Study On Sampled-Data Control Of Networked Systems Based On Event-Triggering Mechanism

The development of computer technology,network communication technology and control theory has promoted the research and development of networked control systems as well as the sampled-data control.For a practical control system,on the one hand,it is not only uncertain and time-varying,but also characterized by complex structure and discrete behavior.On the other hand,if the sampling period is small,it will aggravate the network load,and the system will cause network congestion and waste network resources in a fixed sampling period.At the same time,the control strategy based on event-triggering mechanism can realize the user requirements of system measurement on-demand transmission,and it provides a new control design condition for the non-periodic sampled-data control of networked systems.According to this,the design of appropriate sampling period in complex environment can not only improve the performance of the controlled system,but also realize the efficient utilization of network information transmission channel bandwidth.Therefore,the study of the sampled-data control of networked systems based on event-triggering mechanism has practical significance in improving system robustness and reducing network communication burden.The main contents of this thesis are summarized as follows:1.Input delay control based on full-state and partial-state for common linear networked systems is discussed.Considering the common linear networked systems and the existing network-induced delays and input delays,two full state feedback and partial state feedback controllers based on aperiodic sampling are established.Considering the measurable and partially measurable state of the actual system,a full-state feedback sampling controller based on input delays and a static output feedback control design are further designed.The matrix equality technique is used to design the parameters of the controller.2.A general form of aperiodic sampling fuzzy controller is designed for continuous-time nonlinear networked systems.The nonlinear control system of the pendulum is described by using the fuzzy models.Considering the disturbance in the actual process of actuator,the solution method and H_?performance of the controller satisfying H_?performance index are designed based on H_?control theory.A kind of state feedback variable sampling based on fuzzy rules is designed.The designed controller makes the double inverted pendulum system satisfy H_?interference suppression performance.3.The reference output tracking control of discrete-time linear systems based on event-driven mechanism is studied.The delta operator method is used to obtain discrete-time linear networked systems,and discrete event triggers are used to design event detection cycles of event triggers.Considering the delay of measurable variables from sensors to event triggers,the event-driven control system is transformed into a class of time-delay systems.The stability of the system is analyzed by input-output method.Based on the analysis results,a new reference output tracking controller is designed.4.Event-triggered reference tracking control for discrete-time non-linear networked systems is further studied.The network delay between sensor and control unit is considered on the basis of the designed event-triggering conditions.A full-state event trigger tracking controller based on event trigger is established.The reference model is a simple linear system model,and the time delay is obtained.The stability of the system is analyzed by the utilization of Lyapunov stability theory and small gain theorem.A reference output tracking controller based on event triggering is designed by H_?control technology.Finally,the simulation results show that the designed aperiodic sampling controller can achieve the control goal with less network resources.