◦ Devised a vibration-based damage-detection approach using AI methods on time response simulated data.

◦ Extracted different damage-sensitive features based on signal statistics and modal properties; analyzed signal response time-series using AR and ARX models; proposed statistical model for feature discrimination.

◦ Ranked informative, discriminating and independent damage-sensitive features based on their effectiveness; employed support vector machines to classify structural condition states as damaged and undamaged.

Project Report: The Application of Machine Learning to Structural Health Monitoring