Fault Diagnosis Of Nonlinear Switching System In Finite Network Environment

In the field of industrial production and life,nonlinear switching systems are widely spread.The performance of the switching systems is often determined by the switching subsystems and the switching rules.At present,the research on switching systems is still very popular.Nowadays,research on the stability of switching systems has been very mature,mainly including the methods of dwell time technology and Lyapunov function technology.This paper aims to design a fault detection device for such a sector nonlinear switch-ing system to complete the system fault diagnosis.The network is used for data transmis-sion between the fault detection device and the detected nonlinear switching system.Due to the limited network resources,some factors in network transmission such as data pack-et loss,quantization error,and network delay may exist,which may interfere with the performance of the fault diagnosis device.Therefore,this paper need to design a fault detection device in the presence of network inducing factors,which can accurately detect the occurrence of system faults.Based on the above problems,this paper selects the Kalman filter to design the fault detection link.For the case where there is a network quantization error alone,this paper first models it,and chooses to model the logarithmic quantizer.Then this paper models the nonlinear switching system,propose the Kalman filter model,propose the fault weight-ing model,construct the residual system,and process the residual signal generated by the residual system,so that the residual signal is robust to external interference and sensi-tive to faults.This paper selects the appropriate Lyapunov function to make the residual system satisfy the stability conditions and have the H_? performance,so as to obtain the corresponding filter parameters.On the basis of the existence of quantization error alone,this paper introduces the da-ta packet loss problem and time-delay problem respectively.And abstract the data packet loss phenomenon into the Bernoulli model.Here,this paper still uses the sector nonlin-ear switching system model and the Kalman filter model,and then construct the residual system and process it.The linearized inequality group is obtained by means of Lyapunov function method and linear matrix inequality technology,and the corresponding filter parameters are obtained with the help of toolbox.Finally,through the simulation experi-ments,the corresponding curves are obtained.