讲座题目:Efficient Site Characterisation and Investigationthrough Knowledge-Informed AI
讲座时间:2026.11.17 (周二) 10:30-11:30
讲座地点:华南理工大学广州国际校区G2-101
主讲人:Assist. Prof. Pin Zhang
主讲人简介:
Dr Pin Zhang is an Assistant Professor at the National University of Singapore. His research focuses on resilient and intelligent geotechnics, climate-affected geomaterials, and AI-enabled solutions for geotechnical infrastructure. His international honours include the ICE Geotechnique 75th Anniversary Early Career Award, ISSMGE Bright Spark Lecture Award, FOMLIG Council Lecturer Award and Royal Society Newton International Fellowship. He has been listed among Stanford University's World's Top 2% Scientists since 2022 and serves on editorial boards of leading geotechnical journals.
讲座摘要:
Site investigation and characterisation are essential to geotechnical engineering but remain costly and uncertain because of subsurface variability and limited site-specific data. This study presents the Knowledge-Informed Data-driven Efficient (KIDE) framework, integrating knowledge-constrained neural networks, uncertainty quantification and active learning to improve site investigation and ground modelling. KIDE develops soil correlations from existing geotechnical databases for rapid ground assessment and optimises CPT/borehole locations by targeting areas where new measurements most effectively reduce prediction uncertainty. By progressively updating the ground model, KIDE improves characterisation accuracy, balances investigation effort and supports risk-informed geotechnical decision-making.


