Multi-Lane Load Recognition Theory Based On Computer Vision And Coupled Response Monitoring

In recent years,the rapid increase of vehicles,overload and other problems make the problem of insufficient bearing capacity of some Bridges become increasingly prominent,which is also an important factor restricting the healthy development of Bridges.Bridge dynamic weighing(BWIM)technology can effectively prevent vehicle overload,and most of the developed BWIM methods have been verified to be effective in identifying vehicle loads.However,most of these methods are limited to the load condition of single lane,and it is difficult to realize the identification of vehicle load information of temporal and spatial distribution in actual traffic.Based on this,this paper proposes a multi-lane load recognition method based on computer vision and coupled response monitoring to solve the problem of vehicle load recognition under temporal and spatial distribution.The main research content and work of this paper are as follows:(1)A multi-lane load identification method framework integrating visual sensing and optical fiber sensing technology is introduced.The computer vision technology was used to recognize the information of wheelbase and axle number of vehicles on the bridge.Meanwhile,the coupling strain response of the bridge bottom was monitored in real time by using long-standard fiber Bragg grating(FBG)sensors.The axle information identified by visual sensing was fused with the coupling strain monitoring information of vehicle weight to build a multi-lane load identification framework.(2)A multi-lane load recognition method based on computer vision and coupled response monitoring is proposed.First of all Based on the lane impact surface determined by the calibrated strain influence line,the complex situation that the vehicle deviates from the lane center line was considered.Then,the lane impact surface theory was used to establish the vehicle load identification equation.Tikhonov regularization method was introduced into the inverse solution of load and response to solve the ill-posed problem in the solution of equations.Finally identify moving vehicle loads in multiple lanes.(3)The finite element model was used to verify the effectiveness and robustness of the recommended method.In the numerical simulation analysis,the multi-lane two-vehicle and multi-lane multi-vehicle establishment of multiple working conditions were analyzed respectively,to verify the effectiveness and feasibility of the proposed method.By studying the bridge type,bridge span,sensor location,sensor number,driving speed,measurement noise and other factors on the identification effect of this method.The results show that these factors are not sensitive to the recognition effect.(4)The effectiveness of the recommended method was verified by laboratory scaling model.In laboratory model tests,a camera was first set up to identify information such as wheelbase,number of axles,lane position and running speed of moving vehicles in verified working conditions.The applicability and effectiveness of the multi-lane load identification method proposed in this paper was verified by setting a variety of test conditions.The results show that the axle load identification effect is good and stable,the identification accuracy meets the practical demand of engineering,and provides an effective solution to the problem of multi-lane load identification.