Wireless sensor networks for bridge structural health monitoring: a novel approach

This work presents the ML model in which data collected from the open access repository where experiments conducted on steel structure bridge data for 1-year duration are analyzed. Continuous monitoring of data from sensor nodes helps to monitor bridge health and damage at different load condition. The model combines data from sensors, applications statistics and induced load while monitoring structure. Experiments conducted have tested the ambient vibration test, explored different load condition for vibration test, and artificial damage conditions on bridge structure at different positions to collect enough data for real-time analysis at different environment condition. Five different damage scenarios were considered as a case A with no damage, in case B the vertical section was cut half at the mid-span, case C with fully cut mid-span, in case D damage was recovered by welding the vertical section, in case E 5/8th part of vertical section was cut. Ambient and load-induced vibration data are structured based on different cases using panda’s data frame. The model shows the high accuracy of deformation caused due to load induced. Results show accelerometer measurement as very good feature vectors for real-time monitoring and SARIMAX as a perfect model to evaluate time series data and perform anomaly detection simultaneously. © 2023, The Author(s), under exclusive licence to Springer Nature Switzerland AG.