| Ref.No: | 98003100 |
| Start date: | 12.04.2023 |
| End date: | 08.12.2023 |
| Approval date: | 12.04.2023 |
| Department: | CIVIL ENGINEERING |
| Sector: | STRUCTURAL ENGINEERING |
| Financier: | ΥΠΕΡΓΟΛΑΒΙΑ ΕΣΠΑ, KENTRO IKANOTHTWN GIA ENA ANTHEKTIKO KAI BIWSIMO DOMHMENO PERIBALLON ME TH XRHSH EKSYPNWN TEXNOLOGIW |
| Budget: | 45.632,00 € |
| Public key: | Ψ9ΗΡ46ΨΖΣ4-Τ9Φ |
| Scientific Responsible: | Assoc. Prof. FRAGKIADAKIS |
| Email: | mfragiadakis@gmail.com |
| Description: | THE MAIN OBJECTIVE IS TO DEVELOP TOOLS THAT CONCERN AN INNOVATIVE STRUCTURAL HEALTH MONITORING (SHM) COMPUTATIONAL FRAMEWORK THAT USES OPTIMIZATION ALGORITHMS, NEURAL NETWORKS TRAINED BY HIGH-FIDELITY FINITE ELEMENT MODELS AND MEASUREMENT DATA. THE FIRST DEVELOPED TOOL WILL BE USED TO IDENTIFY FAILURES AND PREDICT FUTURE DAMAGE IN STRUCTURES WITH THE AIM OF OPTIMIZING THE DESIGN OF NEW CONSTRUCTIONS, WHILE THE SECOND WILL BE USED TO MONITOR STRUCTURAL HEALTH AND PREDICT FUTURE DAMAGE IN STRUCTURES WITH THE AIM OF REDUCING COSTS MAINTENANCE AND THE SELECTION OF ACTIONS TO UPGRADE THEM. SPECIFICALLY, THE FOLLOWING TASKS WILL BE CARRIED OUT: 1. DEVELOPMENT OF THE COMPUTATIONAL FRAMEWORK FOR DAMAGE IDENTIFICATION 2. GENERATION OF SHM DATA FOR OPTIMAL FINITE ELEMENT SIMULATIONS WHICH INCLUDES THE DEVELOPMENT OF AN ALGORITHM FOR UPDATING THE FINITE ELEMENT MODELS, THE DEVELOPMENT OF AN ALGORITHM FOR THE GENERATION OF SHM DATA AND THE DEVELOPMENT OF A METHODOLOGY FOR USING MODELS WITH REDUCED DEGREES OF FREEDOM (REDUCED ORDER), 3. SHM TRAINING WITH ARTIFICIAL NEURAL NETWORKS WHICH INCLUDES THE DEVELOPMENT OF A CNN TRAINING METHODOLOGY BASED ON MACHINE LEARNING RULES AND OPTIMAL SENSOR PLACEMENT, 4. EXPERIMENTAL VALIDATION OF THE COMPUTATIONAL FRAMEWORK IN LABORATORY AND REAL STRUCTURES USING OSCILLATORY MEASUREMENTS AND MEASUREMENTS TO DETERMINE THE MECHANICAL CHARACTERISTICS OF THE STRUCTURES, 5. DEVELOPMENT OF CONFIGURATION PROCEDURES OF THE TWO COMPUTATIONAL TOOLS AND DESIGN OF A GRAPHICAL ENVIRONMENT FOR THE USE OF THE SOFTWARE. |
