SimCenter personnel have just published a new article in a special issue of Earthquake Engineering and Structural Dynamics, entitled Surrogate modeling of structural seismic response using probabilistic learning on manifolds. The article examines the use of probabilistic learning on manifolds (PLoM) in seismic design and analysis of structures. The technique, originally developed by Christian Soize and Roger Ghanem, produces in a computationally efficient manner very effective surrogates appropriate for the probabilistic estimation of high dimensional stochastic systems that themselves are highly nonlinear. Download the paper from Wiley at https://doi.org/10.1002/eqe.3839.
The article was authored by Kuanshi Zhong, former SimCenter postdoctoral scholar and now professor at the University of Cincinnati; Sanjay Govindjee, SimCenter domain expert and former SimCenter PI and co-Director; Javi Gual Navarro, visiting student from the Polytechnic University of Catalunya; and Gregory G. Deierlein, SimCenter co-Director.
The PLoM algorithm in the paper is available for users in SimCenter tools EE-UQ and quoFEM. The SimCenter Live Expert Tips session How to Develop and Deploy Surrogate Models for Structural Response Prediction in EE-UQ is a resource for further understanding the available module.
Citation: Zhong K, Navarro JG, Govindjee S, Deierlein GG. Surrogate modeling of structural seismic response using probabilistic learning on manifolds. Earthquake Engng Struct Dyn. 2023;1-22. https://doi.org/10.1002/eqe.3839