Bridge Health Monitoring via Vibration Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Structural health monitoring (SHM) of bridges faces challenges due to complex and unknown physical models, operational, and environmental variations, making it difficult to accurately determine the health condition of structural components without extensive calibration and physical models.
Innovation Solution
A computer-implemented method using vibration data from accelerometers, which involves data calibration through principal component analysis, feature extraction via fast Fourier transform, and classification by a support vector machine classifier, trained with historical data to determine the health condition of bridge components without requiring a physical model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a physical model is used to determine the health condition of structural components, then measurement precision can be improved, but device complexity increases due to the need for extensive calibration and model development
Solution Approach 1:
The patent replaces complex physical models with a machine learning-based system that uses vibration data from accelerometers. Instead of relying on detailed mechanical models of the structure, the system extracts features from vibration data and uses a support vector machine classifier to determine health conditions, thereby reducing device complexity while maintaining or improving detection accuracy
Solution Approach 2:
The patent creates a simplified representation of the structure's health state through vibration data features rather than using complex physical models. The support vector machine classifier learns patterns from vibration data that copy the essential characteristics of structural health without requiring explicit physical model calibration
2Measurement precision
If extensive calibration is performed to improve measurement precision, then health condition determination accuracy improves, but loss of time increases due to calibration requirements
Solution Approach 1:
The system performs preliminary feature extraction and classification model training using historical vibration data. The support vector machine classifier is trained in advance on labeled data representing different health conditions, so that when new vibration data arrives, the system can immediately classify health conditions without requiring time-consuming calibration procedures for each new measurement
3Reliability
If a complex physical model is used to monitor structure health, then reliability can be improved, but ease of operation deteriorates due to difficulty in calibration and model maintenance
Solution Approach 1:
The system performs self-calibration by automatically extracting features from vibration data and training the support vector machine classifier on historical data patterns. The system monitors structural health independently without requiring manual calibration or model maintenance, as the machine learning model continuously adapts to normal variations in the structure's vibration characteristics
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach achieves high accuracy in determining the health condition of bridge components, with over 99% classification accuracy for supervised learning and 85% for unsupervised learning, enabling effective monitoring and maintenance without the need for complex physical models.
Implementation Method 1
receiving or accessing vibration data of the part of the bridge, the vibration data being measured by at least one accelerometer
Implementation Method 2
extracting a feature of the vibration data based on frequency analysis of the vibration data; The frequency analysis may comprise fast Fourier transform of the vibration data
Implementation Method 3
determining the health condition of the part of the bridge by a support vector machine classifier based on the feature of the vibration data
Implementation Method 4
The method may further comprise calibrating coordinate axis directions of the at least one accelerometer based on the vibration data of the part of the structure by principal component analysis to calibrate the vibration data
Data Source
AI summary
The disclosure relates to structural health monitoring (SHM). In particular determining a health condition of a structure, such as a bridge, based on vibration data measured of the bridge. Measured vibration data is calibrated (410-450). Features are then extracted from the calibrated data (610-630) and a support vector machine classifier is then applied (720) to the extracted features to determine (730) the health condition of a part of the structure. Training of the support vector machine classifier by a machine learning process (910) is also described.