Intelligent Framework for Finite Element Analysis with Machine Learning and Back-Analysis Capabilities for Geotechnical Engineering
摘要
To enhance efficiency and optimization in design processes within geotechnical engineering practice, this paper introduces a comprehensive framework that harnesses cutting-edge technology. The objective is to shed light on the intelligent setup of finite element analyses through scripting, integrating machine learning capabilities to augment the execution of geotechnical engineering projects. The study leverages the powerful finite element package, Strand7, in conjunction with the versatile machine learning package, TensorFlow. The framework was applied successfully to a deep excavation case study. Challenges encountered during the automation process to satisfy the objective function were resolved by adopting a sequential approach.