Analysis And Design On Typical Adaptive Controls For Uncertain Nonlinear Systems

Nonlinearities and uncertainties are complex and diverse,and their coupling coexistence seriously affects system performance and control accuracy.For more intrinsic nonlinearities and more types of uncertainties,it is urgent to develop an adaptive control which owns universality and stronger feedback capability.With the rapid development of network communication technology,computer technology,and control technology,the demand for resource conservation in computation/communication/energy has become more urgent.Event-triggered control has significant advantages in reducing information transmission and control updating.Therefore,for uncertain nonlinear systems,it is of great theoretical significance and practical value to construct an adaptive event-triggered strategy with strong feedback ability.In this dissertation,the analysis and design problems of typical adaptive controls for several classes of uncertain nonlinear systems are investigated.A switching adaptive state feedback control with strong feedback compensation capability is constructed and an adaptive output-feedback compensation mechanism with less conservativeness is established.A energy-efficient switching adaptive eventtriggered output-feedback control strategy is further proposed to save resources.In addition,with the aid of adaptive technique and potential functions,adaptive event-triggered connectivity-preserving control strategies are developed,which provides theoretical support for the event-triggered control problems of more general uncertain nonlinear multi-agent systems.The main contents of this dissertation include the following five aspects:(Ⅰ)Switching adaptive state-feedback stabilization of nonlinear systems with unknown control directions and unknown input powersUnknown input powers,which are a new type of system uncertainties,largely challenge the feasibility of continuous feedbacks.For this,a parameterized controller in a specified form is introduced,which can globally stabilize the system once the design parameter involved belongs to a certain specified set.However,the controller cannot directly be implemented since the set depends heavily on the unknown input powers.To make the parameterized controller implementable,a new switching strategy with strong feedback capability is proposed to instantly update the parameters of the controller,thereby achieving the effective compensation of the system uncertainties and excluding Zeno behavior while ensuring the global boundedness and convergence(to zero)of all the closed-loop system signals.(Chapter 3 in the dissertation)(Ⅱ)Adaptive output-feedback stabilization for nonlinear systems with unknown control directions and intrinsic growthIn the context of unknown control directions and intrinsic unmeasurable-statesdependent growth,an adaptive output feedback control strategy with less conservativeness is developed.First,a varying parameter which takes value 1 or-1,is introduced to capture the unknown control direction,and accordingly a switching varying mechanism is delicately constructed to decide when to update the parameter.A dynamic high gain is designed to compensate for the unknown control coefficients and unknown growth rate,and a dynamic-gain-based observer is constructed.Hence,the designed controller ensures no Zeno behavior of the closed-loop systems,and meanwhile,guarantees the boundedness of all the closed-loop system signals and the convergence(to zero)of the system state and observer state.(Chapter 4 in the dissertation)(Ⅲ)Switching adaptive event-triggered output-feedback stabilization for uncertain nonlinear systemsGlobal stabilization via output feedback for uncertain nonlinear systems in the event-triggered framework is studied.The systems under investigation admit unknown control directions,unknown polynomial-of-output growth rate and unmeasurable-states dependent growth.Actually,in the context of the unknown intrinsic growth,there has been no any continuous control strategy that has allowed the unknown control directions so far.Hence,one cannot solve the eventtriggered control problem(based on corresponding continuous feedback)as done in the emulation-based method.In view of the unsolvability,a nonemulationbased strategy is pursued,directly conducting event-triggered control design.First,a parameterized output feedback controller incorporating a dynamic high gain is constructed,and the adjustable parameter therein owns varying sign and keep fixed magnitude.Then,an event-triggering mechanism incorporating switching varying mechanism is developed to not only decide when the controller is sampled/executed but also determine which constant value the adjustable parameter takes.Based on the mechanism,the generated event-triggered control has strong feedback ability,making it possible to solve the control design problem in the event-triggered framework.Particularly,the threshold and suspension time vary with the online adjustment of dynamic gain,thereby ensuring timely enough sampling/execution to cope with system uncertainties and execution error.Moreover,in the performance analysis,infinite switchings of adjustable parameter are excluded and a positive lower bound for inter-execution intervals is ensured.(Chapter 5 in the dissertation)(Ⅳ)Adaptive event-triggered connectivity-preserving consensus control of nonlinear multi-agent systems with unknown control coefficientsAdaptive event-triggered consensus of uncertain nonlinear multi-agent systems with connectivity preservation is studied.The systems under investigation admit nonidentical unknown control coefficients and heterogenous nonlinearities coupling with parameter uncertainties.For this,a group of potential functions are introduced acting as control barrier functions to constrain the relative distances between agents within the communication range for all time.Also,two dynamic gains are specialized for each agent to compensate the system uncertainties,suppress the system nonlinearities and overcome the negative effect of the execution error.Thus,the designed adaptive event-triggered control protocol achieves the connectivity preservation of communication graph and the consensus of system while ensuring a positive lower bound for the inter-execution intervals to exclude Zeno behavior.Particularly,the proposed event-triggering communication mechanism decides asynchronous samplings and executions,which largely reduces information transmission and control updating.Finally,an extended study is conducted on a leader-following scenario.(Chapter 6 in the dissertation)(Ⅴ)Adaptive dynamic event-triggered connectivity-preserving formation control of nonlinear multi-agent systems with unknown control directionsAdaptive dynamic event triggered formation control for uncertain nonlinear multi-agent systems with connectivity preservation is studied.Compared with the systems studied in Chapter 5,the systems under investigation allow unknown control directions.To this end,Nussbaum-gain technique and potential functions are combined to establish an effective connectivity-preserving compensation mechanism.To further reduce information transmission and control updating,a internal dynamic variable is introduced,based on which,a dynamic eventtriggering communication mechanism is developed.Thus,the generated eventtriggered control protocol has strong feedback ability to ensure the connectivity of the communication graph while achieving the desired formation.Through simulation comparison,the designed dynamic triggering mechanism has obvious advantages over the static one in reducing the number of event triggers.(Chapter 7 in the dissertation)...