| Cables are an indispensable component in modern electrical equipment,which takes on the function of power supply and signal transmission.Cable is a key factor affecting the reliable operation of the equipment.Studies have shown that cable hard faults are often developed by soft faults.So,if the soft fault can be detected before the hard fault such as short circuit or open circuit occurs,the operation reliability of the equipment can be effectively improved.In practical applications,the cable is dominated by a network structure,but there is no mature technique to implement the detection and location of the cable network soft fault.Therefore,this paper proposed a soft fault detection method for cable network based on time reversal multiple signal classification(TR-MUSIC).This method is mainly used to study soft fault detection and location of single cable and Yshaped cable network.The specific work of this paper is as follows:Firstly,based on the transmission line theory,the broadband electrical parameters of SYV-50-3 coaxial cable is analyzed.Combined analysis results with 3D electromagnetic simulation results,the effects of extrusion type soft fault,cutting type soft fault and thermal aging on the electrical parameters of coaxial cable are studied.At the same time,the simulation analysis of cable network is carried out to clarify the causes of node reflection phenomenon.According to the basic principle of cable fault detection,the parameters that can be used for fault detection are selected.Secondly,based on the focusing characteristics of time reversal,TR-MUSIC technology is used to implement cable soft fault detection.The soft fault detection algorithm is derived for single cable and Y-shaped network respectively,and the control group is introduced to reduce the adverse impact of node reflection.In order to get the length of the cable and used it to calculate the Green function,Time Domain Reflectometry based S-parameters was proposed.The effectiveness of the proposed soft fault detection method is verified by simulation and experiments.Thirdly,quantitative analysis of the performance of the proposed method is completed.Kurtosis analysis is combined with TR-MUSIC to enhance fault features.Taking SYV-50-3 coaxial cable as the experimental object,the soft fault detection experiments of single cable and cable network under different fault types and fault degrees are carried out.In addition,a twisted pair cable soft fault detection experiment is designed to verify the feasibility of TR-MUSIC for other types of cable soft fault detection.Finally,based on C # programming language,the software of coaxial cable network soft fault detection system is designed and implemented.The software’s function is verified and tested by experiments.In this paper,a soft fault detection method of coaxial cable network based on TRMUSIC technology is proposed,which can implement soft fault fast detection and location in single coaxial cable and Y-shaped coaxial cable network.The soft fault detection result of this method has high sensitivity and precision,and it is also feasible for other types of cable soft fault detection.This method has engineering application prospects. |
