Research On Robust Control Of Networked Systems With Communication Constraints

The rapid development of the computer technology and communication network technology,makes the control systems realize unprecedented stride.The structures of the control systems have been increasingly complex and the space distributions of control systems have become increasingly wider,so that the classic point-to-point control mode has failed to meet the development requirements of the control systems.Under this background,the information transmission mode between the system components has undergone a fundamental change.The traditional point-to-point control structure has shifted to a networked control structure,which has led to the emergence of networked control systems(NCSs).Networked control systems have many advantages such as network resource sharing,low cost,and strong system flexibility,which have been widely used in many fields.However,NCSs also bring the communication constraints that traditional control systems do not consider,such as network-induced delays,packet loss,and quantization.Although the above issues have achieved certain research results,there are still many problems to be solved,and then new challenges are raised for the existing control methods.Based on the above,this dissertation focuses on the robust control problem of the networked systems with packet loss and data quantization.The main research work is as follows:(1)A robust predictive controller is presented for networked uncertain systems with the control input quantization,due to the limitation of the shared network bandwidth.First,logarithmic quantizers are used to characterize the controller-actuator quantization and the sector bound approach is used to establish a networked uncertain system model with quantization.Secondly,based on this predictive control strategy,the sufficient conditions for stabilizing the networked system are obtained.Finally,the effectiveness of the proposed algorithm can be demonstrated by a numerical simulation.(2)For a networked distributed time-delay system with random loss and quantization,a novel robust H? output feedback controller based on state observer is proposed.Assume that the random packet loss satisfies the Bernoulli distribution,and the system quantization is described by the quantizer.Meanwhile,assume that the measured outputs of the timedelay system are first quantized,and then they can be transmitted over the shared network where packet loss can randomly occur.Based on the sector bound approach and the logarithmic quantizers,an augmented networked error system model can be established.Then,the sufficient conditions are deduced for ensuring the system stability and H? control performance,where the optimization problem of the robust H? control is solved by a cone complementary linearization approach.Finally,the stabilization effect of the designed robust controller is verified by a numerical simulation.