Multi-Sensor Fusion Prototype for Offshore Wind Turbine Structural Health Monitoring: Integrating ESP32-CAM Vision-Based Tracking with MPU6050 Inertial Sensing

This repository presents a prototype structural health monitoring (SHM) system for offshore wind turbine applications, integrating low-cost vision-based displacement tracking (ESP32-CAM) with inertial sensing (MPU6050). The system employs Kalman filtering for sensor data denoising and machine learning classification for real-time anomaly detection. Key technical contributions include:- Multi-sensor fusion architecture combining visual and inertial measurements- Discrete Kalman filter implementation for tilt signal stabilization (60% noise reduction)- Machine learning classifier with optimized threshold selection (F1-score: 0.86 at threshold 0.32)- 24-hour experimental validation under simulated offshore conditions- Quantitative performance validation: RMSE = 0.2024°, R² = 0.9871 between sensor modalities The system addresses critical challenges in remote offshore monitoring by providing a cost-effective, real-time solution for detecting compound failure conditions (tilt + vibration + displacement) in wind turbine support structures. All hardware schematics, Python implementation code, and processed datasets are included to ensure reproducibility and adaptation for similar structural monitoring applications.