The Research Of Railway Signal Cable Fault On-line Detection System Based On The STDR/SSTDR Method

The railway communication cable is the main channel of transmission railway signal. However, in recent years, signal cable aging, plus artificial ground excavation and construction and other factors result in increased probability of occurrence of signal cable fault. Therefore, the cable fault detection has attracted increasing attention. Using traditional off-line detection method, there are many drawbacks:in the repair process, it can create additional parking and power loss, besides, the efficiency is very low and the detection precision is not high. On the other hand, the intermittent arc fault of the cable, the maintenance of time is very short, generally it is unable to reproduce when the cable does not work, and the traditional offline detection methods is difficult to detect. This paper uses the STDR/SSTDR’s on-line detection methods for the continuous detection. The low test signal level and high noise immunity test method is very suitable for the intermittent fault of the cable online detection and location.Based on the analysis of the traditional cable fault detection method, the paper points out their respective advantages and disadvantages, and then focuses on spread-spectrum time-domain reflectometry, which is very suitable for on-line fault detection, namely STDR/SSTDR method. First, through theoretical analysis and research on STDR/SSTDR method, the system’s main key technology is discussed in brief. Then the detection system model is established. Next, from the whole system model, the paper will specifically analyse the realization of each module of the system and the parameters on the effect of performance.Around structure module of the STDR/SSTDR system, the author will start to deploy this paper. This system is mainly composed of a signal transmitting module, signal acquisition module, correlated modules and so on basic modules. In the signal transmission module, we need to find out the best launch sequence from a large number of pseudo-random sequences through MATLAB simulation and comparison. After sequence selection, from the point of view of SNR, the paper will theoretically analyse the sequence length and chip rate and so on factors for the influence on the SNR. Then compare the pros and cons of the test signal of SSTDR and STDR from the perspective of bandwidth. Because of the need for sine wave for modulation of SSTDR signal, finally, from the angle of related characteristics, compare different modulation frequency on the correlation’s effect. Correlated module uses software-related, in order to reduce the impact of noise on the signal, on the basis of a correlation algorithm,, the paper preliminarily uses the two correlation algorithm; when SNR is low, the peak of the two correlation will be broadened. Again, on the basis of the two correlation algorithm, Hilbert transform algorithm is proposed; In order to meet the needs of the real-time online detection, the correlation method adopts fast FFT algorithm to achieve. By computation’s analysis, the feasibility of the algorithm gets confirmed in mainstream digital signal processor.Finally, combining with the theoretical research and practical simulation, based on the above-mentioned study, this paper establishes MATLAB simulation model of STDR/SSTDR system. Specific parameters have been set for a specific simulation. The simulation results show that, when the SNR is not too low, STDR and SSTDR can accurately locate the position of fault; but in the case of low SNR, the peak of STDR’s correlation is very weak, and SSTDR can still be for fault detection, SSTDR has more advantages than STDR.