Sensor Selection for Distributed Reflectometry-based Soft Fault Detection using Principal Component Analysis - Archive ouverte HAL Access content directly
Conference Papers Year :

Sensor Selection for Distributed Reflectometry-based Soft Fault Detection using Principal Component Analysis

(1, 2) , (1) , (1) , (2) , (3)
1
2
3

Abstract

In this paper, a new approach is introduced for the selection of the most relevant sensors to monitor and diagnose soft faults in complex wired networks. Although reflectometry offers good results in point to point topology networks, it introduces ambiguity related to fault location in more complex wired networks. As a solution, distributed reflectometry method is used. However, several challenges are imposed, from the computing complexities and sensor fusion problems, to the energy consumption. In this context, the proposed method combines Time Domain Reflectometry (TDR) method with Principal Component Analysis (PCA) approach. To do so, a distributed reflectometry approach is considered for a CAN BUS network where sensors perform their reflectometry measurements consecutively. The simulated TDR responses are then arranged into a database. With this database, a PCA model is developed and used to detect the existing soft faults. Coupled with statistical analysis based on Hotellings T 2 and squared prediction error, the most relevant sensors to monitor and diagnose the soft faults present in the network are highlighted with a high accuracy.
Fichier principal
Vignette du fichier
AUTOTESTCON2019.pdf (285.91 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-02295381 , version 1 (12-03-2020)

Identifiers

  • HAL Id : hal-02295381 , version 1

Cite

Nour Taki, Wafa Ben Hassen, Nicolas Ravot, Claude Delpha, Demba Diallo. Sensor Selection for Distributed Reflectometry-based Soft Fault Detection using Principal Component Analysis. IEEE AUTOTESTCON 2019, Aug 2019, Maryland, United States. ⟨hal-02295381⟩
131 View
49 Download

Share

Gmail Facebook Twitter LinkedIn More