CN-122009556-A - Method and system for monitoring new rail state before rail change

CN122009556ACN 122009556 ACN122009556 ACN 122009556ACN-122009556-A

Abstract

The invention discloses a method and a system for monitoring a new rail state before rail change of a railway, and relates to the technical field of railway engineering construction monitoring. The method comprises the steps of collecting image data of a new rail and an old rail through an unmanned aerial vehicle, collecting rail temperature data of the new rail, accurately registering image frames with rail mileage, extracting rail structure characteristic lines of the new rail and the old rail from images based on standard structural parameters of railway rails, quantitatively calculating structural deformation of the new rail relative to the old rail, comprehensively judging whether the new rail has a rail expansion phenomenon and the degree of the rail expansion phenomenon according to the structural deformation and the rail temperature data, and generating quantized construction treatment measure instructions according to judging results. The invention also provides a corresponding monitoring system. According to the invention, manual inspection is replaced by unmanned aerial vehicle inspection, and image and temperature two-dimensional intelligent analysis is combined, so that the high-efficiency, accurate and safe monitoring of the new rail state before rail replacement is realized, a quantization processing instruction can be automatically generated, and the preparation efficiency and the intelligent level of railway rail replacement construction are remarkably improved.

Inventors

YANG HAIPENG
GONG HUAN
WANG QI
SU ZEHUA
SHI PENGFEI
LANG ZEJUN
HE JIANJIA
ZHANG YANYUN
FANG LEI
CHENG PENGFEI
XIN JIE
XU XINHUA
WANG JIAWEI

Assignees

中交第二航务工程局有限公司

Dates

Publication Date: 20260512
Application Date: 20260309

Claims (10)

1. A method for monitoring the state of a new rail before changing the rail of a railway is used for monitoring the state of the new rail to be paved, Acquiring video image data of a new rail and an old rail of a monitoring rail section through image acquisition equipment carried by an unmanned aerial vehicle, and acquiring rail temperature data of the new rail; preprocessing the video image data, and accurately registering the preprocessed image frame with track mileage information; based on standard structural parameters of the railway track, directionally extracting track structural feature lines of the new track and the old track from the registered images; Based on the extracted characteristic lines, quantitatively calculating the structural deformation of the new track relative to the old track or the standard track; Combining the structural deformation and the rail temperature data, and judging whether the new rail has a rail expansion phenomenon and the rail expansion degree according to a preset threshold value; and generating a corresponding construction treatment measure instruction according to the judging result.
2. The method for monitoring the new rail state before rail change of the railway is characterized in that the method for registering comprises the steps of converting position information acquired by an unmanned aerial vehicle positioning module into rail mileage and binding the rail mileage with video frames to ensure that each frame of image corresponds to a unique rail mileage section.
3. The method for monitoring the state of a new rail before rail change of a railway is characterized in that the method for extracting the structural characteristic lines of the new rail and the old rail comprises the steps of adopting a Canny edge detection algorithm, taking preset standard section parameters of the steel rail as constraint conditions, extracting rail head central axes, rail waist outer straight line segments and rail bottom upper edges of the new rail and the old rail from preprocessed images as core structural characteristic lines, and fitting the extracted structural characteristic lines through a Hough straight line detection algorithm to obtain linear extension tracks of the new rail and the old rail.
4. The method for monitoring the state of the new rail before rail change of the railway is characterized in that the quantitative analysis method comprises the steps of equally dividing structural feature lines of the new rail and the old rail into feature matching points according to mileage segments, matching the new rail with the old rail feature points in the same mileage segments, calculating the transverse offset and the vertical lifting of each matching point, and calculating the curvature of the structural feature lines of the new rail in continuous mileage segments to judge whether the new rail has arch deformation or not.
5. The method for monitoring the state of a new rail before rail change of a railway according to claim 4, wherein the method for judging whether the new rail has a rail expansion phenomenon according to a preset threshold value comprises the steps of marking the new rail as an image suspected rail expansion phenomenon if any one of the lateral offset, the vertical lifting amount or the bending degree of a certain mileage section exceeds a corresponding first threshold value; for the mileage section marked as the suspected expanded rail of the image, if the rail temperature data exceeds the upper limit of the construction allowable rail temperature and the distance between the new rail and the old rail calculated according to the characteristic lines of the new rail and the old rail is smaller than the standard distance, finally judging that the expanded rail is determined; And grading the rail expansion degree according to the ratio of the actual values of the transverse offset, the vertical lifting amount and the bending degree to the corresponding threshold values.
6. The method for monitoring the state of a new rail before rail change of a railway according to claim 1, wherein the method further comprises the steps of constructing a calculation model of an abnormality grade coefficient for representing the abnormality degree of the state of the new rail: Carrying out normalization and standardization treatment on basic acquisition indexes, wherein the basic acquisition indexes comprise rail temperature, new and old rail spacing and comprehensive rail expanding offset, and the comprehensive rail expanding offset is obtained by comprehensive calculation according to the transverse offset, the vertical lifting amount and the bending degree; Constructing a calculation formula of an abnormal grade coefficient by adopting a weighted comprehensive membership method: k is an abnormal grade coefficient; w 1 -weight coefficient of rail temperature index; t Label (C) -a normalized value of rail temperature; w 2 -the weight coefficient of the spacing index; D Label (C) —a normalized value of the spacing; w 3 -weight coefficient of the rail expansion offset index; Δs Label (C) —a normalized value of the rail-expanding offset; and judging the new track state as normal, slight abnormality, moderate abnormality or severe abnormality according to the value range of the calculated abnormality grade coefficient.
7. The method for monitoring the state of a new rail before rail change of a railway according to claim 6, further comprising constructing a calculation model of a processing intensity coefficient for quantifying the intensity of a construction processing measure, and calculating the processing intensity coefficient according to the following formula: M is the treatment intensity coefficient; k, an abnormal grade coefficient; t Label (C) -a normalized value of rail temperature; D Label (C) —a normalized value of the spacing; Δs Label (C) —a normalized value of the rail-expanding offset; alpha-rail Wen Xiuzheng coefficient; beta-pitch correction factor; Gamma, an expanding track offset correction coefficient; And dividing the treatment measure grade into treatment-free, on-site fine-tuning, mechanical correction or shutdown correction according to the value range of the calculated treatment intensity coefficient.
8. The method for monitoring the new rail state before rail change of a railway according to claim 7, wherein the method for generating the corresponding construction treatment measure instruction according to the judging result comprises the following steps: when the processing measure grade is on-site fine adjustment, generating an instruction comprising manual local adjustment and specific adjustment quantity by using a simple tool, wherein the adjustment quantity is not more than half of the comprehensive rail expansion offset threshold value; when the processing measure grade is mechanical correction, generating an instruction containing accurate correction and specific correction force by adopting an orbit correction machine, wherein the correction force is determined according to comprehensive rail expansion offset and rail temperature calculation; when the treatment measure grade is shutdown correction, generating an instruction for immediately suspending construction, removing a new rail of the mileage section, and re-paving and adjusting a rail temperature adaptation scheme; And inputting the retested data after the renovation into an algorithm, comparing the index changes before and after the treatment, and automatically improving the treatment intensity coefficient and iteratively updating the weight coefficient and the correction coefficient if the index changes do not return to normal.
9. The method for monitoring the state of a new rail before rail change of a railway according to claim 1, further comprising an intelligent decision algorithm flow, wherein the flow comprises: receiving original acquisition data and calculating a standardized value thereof; substituting the standardized value into an abnormal grade coefficient calculation model, and judging the abnormal grade according to the abnormal grade coefficient; Substituting the abnormal grade coefficient and the standardized value into a processing intensity coefficient calculation model, and matching corresponding quantized construction treatment measures according to the processing intensity coefficient; generating and outputting a quantization processing instruction containing specific mileage and construction parameters; and the processing instruction is issued to the construction instruction receiving terminal through the wireless communication module.
10. A new rail condition monitoring system before a change of rail for performing the method of any one of claims 1 to 9, comprising: The unmanned aerial vehicle platform is provided with image acquisition equipment and a positioning module, and is used for acquiring video image data of a new track and an old track and acquiring track mileage information; The rail temperature acquisition equipment is used for acquiring rail temperature data of the new rail; A ground control terminal comprising: The data processing module is used for receiving and processing video image data, track mileage information and track temperature data; An analysis module for performing the steps of the method according to any one of claims 1 to 9; And the communication module is used for sending the quantization processing instruction to the construction instruction receiving terminal.

Description

Method and system for monitoring new rail state before rail change Technical Field The invention relates to the technical field of railway engineering construction monitoring, in particular to a method and a system for monitoring a new rail state before rail change of a railway. Background Before the existing line rail change construction of the common railway, the state of the newly paved rail needs to be comprehensively monitored, and the method mainly comprises rail temperature monitoring, new and old rail distance monitoring, rail expansion or not and the like. The traditional monitoring method relies on a guard to take workers to walk on the track, and has the following remarkable defects: The manpower and material resources are consumed greatly, a large number of people are required to carry out grouping operation, and the cost is high; the efficiency is low, the track inspection speed is low, and the track of 8 km takes several hours or even longer; The data feedback is delayed, namely the time for manually observing the data record and transmission is consumed, and hidden danger is difficult to discover in time; The safety risk is high, and hiking and rail-patrol personnel need to operate along the railway and face safety threats such as train passing. These problems seriously affect the preparation efficiency before the rail changing construction, and bring great hidden trouble to the running and construction safety of the railway train. Therefore, a new track condition monitoring technology that is efficient, accurate and safe is needed to replace the traditional manual mode. Disclosure of Invention The method solves the problems of high manpower consumption and high safety risk by replacing manual inspection by unmanned aerial vehicle acquisition, low efficiency and data feedback lag by accurate registration and automatic analysis of images and mileage, and high subjectivity and lack of quantitative basis by introducing a two-dimensional comprehensive judgment and quantitative decision model. The technical proposal of the application is that a method for monitoring the state of a new rail before changing the rail of a railway is used for monitoring the state of the new rail to be paved, comprising, Acquiring video image data of a new rail and an old rail of a monitoring rail section through image acquisition equipment carried by an unmanned aerial vehicle, and acquiring rail temperature data of the new rail; preprocessing the video image data, and accurately registering the preprocessed image frame with track mileage information; based on standard structural parameters of the railway track, directionally extracting track structural feature lines of the new track and the old track from the registered images; Based on the extracted characteristic lines, quantitatively calculating the structural deformation of the new track relative to the old track or the standard track; Combining the structural deformation and the rail temperature data, and judging whether the new rail has a rail expansion phenomenon and the rail expansion degree according to a preset threshold value; and generating a corresponding construction treatment measure instruction according to the judging result. The registration method comprises the steps of converting position information acquired by an unmanned aerial vehicle positioning module into track mileage and binding the track mileage with video frames to ensure that each frame of image corresponds to a unique track mileage section. The method for extracting the track structure characteristic lines of the new track and the old track comprises the steps of adopting a Canny edge detection algorithm, taking preset standard section parameters of the steel rail (obtained through railway track design specifications) as constraint conditions, extracting the central axes of the rail heads, the outer straight line sections of the rail web and the upper edge of the rail bottom of the new track and the old track from the preprocessed image to serve as core structure characteristic lines, and fitting the extracted structure characteristic lines through a Hough straight line detection algorithm to obtain the linear extension tracks of the new track and the old track. The method for quantitatively analyzing the state of the new rail before rail change of the railway comprises the steps of equally dividing structural feature lines of the new rail and the old rail into feature matching points according to mileage segments, matching the new rail and the old rail in the same mileage segment with the feature points of the old rail, calculating the transverse offset and the vertical lifting of each matching point, and calculating the curvature of the structural feature lines of the new rail in the continuous mileage segments to judge whether the new rail has arch deformation or not. The method for judging whether the new rail has the rail expansion phenomenon according to the preset threshold val