Study On Dynamic Quantized Control Of Networked Systems With Actuator Fault

In an Industry 4.0 era,cyber-physical systems will have a disruptive impact on manufacturing companies.The fusion of the physical and the virtual world requires connections between physical objects and units of computation.That is where networked control systems come into play.Compared with traditional point-to-point feedback control systems,the introduction of communication networks between distributed system components brings new challenges.Firstly,when communication resources are in short supply,the quality of control might be degraded due to the heavy communication burden.Until now,there are two promising ways to save communication consumption: data quantization and event-triggered mechanism.It is interesting and important to investigate the way to combine above methods and bring their potentialities into full play.Secondly,affected by aging and disturbances,the reliability of the actuators cannot always be guaranteed.The research on accommodating actuator fault and maintaining acceptable system performance has attracted much attention.In this study,the author mainly focuses on the dynamic quantized control design problem of networked control systems(NCSs)with event-triggered mechanism and actuator fault.The main contributions can be summarized as follows:1.Assuming that a finite-level dynamic uniform quantizer is used and the controller is event-triggered,asymptotic stability analysis and feedback controller design of NCSs with constant sensor fault and actuator fault are accomplished.2.Assuming that a finite-level dynamic uniform quantizer is used and the actuator is event-triggered,asymptotic stability of NCSs with parameter uncertainty and stochastic actuator fault is analyzed,and a robust feedback controller is designed.3.When the actuator is subject to deterministic fault and time delay occurs in the network between the sensor and the controller,stability analysis and stabilization of the system with finite-level dynamic logarithmic quantization are accomplished.4.Suppose the actuator is subject to stochastic fault and there is packet dropout in the network between the sensor and the controller,asymptotic stability of the system with finite-level dynamic logarithmic quantization is analyzed,and a mean-square feedback controller is designed.