DRF Lunchtime Clinic - Dr Jin Di

We look forward to introducing Dr Jin Di, visiting Fellow at University of Cambridge, and hearing his talk on Intelligent Evaluation of Cable-Stayed Bridge Service Performance.

Dr. Jin Di is the professor and the director of Intelligent Diagnosis and Rehabilitation Research Center of Civil Engineering Structures at Chongqing University, the director of the Engineering Structures Institute of The National Center for International Research of Low-carbon and Green Buildings, adjunct professor at Zhejiang University and Chang'an University. He is also the Chairman of Xi'an Changchang Bridge Engineering Technology Co., Ltd and Zhejiang Tianxi Civil Engineering Technology Co., Ltd. He has also held visiting posts at the Department of Civil and Environmental Engineering, Stanford University (2016-2017) and at the Construction Information Technology Laboratory at the Division of Civil Engineering of the Department of Engineering at the University of Cambridge.  (2025-now) as a Visiting Professor. His current research and scholarly interests focus on High-Performance Bridge Structure, Nondestructive testing and Structural Health Monitoring of Bridge Structures, Service Performance Evaluation and Improvement of Bridge Structures. 

Abstract

Intelligent Evaluation of Cable-Stayed Bridge Service Performance

There is a large stock of old bridges in the world, which faces durability problems and catastrophic risks, and has great service safety problems. Efficient and accurate diagnosis is the basis for ensuring the safety of service throughout their life cycle. However, traditional diagnosis techniques have bottlenecks such as "undiagnosed, inaccurate, and untimely". 

Aiming at the "inaccuracy" bottleneck problem, the theoretical research, numerical analysis, model test and artificial intelligence methodologies were adopted to implement basic scientific research related to multi-dimensional characteristic indicators of structural service performance and intelligent assessment theories and methodologies based on the data of a Cable-stayed bridge. A method to improve the quality of multi-source heterogeneous big data, and to clarify the coupling relationship of multi-dimensional characteristic indicators was proposed. Then, the mapping mechanism of structural service safety performance and indicators was established, and the mapping mechanism was analyzed from the perspective of physical mechanics, so as a multi-scale and multi-dimensional characteristic index system was built. Finally, a structural service safety performance assessment methodology based on physical reinforcement learning algorithm was established. 

A new path to assess the structural safety performance in service that integrates artificial intelligence and physical mechanics was explored, and a "data-physics" dual-driven quantitative assessment theory and methodology for multi-scale structural service performance in an efficient, precise and intelligent way was preliminarily established. The research achievements have important scientific significance and application value for the intelligent diagnosis of the service safety of major infrastructures