Performance Evaluation of an Entropy-based Structural Health Monitoring by Utilizing Composite Cross-sample Entropy

Performance Evaluation of an Entropy-based Structural Health Monitoring by Utilizing Composite Cross-sample Entropy

This study aimed to solve the issue of unstable entropy values when multiscale cross-sample entropy (MSCE) was employed to assess the damage of real structure. Therefore, composite multiscale cross-sample entropy (CMSCE) was utilized to enhance the reliability of entropy value under every scale. Additionally, the first mode of structure was extracted by using ensemble empirical mode decomposition (EEMD) to conduct entropy analysis and evaluate the performance of damage assessment. A seven-story model was created to validate the efficiency of proposed method and damage index. Subsequently, a seven-story steel benchmark experiment, including fifteen damage cases, was conduct to compare the difference between numerical and experimental model. The confusion matrix was applied to classify the results and evaluate the performance by three indexes: accuracy, precision, and recall. The result showed the feasibility of modified structural health monitoring system and demonstrated the potential on the field of long-term monitoring.