🤖 Applied Machine Learning

✅ Machine Learning for Structural Health Monitoring Systems

During my 6-month-long study at the prestigious SMSS Lab, IIT Kanpur, I focused on leveraging Machine Learning techniques to address the critical area of Structural Health Monitoring Systems. The primary objective was to develop predictive models capable of assessing and predicting the structural integrity of various mechanical structures.

The study involved the classification of defects based on Time Domain and Frequency Domain features obtained from vibration response data, acquired through the use of a Laser Doppler Vibrometer. The research encompassed Civil Engineering structures such as bridges and dams, as well as Mechanical Engineering structures, including aircraft parts and windmill fan blades.

Key highlights of the project include:

  1. Feature Extraction and Classification: I extracted essential Time Domain and Frequency Domain features from the vibration response data. These features served as critical inputs for the Machine Learning models used for defect classification.

  2. Machine Learning Model Selection: The project involved an in-depth exploration of traditional Machine Learning algorithms, including Linear Regression, Logistic Regression, k-Nearest Neighbors, Naïve Bayes, Support Vector Machines, and Tree-based methods such as Decision Trees, and Random Forest.

  3. Results: By employing the Random Forest algorithm, I achieved an impressive accuracy of approximately 92% in the defect classification task.

Overall, this project allowed me to gain a comprehensive understanding of applying Machine Learning techniques to practical applications like Structural Health Monitoring systems. It provided valuable insights into how Machine Learning can be harnessed to advance the safety and reliability of critical engineering structures.