Risk-based maintenance strategy for deteriorating bridges using a hybrid computational intelligence technique: a case study

Abstract

The current bridge inspection and maintenance protocol that is used in most countries focuses primarily on the visible aspects of bridge fitness and underestimates the invisible aspects, such as resistance to scouring and earthquake hazards. To help transportation authorities to better consider both aspects, the present study developed a new computational intelligence system, the so-called risk-based evaluation model for bridge life-cycle maintenance strategy (REMBMS). This model considers the three main risk factors of component deterioration, scouring and earthquakes in order to minimise the expected life-cycle cost of bridge maintenance. Monte Carlo simulation is used to estimate the probability of bridge maintenance. The evolutionary support vector machine inference model (ESIM) was applied to estimate the risk-related maintenance cost using historical data from the Taiwan Bridge Management System (TBMS) database. The time-influenced expected costs were obtained by multiplying each maintenance probability with its associated cost. Finally, the symbiotic organisms search (SOS) algorithm is used to identify the bridge maintenance schedule that optimises the life-cycle maintenance cost. The present study provides to bridge management authorities an effective approach for determining the optimal timing and budget for maintaining transportation bridges.

Keywords:

• Bridge maintenance strategy
• artificial intelligence
• optimisation
• bridge risk evaluation
• evolutionary support vector machine inference model
• symbiotic organisms search

Disclosure statement

No potential conflict of interest was reported by the authors.