Table of Content

Open Access iconOpen Access

ARTICLE

Sensor Fault Detection in Large Sensor Networks using PCA with a Multi-level Search Algorithm

A. Rama Mohan Rao1, S. Krishna Kumar1, K. Lakshmi1

CSIR-Structural Engineering Research Centre, CSIR Campus, Taramani, Chennai

Structural Durability & Health Monitoring 2012, 8(3), 271-294. https://doi.org/10.32604/sdhm.2012.008.271

Abstract

Current advancements in structural health monitoring, sensor and sensor network technologies have encouraged using large number of sensor networks in monitoring spatially large civil structures like bridges. Large amount of spatial information obtained from these sensor networks will enhance the reliability in truly assessing the state of the health of the structure. However, if sensors go faulty during operation, the feature extraction techniques embedded into SHM scheme may lead to an erroneous conclusion and often end up with false alarms. Hence it is highly desirable to robustly detect the faulty sensors, isolate and correct the data, if the data at faulty sensor locations are sensitive. Several sensor fault detection algorithms have been reported in the literature and among them the PCA based sensor fault detection algorithm [Kerschen et al (2005)] appears to be robust in isolating all types of sensor faults. However, in a large sensor network, the computational time in isolating the faulty sensor is prohibitive for online fault detection and isolation. In this paper we propose a multi-level search algorithm, which improves the performance of the PCA based sensor fault detection algorithm quite appreciably. Numerical simulation studies have been carried out to demonstrate the effectiveness of the proposed algorithm. The sensitivities of the proposed algorithm with parameter settings are also presented.

Keywords


Cite This Article

Rama, A., Kumar, S. K., Lakshmi, K. (2012). Sensor Fault Detection in Large Sensor Networks using PCA with a Multi-level Search Algorithm. Structural Durability & Health Monitoring, 8(3), 271–294.

Citations




cc This work is licensed under a Creative Commons Attribution 4.0 International License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
  • 1716

    View

  • 1431

    Download

  • 0

    Like

Share Link