The National Science Foundation awarded a $649,800 Project Grant to the Georgia Tech Research Corporation from January 1, 2023 to December 31, 2025 under the Engineering program (CFDA 47.041). The grant will support research into developing convex relaxation and branch-and-bound algorithms to solve non-convex model updating problems in structural engineering simulations. Model updating aims to improve simulation accuracy by identifying parameter values that optimize agreement with measured structural response data. The proposed algorithms seek to guarantee finding globally optimal solutions for model updating, offering certifiable optimality. Multiple convex relaxations will be comparatively studied, including quadratic, semi-definite, and second-order cone programming approaches. The algorithms will be applied to deterministic and probabilistic modal property difference formulations in model updating. Validation will utilize sensor data from instrumented structures. Outcomes have potential to significantly advance simulation response accuracy for structures subjected to dynamic loads like earthquakes and impacts.
Generated 1/7/24, 10:18 AM