Measurement of Bridge Influence Lines Based on Machine Vision and Interval Affine Algorithm

The influence line (IL) is a crucial indicator for evaluating bridge conditions. Traditional methods for measuring ILs rely on weigh-in-motion (WIM) and contact sensors, which are limited by issues such as high cost, low efficiency, high risk, and traffic obstruction. In this study, we propose a method for measuring the order that integrates machine vision and an interval affine algorithm to achieve intelligent bridge detection without the need for contact sensors, closed traffic, or WIM. This method employs machine vision to capture the dynamic displacement responses of bridge measurement points in various testing cases. Subsequently, a vehicle axle load interval mAtrix is established based on the vehicle factory information, and the IL interval under multiple testing cases is calculated using the interval affine algorithm. Finally, a support vector machine (SVM) is used to identify the actual bridge IL from the IL interval. The effectiveness of this method was demonstrated through a real bridge test, where the performance was evaluated by controlling the load and travel speed of the calibration vehicle to obtain the displacement response data under different testing conditions. The results indicate that this method can accurately identify the real bridge IL from the IL interval, with an error of 8. 48% in mixed testing cases. Furthermore, the error of the IL increases with vehicle speed, with errors of 9. 22 % . 10. 23%. and 12. 38% observed at speeds of 10 km • h-1, 20 km • h-1, and 30 km • h-1, respectively. The proposed measurement method for bridge ILs has several advantages, including noncontact operation, high accuracy, affordability, and flexibility. This effectively overcomes the drawbacks of existing contact-based bridge IL measurement methods and exhibits excellent prospects for engineering applications. © 2024 Chang'an University. All rights reserved.