loop is guaranteed.The simulation result demonstrates the given control scheme has good performance for the change of the stator resistance and the load torque.