Diagnosis of Multiple Wiring Faults Using Time-Domain Reflectometry and Teaching–Learning-Based Optimization

Abstract

Time-domain reflectometry has proven to be one of the best methods for wiring network diagnosis, and it can easily be applied to the detection and localization of defects, while only requiring one access point to the wiring network. In this article, a novel approach combining the time-domain reflectometry response extracted from vector network analyzer measurements and the teaching-learning-based optimization technique is developed and applied to the diagnosis of wiring networks. The proposed approach consists of two steps. In the first step, propagation along the cables is modeled using the forward model. In the second step, teaching-learning-based optimization is used to solve the inverse problem to deduce physical information about the defects in the wiring network. The proposed approach has been successfully tested on several cases and for different configurations. Comparisons of the proposed timedomain reflectometry/teaching-learning-based optimization approach results with measurements reveal that this approach has a high potential and is very effective for wiring network diagnosis.

Locating Multiple Soft Faults in Wire Networks Using an Alternative DORT Implementation

“…In fact, these are known through the healthy NUT model used for baselining. The fact that only the moduli of the τ i are required in (13) implies that errors in the positioning of the dummy fault have a minor impact on the accuracy of the estimated |Γ i |, as long as the NUT is tested at frequencies where propagation can be regarded as subject to weak dissipation.…”

“…In fact, soft faults are characterized by a very weak reflectivity, which produce echoes that can pass unnoticed compared to those caused, e.g., by junctions within a network under test (NUT) [3], [4], [11], [12]. A recent work [13] aiming at locating multiple faults falls in this same group, as it mainly works for relatively hard faults.…”

A Performance Analysis of Optimization Algorithms for Wiring Network Reconstruction and Diagnosis based on Reflectometry

“…In literature, many authors have been working on merging reflectometry with different optimization based techniques for the sake of EWIS mapping [13]- [15]. However, most relied on the prior knowledge of the network's topology and were confined to estimating the branch lengths.…”

Locating Faults With High Resolution Using Single-Frequency TR-MUSIC Processing

“…As a matter of fact, soft faults are usually characterized by weak reflectiveness which produce echoes that can pass unnoticed compared to those caused, e.g., by junctions within an NUT [7], [8], particularly when noisy conditions are present, e.g., in live testing [9]. Add to that, the detection of multiple faults is also mostly limited to hard faults [10].…”