Adaptive Output Feedback Tracking Control Method For Smart Material-Driven Systems

In the background of high precision equipment manufacturing applications,smart material-driven systems have attracted considerable amount of attention in the field of the micro and nano positioning.It is due to the fact that the smart material drivensystems have overcome the drawbacks of the conventional actuation devices by virtue of the intelligent characteristics of the smart materials in sensing and responding to stimuli.However,the complex hysteresis behavior,which inherently occurs in the smart material-driven systems,is the main nonlinear trait that affects the tracking performance of the system,rendering the modeling and control for the system more challenging.In addition,when controlling the system for high precision trajectory tracking,the main difficulty is how to design an effective control strategy for the smart material-driven systems with input hysteresis to arrive at a trade-off between the control performance and device transmission burden,thus guaranteeing the effectiveness and applicability of the system.In this thesis,to address the primary challenge of the control performance constraint and device resource constraint in smart material-driven systems,an extensive study of the comprehensive modeling and control approaches is developed.The proposed control strategies are in the framework of backstepping method by using techniques such as adaptive control and neural networks.This thesis mainly focuses on three aspects: the system hysteresis characteristics,system transient performance,and device burden.The main research contents are summarized as follows.Firstly,with different excitation signals applied to the smart material-driven system,the multi-valued mapping property and dynamic characteristic of the hysteresis behavior are analyzed from the perspective of experiment.Based on this,by designing the auxiliary function with respect to the model features,a generalized Duhem model(RF-Duhem)with relaxation function and modified function-based Prandtl-Ishlinskii(MFPI)model are proposed,respectively.The parameter identification of these two models is implemented by the modified bacterial foraging algorithm with an adaptive adjustment mechanism.The proposed two models can accurately represent the hysteresis nonlinearity with the help of the auxiliary function;and their effectiveness is demonstrated by experimental results.The problem of the tracking control for the smart material-driven system with unmeasurable hysteresis and state information is investigated.An extended hysteresis observer is constructed by incorporating the unknown hysteresis as an extended state.Then,an adaptive dynamic surface control algorithm is developed for the smart material-driven system,in which the system states can be estimated by means of the proposed observer.The asymptotic convergence of the closed-loop system signals is demonstrated through the theoretical analysis.Comparative experimental results are provided to illustrate the validity of the constructed controller.To tackle the problem of achieving high-precision tracking control task for the smart material-driven system that is constrained by device resources,a finite-time quantization control strategy is proposed.By taking the system output as a feedback,a new composite quantizer is designed.This quantizer can not only achieve the state estimation but also serve the purpose of adjusting the device transmission bandwidth by the adaptive parameter.To eliminate the negative effect of the input hysteresis on the control performance,an estimated inverse compensator is constructed based on the RF-Duhem model.In addition,considering that the time delay in the signal transmission process is unavoidable,the adaptive output feedback control scheme is designed to eliminate the coupling effects of the quantization error,uncompensated hysteresis error,and time delays.Moreover,by introducing the power in the virtual control law,error compensation signal,and control law,the finite-time convergence of the tracking error is guaranteed.Experimental results indicate that the proposed controller achieves superior transient and steady-state tracking performance under the finite time effect,meanwhile taking full use of the limited device resources with the new composite quantizer.The problem of how to conserve the device resource of smart material-driven system with uncertain initial values while balancing transient performance is investigated.An extended fixed time lemma is developed by introducting a scaling function.The upper bound of the convergence time based on this lemma is only related to the system parameters;and there is no need to include the system initial value as a priori.A nonlinear auto-regressive moving average model with a hysteresis term(HTNARMAX)is provided to portray the hysteresis characteristics in the system.The unknown parameters of the hysteresis model and system states are simultaneously estimated by a composite hysteresis observer.In addition,an improved tracking differentiator is designed to cope with the insufficient performance of the first-order filter in dynamic surface control.In accordance with the relative threshold eventtriggered mechanism and designed fixed time lemma,an adaptive fixed time event triggering controller is constructed.From the theoretical analysis and experimental results,it is concluded that all signals of the closed-loop system can achieve the fixedtime convergence.The signal transmission is performed only when the event-triggered condition is satisfied,thus reducing the device transmission burden.The problem of how the smart material driven-system can achieve less computational complexity and save device resource utilization with the guarantee of control performance is investigated.A prescribed settling time regulator is constructed;it allows the designer to predetermine the controller convergence time and tracking error band.By means of this regulator,the fractional power terms in the observer or controller design can be avoided,and the computational burden of the control algorithm can be significantly reduced.The hysteresis nonlinearity is eliminated by constructing a new coordinate transformation and hysteresis auxiliary system.The hysteresis modeling error is used as the input of this auxiliary system.A predefined time control scheme is designed with the aid of the adaptive technique and backstepping control method.An event-triggerred mechanism is introduced in the controller implementation to improve the utilization of device resources.The experimental results reveal that the proposed approach maintains the system transient performance within the preset regulation time and tracking error band with the effect of the prescribed settling time regulator.Moreover,the cumulative number of the control signal with the eventtriggered effect is obviously less than that of the time-triggered control,thus realizing the conservation of device transmission resources.