An exponential chaotic differential evolution algorithm for optimizing bridge maintenance plans

doi:10.1016/j.autcon.2021.104107

Return

An exponential chaotic differential evolution algorithm for optimizing bridge maintenance plans

2022-02-01

Eslam Mohammed Abdelkader *

Osama Moselhi

Mohamed Marzouk

Tarek Zayed

DOI：10.1016/j.autcon.2021.104107

Original

Save

View PDF

Abstract

En 中文

Bridges are one of the fundamental infrastructure assets that are vital for economic growth and public welfare. Over the past few decades, the numbers of deteriorating bridges have drastically escalated raising concerns for serviceable, safe and functional transportation networks. This state of affairs poses a paramount challenge especially when coupled with the need to address social and environmental constraints. Accordingly, this current research paper proposes an automated three-component model for bridge maintenance optimization at both project and network levels. The first component aims at identifying the physical characteristics of the tackled bridge inventory. The second component encompasses designing a multi-objective optimization model to determine the optimal set of maintenance plans through four principal objective functions. These functions comprise maximization of performance condition of bridge elements, minimization of agency and user costs, minimization of duration of traffic disruption and minimization of environmental impact. In the multi-objective optimization model, an exponential chaotic differential evolution (ECDE) algorithm is introduced in an attempt to circumvent the drawbacks of convergence speed and search behavior of classical meta-heuristics. The third component combines criteria importance through inter-criteria correlation (CRITIC), complex proportional assessment (COPRAS) and grey relational analysis (GRA) to select the most optimum maintenance plan for each study period. Comparison results revealed that ECDE-based Sinusoidal algorithm managed to improve the performance diagnostics of classical meta-heuristics by values ranged from 49.2% to 73.1% over the multi-year maintenance plans. The results of benchmark test functions exemplified that ECDE-based Sinusoidal algorithm performed better than genetic and differential evolution algorithms by 114.2% and 79.5%, respectively. The developed integrated model is expected to assist infrastructure managers in executing optimized and sustainable maintenance budget plans within various planning scenarios.

Keywords:

Bridges

Maintenance optimization

Project and network levels

Multi-objective

Exponential chaotic differential evolution

Multi-criteria decision making

Complex proportional assessment

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.