Physics-based and AI-informed design and evaluation of infrastructure subjected to multiple hazards (Passed)

Abstract

Civil infrastructure systems underpin a broad range of services that are foundational to the well- being of a community. Given their overall importance to social and economic development, it is essential to introduce more advanced design and evaluation solutions to ensure that these infrastructure systems are sustainable as well as resilient to natural or man-made hazards. In the past, these tasks were primarily accomplished by using extensive physics-based simulations such as finite or discrete element analysis. However, recent advancements in artificial intelligence (AI) combined with breakthroughs in computing techniques have made it feasible to propose alternative design and evaluation approaches. This research meticulously combines automation, physics- based simulation, and artificial intelligence to formulate the novel design and evaluation methods that allow stakeholders to target multiple performance objectives in a computationally efficient manner. An automatic control paradigm for performing structural design is introduced and combined with extensive physics-based simulations to create a large representative database. By integrating mechanics-based and machine learning methods, a set of data-driven and hybrid models are developed to characterize hazard-induced impacts on infrastructure. Finally, different models are comparatively assessed, and their applications are discussed.