Font Size: a A A

Development Of Fault Diagnosis System For Nonlinear Analog Circuit Based On Multi Port Network

Posted on:2022-10-02Degree:MasterType:Thesis
Country:ChinaCandidate:J B YeFull Text:PDF
GTID:2518306314981059Subject:Instrumentation engineering
Abstract/Summary:PDF Full Text Request
Due to the manufacturing process and other reasons,the vast majority of electronic components have nonlinear characteristics.The signal continuity,component parameter tolerance,nonlinear phenomena and complex fault characteristics of nonlinear analog circuits lead to great difficulties in testing and diagnosing nonlinear analog circuits.In the nonlinear analog circuit,more achievements have been made in the diagnosis of single input and single output system,while the research on the diagnosis of multiple input and multiple output system is still in the theoretical stage,which is not systematic and perfect in practical application.In this paper,the multi-port network is a multi-input and multi-output nonlinear analog circuit.Based on the existing Volterra kernel diagnosis theory,this paper discusses the Volterra functional series model of multi-port network nonlinear analog circuit,and the intelligent diagnosis principle based on Volterra kernel and neural network.The optimal search theory is used to search and optimize the multi tone test excitation signal.By selecting the appropriate frequency base to apply the multi tone excitation signal to each input of the circuit to be diagnosed,the output frequency components of the same order Volterra core do not coincide with each other.According to the frequency domain output characteristics of the Volterra core under multi tone excitation,the Volterra core sequence is obtained.In the process of Volterra kernel processing in nonlinear analog circuit diagnosis,the circuit fault feature selection is studied The improved annealing genetic kernel feature extraction algorithm is used to extract fault features,which can achieve faster feature extraction speed and global optimization ability,and improve the efficiency and accuracy of fault diagnosis of multi port network nonlinear analog circuit.Finally,the software and hardware design of the system is completed,and the multi port nonlinear analog circuit fault test and diagnosis system with signal source acquisition unit,multi tone signal generation unit and data acquisition unit is developed.The specific circuit is diagnosed,and the effectiveness and practicability of the system are verified by the actual circuit fault diagnosis.
Keywords/Search Tags:multiport network, nonlinear analog circuit, Volterra kernel, improved annealing genetic algorithm, feature selection and extraction
PDF Full Text Request
Related items