Structural system reliability, reloaded

Abstract

Over the last four decades, structural system reliability (SSR) has been an active research topic as engineering systems including structures and infras tructure networks become more complex, and the computational power has been remarkably advanced. Among many efforts to advance SSR technologies, the research by Prof. Armen Der Kiureghian in the early 2000s is considered to have built critical pathways to the development of next-generation SSR methods by revisiting the topic of SSR. As demonstrated in Der Kiureghian (Structural system reliability, revisited. Proceedings, 3rd ASRANet International Colloquium, Glasgow, UK, July 10-12, 2006), his research efforts led to new insights and perspectives on critical topics of structural system reliability. This has encouraged researchers to perform a variety of research activities to develop new SSR technologies that can address challenges and needs in risk management of real-world systems. This chapter provides a review on such efforts to reload the research community with SSR technologies, which were made possible by the revisit. The chapter reviews the following SSR methods: linear programming bounds, matrix-based system reliability method, sequential compounding method, cross-entropy-based adaptive importance sampling, selective recursive decomposition algorithm, branch-and-bound employing system reliability bounds, and genetic-algorithm-based selective search for dominant system failure modes. For each of the reviewed methods, their main concepts, applications and follow-up research activities are presented along with discussions on merits, remaining challenges, and future research needs.