Sampled-data And Event-triggered Control Of Stochastic Systems Subject To Incomplete Transmission Of Network Information

With the remarkable improvement of computer's performance and reliability,sampled-data control system,the application of discrete digital controller to control continuous controlled object,has been widely studied and applied in modern control theory.Tradi-tionally,the sampling interval of sampled-data control system is assumed to be a constant.However,the assumption is often invalid in some practical systems.For example,in net-worked or embedded control systems,unpredictable network-induced phenomena often lead to uncertainty in sampling intervals.In addition,due to the limitation of commu-nication network resources,sampling signals or control commands may suffer from some inherent imperfection of communication network in the transmission process,such as delays,packet dropouts and disorder,which seriously affect the performance and stabil-ity of considered systems.Therefore,the modeling and control of periodic or aperiodic sampled-data systems subject to communication constraints has become one of the re-search hotspots.In the way of modeling,the stability criterion with lower conservativeness can be obtained by using the inherent characteristics of communication networks to model the controlled systems.For example,the probabilistic characteristics of stochastic sam-pling,the probabilistic characteristics of network-induced delays,and the non-uniform distribution characteristics of delays caused by successive packet dropouts,etc.In the way of control,using event-triggered mechanism to control considered systems can effec-tively save network resources on the premise that stability or related performance can be guaranteed.Therefore,under the guidance of these two main principles,the thesis mainly studies the sampled-data control and event-triggered control of stochastic systems with communication constraints.Based on periodic and aperiodic sampling,the problem of modeling packet dropouts with intrinsic characteristics of successive packet dropouts is studied.Based on event-triggered mechanism,the problem of event-triggered control for discrete-time stochastic systems with packet dropouts is studied.The research contents of the thesis are as follows:1.Based on periodic sampling,the modeling of successive packet dropouts and robust H_?control of stochastic systems are studied.In order to explore the inherent characteristics of successive packet dropouts,the considered networked control system with uncertain parameters is first transformed into a stochastic system with stochastic delay,in which the randomness of the delay comes from successive packet dropouts,and the non-uniform distribution characteristics of the delay can be verified by the proposed interval probability,which depends on packet loss rate and the upper bound of successive packet dropouts.Based on this,robustly exponentially mean-square stability of the system with an H_?performance is guaranteed,an H_?controller design procedure is then proposed.Finally,a simulation example is given to verify the effectiveness of the proposed method.2.Based on stochastic sampling,the modeling of successive packet dropouts and control of stochastic networked systems are studied.By input delay approach,the in-put delay induced by successive packet dropouts is subject to double randomness.The randomness of the input delay comes from not only successive packet dropouts,but also stochastic sampling.In order to explore the intrinsic characteristics of packet dropouts and stochastic sampling,the probability distribution values of the stochastic delay taking values in two given intervals can be explicitly obtained by using the total probabili-ty formula.Based on existing methods and the probability distribution values,another model which takes full advantage of probability distribution characteristics of both packet dropouts and sampling periods is then established.Based on the new model,the stability theorem with lower conservativeness and the controller design procedure are given,re-spectively.Finally,a numerical simulation example is exploited to show the effectiveness and applicability of the results derived and some less conservative results are obtained.3.Based on multiple stochastic sampling,the modeling and stabilization problems of networked control systems under simultaneous consideration of bounded packet dropouts and occasionally missing control inputs are studied.Assuming that the sampling errors of every sampling period have an upper bound and the upper bound satisfies some con-dition,two probability distribution values to characterize the non-uniform distribution characteristics of successive packet dropouts is first obtained by the total probability for-mula.By introducing new stochastic variables and using switched system method,a new stochastic model with switched delay is established.Second,some sufficient conditions in terms of linear matrix inequalities for the existence of desired stabilizing controller are derived by using the average dwell time method.Finally,the effectiveness of controller and the lower conservativeness of proposed method are verified by a numerical simulation example.4.Based on discrete-time systems,the intrinsic characteristics of packet dropouts and synchronization of stochastic complex dynamical networks are studied.First,an error dynamical network with stochastic and bounded delay is established by step-delay method.By the probability theory method,the accurate probability to characterize the non-uniform distribution characteristics of discrete time-delay is obtained.By introducing two new mapping functions and a new random variable,a new model is established to reflect the probability characteristic of consecutive packet dropouts.Based on the proposed model,some sufficient conditions are proposed under which the error dynamical network is globally exponentially synchronized in the mean square sense.Subsequently,a probability-distribution-dependent controller design procedure is then proposed.Finally,two examples are given to demonstrate the effectiveness of the controller and the lower conservativeness of the proposed method,respectively.5.Based on event-triggered mechanism,the modeling and control problems of dis-crete stochastic systems with packet dropouts are studied.Considering the difference between event-triggered mechanism and time-triggered mechanism,stochastic sequences satisfying Bernoulli distribution and not satisfying Bernoulli distribution are used to model the packet dropouts in triggered packets in the transmission networks,respec-tively.Then,the mean-square exponential stability of resulting augmented system is guaranteed and the prescribed H_?performance level is achieved by solving resulting dis-crete P-problem.Finally,two examples are given to illustrate the effectiveness of the proposed method.