CN-122016117-A - Suspension bridge full-bridge cable force automatic inspection and identification system and method based on sliding rail moving platform and machine vision

CN122016117ACN 122016117 ACN122016117 ACN 122016117ACN-122016117-A

Abstract

The invention belongs to the technical field of automatic inspection of full-bridge cable force of a suspension bridge, and particularly relates to an automatic inspection identification system and method of full-bridge cable force of a suspension bridge based on a sliding rail moving platform and machine vision. The system comprises a fixed track subsystem fixedly arranged on the side surface of a stiffening girder of a suspension bridge, a mobile detection vehicle subsystem movably arranged on the fixed track subsystem and used for autonomous walking along a track, collecting vibration video and environmental parameters of a sling and transmitting data in real time, and a background data processing and analyzing subsystem which is in remote communication connection with the mobile detection vehicle subsystem and used for receiving, storing and processing the data transmitted by the mobile detection vehicle subsystem. According to the invention, a three-level cooperative framework of 'fixed track-mobile detection vehicle-background processing' is constructed, full-bridge detection coverage is realized through the fixed track subsystem, autonomous movement and accurate detection are completed by the mobile detection vehicle subsystem, data integration and health evaluation are realized by the background data processing and analyzing subsystem, and detection efficiency is improved.

Inventors

Yuan Haoyun
YANG BINGCHEN
San Jianping
MA JIANYONG
LI HAOTIAN
ZHANG JIAHAO
XIAO TIANBAO
LI SIRUN

Assignees

中交第二公路工程局有限公司

Dates

Publication Date: 20260512
Application Date: 20260130

Claims (10)

1. A full-bridge cable force automatic inspection and identification system of a suspension bridge based on a sliding rail moving platform and machine vision is characterized by comprising the following components: The fixed track subsystem (1) is fixedly arranged on the side surface of the stiffening girder of the suspension bridge and provides a running track and a full-bridge space coverage for movement detection; The mobile detection vehicle subsystem (2) is movably arranged on the fixed track subsystem (1) and is used for independently walking along a track, collecting sling vibration videos and environment parameters and transmitting data in real time; The background data processing and analyzing subsystem (3) is in remote communication connection with the mobile detection vehicle subsystem (2) and is used for receiving, storing and processing data transmitted by the mobile detection vehicle subsystem (2), analyzing and calculating cable force through an algorithm, and carrying out visual display and health state evaluation; The fixed track subsystem (1), the mobile detection vehicle subsystem (2) and the background data processing and analyzing subsystem (3) form a three-level automatic inspection and identification framework in a cooperative mode.
2. The full-bridge cable force automatic inspection and identification system for the suspension bridge based on the sliding rail moving platform and the machine vision, which is disclosed in claim 1, is characterized in that the fixed track subsystem (1) comprises two tracks (11) which are arranged on the side face of a stiffening beam of the suspension bridge in parallel, the length of each track (11) covers the detection area of all slings of the full bridge, the cross section of each track (11) is I-shaped, a guide groove (111) and a rack (112) are arranged on the inner side of each track (11), the guide groove (111) is used for guiding a mobile detection vehicle to walk, the rack (112) is used for being meshed with a driving gear of the mobile detection vehicle, and safety stop tables (12) are arranged at two ends of each track (11) and used for parking, charging and maintaining the mobile detection vehicle.
3. The full bridge cable force automatic inspection and recognition system of a suspension bridge based on a sliding rail moving platform and machine vision according to claim 2, wherein the movement detection vehicle subsystem (2) comprises a vehicle body (21) and the following modules integrated on the vehicle body (21): The driving and walking module (22) comprises a driving motor (221), a gear set (222) meshed with the track rack (112), four guide wheels (223) embedded into the guide grooves (111) and an electromagnetic braking unit (224), a transmission chain (225) is arranged between the guide wheels (223) adjacent to each other in front and back, and the gear set (222) is meshed with the transmission chain (225) and is used for driving the vehicle body (21) to move along the track (11) and realizing accurate stopping; The high-precision positioning module (23) is integrated with a GPS/Beidou dual-mode satellite positioning unit and a laser ranging sensor and is used for acquiring and fusing absolute position and relative displacement information of the vehicle body (21) in real time; The environment sensing and obstacle avoidance module (24) comprises an environment sensor group (241) for acquiring wind speed, temperature and humidity and illumination intensity parameters, and a composite obstacle avoidance unit formed by a laser radar (242) and an ultrasonic sensor (243) and used for sensing an environment state and detecting obstacles on a travelling path; the visual acquisition module (25) comprises a bracket (251) capable of being adjusted at multiple angles, a high-definition industrial camera (252) arranged on the bracket, and a night light supplementing unit (253) matched with the camera, and is used for acquiring a vibration video sequence of a target adhered to the sling; The data transmission module (26) adopts a 5G communication module and is internally provided with a data cache memory, and is used for realizing bidirectional data communication between the mobile detection vehicle subsystem (2) and the background data processing and analyzing subsystem (3); the energy module (27) comprises a main lithium battery pack (271) and an auxiliary solar charging panel (272) arranged at the top of the vehicle body, and is used for providing power for the whole movement detection vehicle subsystem (2) and managing the charging and discharging process.
4. A full bridge cable force automatic inspection and recognition system for a suspension bridge based on a sliding rail moving platform and machine vision according to claim 3, wherein the background data processing and analyzing subsystem (3) comprises: The data receiving and storing server (31) adopts a cloud and local dual-backup architecture and is used for receiving and classifying and storing vibration videos, position information and environment parameter data from the mobile detection vehicle subsystem (2) and storing bridge structure parameters and history detection data; The high-level cable force analysis algorithm library (32) is integrated with a high-efficiency phase motion amplification algorithm module, a cable force calculation correction model module and a trend analysis algorithm module and is used for processing the received vibration video to extract the frequency of the sling, calculate and correct the cable force value and analyze the cable force long-term change trend; The full-bridge cable force visualization and health assessment platform (33) provides a human-computer interaction interface, supports the display of full-bridge cable force distribution in the form of a two-dimensional thermodynamic diagram and a three-dimensional model, and has the functions of automatic comparison early warning, trend curve generation, automatic generation of a detection report and remote control.
5. The full-bridge cable force automatic inspection and recognition system for the suspension bridge based on the sliding rail moving platform and the machine vision according to claim 4 is characterized in that the efficient phase motion amplification algorithm module is specifically used for converting an RGB video into a YIQ color space and extracting a brightness component, performing double-tree complex wavelet transform DTCWT decomposition on a brightness signal, amplifying phase information of a specific frequency band in a complex domain, performing double-tree complex wavelet transform DTCWT reconstruction on an amplified coefficient to obtain an amplified brightness signal, recombining the processed brightness signal with an original chroma signal and converting the processed brightness signal into the RGB space to obtain a motion amplification video, and extracting a sling natural vibration frequency through spectrum analysis on the amplification video.
6. The automatic inspection and recognition system for full-bridge cable force of a suspension bridge based on a sliding rail moving platform and machine vision according to claim 4, wherein the cable force calculation and correction model module establishes a relation between cable force and temperature based on a linear expansion coefficient and an elastic modulus temperature coefficient, and for a single sling, a cable force correction value F corr caused by temperature change is expressed as: F corr =F meas - E⋅A⋅α⋅ΔT Wherein F meas is the cable force value at the reference temperature (20 ℃), E is the elastic modulus of the material, A is the cross section area of the sling, alpha is the linear expansion coefficient of the sling material (high-strength steel wire alpha is approximately equal to 1.2 multiplied by 10 −5 /DEGC), and delta T is the variation of the measured temperature and the reference temperature.
7. The automatic inspection and recognition system for full bridge cable force of suspension bridge based on a sliding rail mobile platform and machine vision according to claim 4, wherein the trend analysis algorithm module adopts a dual early warning mechanism of trend slope early warning and cable force absolute value early warning: trend slope early warning, namely setting a primary early warning threshold value and a secondary early warning threshold value based on historical rope force attenuation slope statistical data of a healthy sling, wherein the primary early warning threshold value is set When the comprehensive attenuation slope of the target sling When the method is used, the healthy sling with the attenuation rate faster than 95% is judged, and the primary trend early warning is triggered, and the secondary early warning threshold value is obtained When (when) When the attenuation rate is abnormal, triggering secondary trend early warning, wherein, For instantaneous attenuation slope, the linear attenuation slope of healthy slings installed on the same bridge and in the same batch in the stable attenuation stage is counted The average value is Standard deviation of Triggering early warning when the instantaneous or fitting attenuation slope of the target sling exceeds a corresponding threshold value; and the cable force absolute value early warning is that a primary early warning threshold value is set to be 95% of a designed cable force value of the sling, a secondary early warning threshold value is set to be 90% of the designed cable force value, and the early warning is triggered when the actually measured cable force value of the target sling is lower than the corresponding threshold value.
8. The automatic inspection and recognition system for full-bridge cable force of suspension bridge based on a sliding rail moving platform and machine vision according to claim 3, wherein the target stuck on the sling is a regular geometric figure with high contrast, and an anti-reflection coating is sprayed on the surface of the target.
9. The automatic inspection and recognition method for the full-bridge cable force of the suspension bridge based on the sliding rail moving platform and the machine vision is characterized by comprising the following steps of: S1, initializing a system, starting a patrol task through a background data processing and analyzing subsystem (3), loading a sling database, and enabling a mobile detection vehicle subsystem (2) to start from a safe parking platform (12) after self-inspection; s2, the movement detection vehicle subsystem (2) automatically moves to the position of a target sling along the fixed track (11) according to a preset detection sequence, and monitors the environment and path safety in real time through the environment sensing and obstacle avoidance module (24); s3, after reaching a designated position, the mobile detection vehicle subsystem (2) accurately positions and brakes, and adjusts the vision acquisition module (25) to aim at a sling target to acquire current environmental parameters; s4, a vision acquisition module (25) acquires a vibration video of a sling target and uploads the vibration video to a background data processing and analyzing subsystem (3) through a data transmission module (26); S5, a background data processing and analyzing subsystem (3) calls an advanced cable force analysis algorithm library (32), performs motion amplification processing and spectrum analysis on the vibration video to extract frequency, and calculates a corrected cable force value by combining a cable force calculation correction model; S6, displaying the calculated cable force value on a visual platform (33), and calling a trend analysis algorithm to evaluate and early-warn the health state; S7, after the current sling detection is completed, the movement detection vehicle subsystem (2) is unlocked and moves to the next sling, and the steps S3 to S6 are repeated until the detection of all slings of the full bridge is completed; And S8, after the full-bridge detection is finished, the mobile detection vehicle subsystem (2) automatically returns to the safety stop table (12), and the background data processing and analyzing subsystem (3) generates and outputs a complete inspection report.
10. The automatic inspection and recognition method for full bridge cable force of a suspension bridge based on a sliding rail moving platform and machine vision according to claim 9, wherein in the moving process of steps S2 and S7, if the environment sensing and obstacle avoidance module (24) detects that the wind speed continues to exceed 8m/S or recognizes that an obstacle exists in front, the movement detection vehicle (2) automatically executes the operation of suspending, waiting or returning to the safety stop (12).

Description

Suspension bridge full-bridge cable force automatic inspection and identification system and method based on sliding rail moving platform and machine vision Technical Field The invention belongs to the technical field of automatic inspection of full-bridge cable force of a suspension bridge, and particularly relates to an automatic inspection identification system and method of full-bridge cable force of a suspension bridge based on a sliding rail moving platform and machine vision. Background The suspension bridge is used as one of the main flow forms of the large-span bridge, and is widely applied to traffic construction of complex geographic environments such as rivers, lakes, seas, deep mountain canyons and the like by virtue of the advantages of strong spanning capacity, reasonable stress and the like. The sling is used as a core stress member of the suspension bridge, plays a key role in transmitting the dead weight of the stiffening girder and the bridge deck load to the main cable, and the health state of the sling directly determines the overall structural safety and the service life of the bridge. The cable force is a core index reflecting the health state of the sling, and the sling is easily affected by factors such as alternating load, environmental corrosion (such as wind and rain corrosion and salt spray corrosion), material aging, fatigue damage and the like in the long-term service process, and is easy to suffer from cable force attenuation, looseness, wire breakage and the like. If the slings with abnormal cable force cannot be found and processed in time, the load distribution is possibly unbalanced, the interlocking damage of other components such as a main cable, a stiffening girder and the like is caused, and even serious safety accidents such as bridge collapse and the like can be caused. Therefore, the accurate detection of the rope force of the full-bridge sling of the suspension bridge is a key link for guaranteeing the safe operation and maintenance of the bridge structure. Currently, suspension bridge cable force detection has become a research focus and engineering difficulty in the field of bridge health monitoring. Along with the continuous increase of traffic flow and the increase of service life of bridges, higher and higher requirements are put forward on the accuracy, efficiency, automation degree and safety of cable force detection. The traditional detection method and the prior art scheme gradually expose a plurality of limitations in practical engineering application, and are difficult to meet the requirements of high-efficiency, accurate and automatic inspection of full bridge cable force, and development of an automatic inspection and identification system of full bridge cable force with autonomous movement, accurate identification, data real-time uploading and visual analysis is urgent. At present, the suspension bridge cable force detection technology is mainly divided into two main types of contact detection and non-contact detection, and various technologies have obvious defects in practical application, and the specific steps are as follows: 1. The contact detection technology (such as vibration method, pressure sensor method and magnetic flux method) ① has complex operation and extremely low detection efficiency. The method needs to be used for detecting personnel to approach slings by means of an overhead working platform (such as a bridge detection vehicle and a hanging basket), manually installing a sensor or a detection device, has a full-bridge detection period of hundreds of slings for a large-span suspension bridge, and is long in days and even weeks, so that normal traffic of the bridge is seriously influenced, ② is poor in safety, the overhead working environment is complex, the influence of natural environments such as wind power and precipitation is easily caused, potential safety hazards such as personnel falling and equipment falling exist, ③ detection accuracy is greatly influenced by manual operation, the installation positions and operation methods of different detection personnel can cause large discreteness of detection data, the consistency and reliability of the data are difficult to ensure, ④ cannot realize real-time dynamic detection, traffic is required to be interrupted or traffic flow is limited in the detection process, and great interference is caused to traffic. 2. The traditional non-contact detection technology (such as machine vision detection based on a fixed camera and laser Doppler vibration detection) has the advantages that the detection range is limited, the fixed camera can only cover a few slings in a specific area, if full-bridge detection is to be realized, a large number of cameras are required to be arranged on a bridge, equipment cost is high, installation and maintenance difficulties are high, the equipment is easily influenced by shielding of bridge structures, detection blind areas exist, the ② environmental adaptab