Imperfect inspection of a system with unrevealed failure and an unrevealed defective state

Cavalcante, C, Scarf, PA ORCID: https://orcid.org/0000-0001-5623-906X and Berrade, MD 2019, 'Imperfect inspection of a system with unrevealed failure and an unrevealed defective state' , IEEE Transactions on Reliability .

[img]
Preview
PDF - Published Version
Available under License Creative Commons Attribution.

Download (1MB) | Preview
[img] PDF - Accepted Version
Restricted to Repository staff only

Download (1MB) | Request a copy

Abstract

This paper proposes a model of inspection of a protection system in which the inspection outcome provides imperfect information of the state of the system. The system itself is required to operate on demand typically in emergency situations. The purpose of inspection is to determine the functional state of the system and consequently whether the system requires replacement. The system state is modeled using the delay time concept in which the failed state is preceded by a defective state. Imperfect inspection is quantified by a set of probabilities that relate the system state to the outcome of the inspection. The paper studies the effect of these probabilities on the efficacy of inspection. The analysis indicates that preventive replacement mitigates low quality inspection and that inspection is cost-effective provided the imperfect-inspection probabilities are not too large. Some derivative policies in which replacement is “postponed” following a positive inspection are also studied. An isolation valve in a utility network motivates the modeling.

Item Type: Article
Schools: Schools > Salford Business School
Journal or Publication Title: IEEE Transactions on Reliability
Publisher: IEEE
ISSN: 0018-9529
Related URLs:
Depositing User: Dr Philip Scarf
Date Deposited: 24 Jan 2019 14:30
Last Modified: 17 Apr 2019 08:00
URI: http://usir.salford.ac.uk/id/eprint/49849

Actions (login required)

Edit record (repository staff only) Edit record (repository staff only)

Downloads

Downloads per month over past year