A Delay Time Model With Multiple Defect Types and Multiple Inspection Methods

Mohamad Mahmoudi

doi:10.1109/TR.2017.2754519

2017-12-01

PRE

Mohamad Mahmoudi

Alaa Elwany *

Kamran Shahanaghi

Mohammad Reza Gholamian

DOI：10.1109/TR.2017.2754519

原文链接

原文求助

摘要

En 中文

We develop delay time models to determine optimal inspection policies for deteriorating infrastructures. We consider the case where complex infrastructures can fail due to different causes (defects) originating from various environmental or operational conditions, and capture this through modeling the arrival of different types of defects as nonhomogeneous Poisson processes with distinct rates of occurrence of defects. Additionally, we assume at each inspection epoch, there are multiple inspection methods available for use from which one is to be selected for use at that particular epoch. The key contribution that distinguishes our proposed models from previous works on delay time modeling is simultaneously considering multiple defect types and multiple inspection methods. Two mixed-integer nonlinear programming models are introduced to address the problem described above. The first model focuses on determining the optimal inspection policy that maximizes the reliability of the system over its useful life subject to a minimal threshold value of this reliability term. The second model determines the optimal policy thatminimizes the system downtime. The two models are solved using a branch-and-cut global optimization approach. Two separate numerical studies are conducted to demonstrate the performance of the models and validate it through benchmarking these results against a prior study in the literature.

Keyword:

Delay time model

infrastructure

maintenance optimization

mixed-integer nonlinear programming (MINLP)

multiple defect types

multiple inspection methods