Real-Time Structural Health Monitoring: A Comparison of Analytical Techniques

This paper discusses a Real-Time Structural Health Monitoring (SHM) technique developed to predict the presence and extent of multiple delaminations in a composite structure. This technique was developed integrating Artificial Neural Networks (ANN) and the Damage Relativity Analysis Technique (DRAT). This SHM technique uses strain signatures emanating from the surface of the structure as damage features to facilitate damage detection and classification. Damage signatures were created using Finite Element models. The DRAT was found to reduce computational time and effort drastically. This real-time system was also made independent of the magnitude of the load acting on the structure. The efficiency of the Neural Network technique was then compared with a statistical technique known as the outlier analysis technique. These techniques were applied to the simulated data, and the relative merits of the respective techniques are discussed