Font Size: a A A

Intelligent Control Of Uncertain Nonlinear Systems Based On Event-Triggered Mechanism

Posted on:2020-07-23Degree:DoctorType:Dissertation
Country:ChinaCandidate:L CaoFull Text:PDF
GTID:1368330602956210Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
With the maturity and wide application of computer technology and communication technology,communication network is used to transmit and share data among many systems.The communication networks are usually shared by different nodes of the system,and the transmission channel bandwidth is limited between each node,which may lead to the waste of communication resources.Therefore,how to solve the control problem in the presence of network resource constraints,which include the bandwidth and computing power of communication channels,is of great significance in both theoretical and practical applications.At the same time,the complex systems present the characteristics of large-scale,high-speed,complex and intelligent,for example,industrial production,automobile autopilot,spacecraft flight,smart grid and other fields.The non-linearity of the controlled system is stronger so that it is difficult to describe by accurate mathematical models.However,the traditional control methods can hardly be directly applied in these practical non-linear systems.Even thoughthere are some solutions,they can not perfectly compensate for the influence of uncertainties,strong coupling,dynamic mutation and other factors in the nonlinear systems.Intelligent control is the advanced stage of the development of control theory.It is mainly used to solve the nonlinear and complex control problems which are difficult to be solved by traditional control methods.Therefore,under the constraints of communication network bandwidth and other complex constraints,it is worthy studying to design event-triggered mechanism,improve the utilization of communication network,and handle various constrained problems for complex nonlinear systems.Under the framework of adaptive neural and fuzzy control,based on event-triggered strategy,this paper co-designs the adaptive neural/fuzzy controllers and event-triggered mechanism for nonlinear single-input single-output(SISO)systems,interconnected large-scale systems and multi-agent systems in the presence of input constraints,output constraints or full-state constraints,respectively.The main results are summarized as follows:1.For strict feedback nonlinear systems with multiple actuator faults and measurable states,based on event-triggered mechanism,the adaptive fuzzy control problem is studied.The considered fault modes contain the bias fault,partial loss of effectiveness and total loss of effectiveness.The unknown nonlinear functions are approximated by fuzzy logic systems.In addition,considering adaptive backstepping technology and event-triggered mechanism,the designed fault tolerant controller and adaptive laws are correlated with virtual control coefficients.Finally,the feasibility and effectiveness of the control scheme are verified by numerical and practical simulation results2.Based on the first Chapter,the problem of adaptive event-triggered fuzzy decentralized control is studied for non-strict feedback interconnected large scale systems with multiple actuator faults and unmeasurable states.The state observer is established to estimate the unmeasurable states by using the fuzzy logic systems,where nonlinear functions don't need to satisfy Lipschitz condition.In the framework of backstepping,the relative threshold-based event-triggered algorithm is constructed such that the updating frequency of controller is reduced,and the adaptive controller and adaptive laws with virtual control coefficients are designed.Based on the designed control strategy,all closed-loop signals can be ensured to be semi-globally uniformly ultimately bounded,and Zeno behavior can be avoided.Finally,the effectiveness of the proposed control strategy is verified by numerical and practical simulation results.3.Based on neural network,an adaptive dynamic surface output feedback control strategy is proposed for non-strict feedback nonlinear systems with output constraints,actuator faults and unmeasurable states.In the framework of adaptive fuzzy control and backstepping technique in the first Chapter and the second Chapter,the dynamic surface control method is employed to solve the problem of "complexity explosion" and the effect of faults is compensated by adaptive control method.At the same time,in order to avoid the waste of communication resources,a relative threshold-based event-riggered mechanism is constructed.In addition,a barrier Lyapunov function is used to solve the output constrained problem Finally,it is proved that all closed-loop signals in the system are semi-globally uniformly ultimately bounded by the stability theory of Lyapunov.4.For interconnected large-scale nonlinear systems with unknown hysteresis and full-state constraints,the problem of event-triggered adaptive neural control is solved.The characteristics of radial basis function are utilized to construct a state observer and solve the algebraic loop problem caused by nonstrict-feedback structure.To reduce communication burden and control signal transmission frequency,an adaptive event-triggered control strategy is proposed by using dynamic surface control technology and backstepping control method Then,by estimating the unknown coefficients in the hysteretic differential equation,the influence of unknown hysteresis is compensated.In addition,the barrier Lyapunov function is utilized to deal with full state constraints problem.Finally,all signals of the closed-loop system can be proved to be semi-globally ultimately uniformly bounded.5.For a class of non-strict feedback multi-agent systems with external disturbances,sensor faults and input saturation,the problem of adaptive cooperative control is addressed.The mean-value theorem and Nussbaum type function are applied to deal with the saturation term.By combining various sensor fault models,an adaptive neural network controller is designed to compensate for the effect of sensor faults.Based on neural network,the disturbance observer is designed to observe the complex disturbance signal and restrain the effect of disturbance.The event-triggered mechanism is considered in the controller design,which reduces the communication burden and adjusts the update frequency of the controller.Finally,some simulation results verify the effectiveness of the proposed method.6.Based on aforementioned Chapters,the proposed adaptive neural network control method is applied to a six-rotor nonlinear dynamic model with time-varying output constraints in the eighth Chapter,where the model is divided into the position and attitude system.Due to the limited hardware of UAVs,the event-triggered mechanism is used to control the update frequency of the controller.At the same time,the consistency tracking problem of multiple UAVs is studied.The stability of multiple UAVs with time-varying output constraints is guaranteed by numerical simulation,and the cooperative control problem with time-varying output constraints are addressed.Finally,simulation results show that the control strategy can ensure that UAVs operate stably in the confined area and external disturbances.
Keywords/Search Tags:Nonlinear systems, adaptive fuzzy control, neural network control, event-triggered mechanism, backstepping
PDF Full Text Request
Related items