Structural Assessment System for Damage and Degradation: Two-Stage Damage Identification Based on Neural Networks and Improved MDLAC Method

A “smart” structure has many functions, including monitoring, repairing, shape formation, and learning. Recently, interest in applying a monitoring system to structures for quality assurance and for evaluating seismic risk has been strong. Monitoring system is useful to diagnose the structural condition, and detect structural damage and degradation. In this study, we developed a monitoring system to assess the structural integrity. This system includes a diagnostic system for structural damage and degradation based on neural networks and improved MDLAC method, say, to detect the damage sites globally by applying neural networks and then to narrow the damage sites by using improved MDLAC method. To validate this system, we then use the 5-story structure in which the beams are fixed at both ends in order to confirm the performance of our proposal damage detection methods. As a result, it is pointed out that there are some possibilities to confirm the diagnostic system by utilizing these two methods.