A Stochastic Optimization Model for Infrastructure Reliability Assessment Under Environmental Uncertainty

Abstract

Environmental uncertainties, degradations, and hazards have an impact on the behavior of infrastructure systems; hence, they pose major challenges in the process of reliability assessment and sustainable maintenance management. The existing deterministic methods tend to miss out on the stochastic nature of environmental loads and behaviors of the infrastructure system, leading to inefficient risk assessment and improper use of resources. This paper presents a stochastic optimization model for infrastructure reliability assessment under environmental uncertainties. In this regard, the model incorporates probabilistic hazard modeling, reliability-based optimization, and uncertain decision making into the assessment of infrastructure reliability. The infrastructure system reliability is formulated using a stochastic optimization model, which involves environmental loads, system degradations, and maintenance actions. This proposed model has been applied to a multi-component infrastructure test problem that is vulnerable to correlated environmental risks and has proven its efficiency in reliability evaluation and reduction of unnecessary cost investment. It can be seen from the results that the model ensures system reliability of about 0.949 at optimized cost of $591.3k and performs better than deterministic equivalents since it provides higher reliability values at a smaller cost. These results have proved that uncertainty-based optimization is more efficient in terms of resilience planning, risk management, and decision-making regarding infrastructure.