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 time-domain 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.