On the Diagnosis of Incipient Faults in Transmission Lines Using a Projection Approach Based on Phase Patterns

Throughout the last decade, the problem of soft fault detection in transmission lines have been overflown with contemporary powerful technologies. However, a vast majority of state-of-the-art techniques including the well-known reflectometry methods, require bandwidths in the order of hundreds of megahertz for providing spatial resolution in the millimeter range of faults' locations. On the other hand, an emerging technique based on the analysis of multi-port transmission and reflection parameters, often referred to as the Time-reversal multiple signal classification (TR-MUSIC) ensured location accuracy and sub-millimeter resolution of multiple soft faults in complex wire networks. More importantly, this was made possible using continuous wave excitations even at low frequencies. However, as any other existing method, it suffered from the problem of attenuation inherent to transmission lines. In this paper, we will introduce a method based on Green function phase pattern analysis of tested cables relying only on reflection parameters, with no need for acquiring the transmission ones. In effect, testing long cables as in the case of power grids becomes possible. The proposed technique is shown to be robust against attenuation. Besides, it appears to be readily adapted for lively monitoring transmission lines, thanks to its continuous wave excitation abilities. The technique is shown to operate frequency by frequency, which allows the choice of specified frequency samples if distortion is present. Moreover, the proposed processing is shown to reinstate precise super-resolved estimates of soft fault locations. Experimental results based on coaxial cable implementation is provided to validate the method's feasibility.