Model-Free Adaptive Control Of Networked Systems With Event-Triggered Mechanism

Along with the rapid development of networked communication technologies as well as the increasing scale of control systems,networked control systems have broken many limitations such as geographical area and resources.As such,networked control systems become exceedingly important due to the inherent merits including low cost,simple installation and resource sharing.However,transmitting real-time data via a shared network would increase the communication burden and thus,result in packet dropouts,which will degrade the performance of control systems or even make the underlying systems unstable in some cases.In order to save network bandwidth and reduce redundant data transmission,a communication strategy called event-triggered mechanism was proposed and attracted the attention of many researchers.Motivated by the above-mentioned engineering problems,this thesis focuses on the corresponding data-driven control problems for networked systems with event-triggered mechanisms.The main contents are given as follows:First,a novel event-triggered model-free adaptive control method for networked nonlinear systems is proposed to reduce the transmission burden of communication network.Different from the traditional model-free adaptive control method which calculates the control signal at fixed sampling instants,an event-triggered sampling method is proposed to calculate the new control signal only when the output data changes sufficiently large.As such,the proposed method can avoid the unnecessary waste of communication and computation resources.Furthermore,to achieve a satisfactory control performance subject to packet dropouts while improving the efficiency in resource utilization,an event-triggered model-free adaptive control method combined with the packet dropouts compensation is proposed.Sufficient conditions are established to guarantee that the tracking error is uniformly ultimately bounded.Then,the distributed model-free adaptive control method based on the event-triggered strategy for leader-following networked multiagent systems is proposed.In such a communication scheme,the signal transmission of each agent is only triggered when the certain triggering condition is satisfied.Furthermore,in terms of the distributed measurement outputs obtained by each agent's information interaction with neighboring agents,an effectively distributed data-driven controller is designed in order to achieve the prescribed performance requirement for multiagent systems with fixed communication topology.Finally,simulation examples are carried out to illustrate the effectiveness of the proposed control methods.The simulation results show that,compared with the traditional model-free adaptive control,the proposed methods have a strong ability to save communication resources and reduce the negative effects caused by packet dropouts.