CN114379598B

CN114379598B - Railway comprehensive inspection system

Info

Publication number: CN114379598B
Application number: CN202210096400.1A
Authority: CN
Inventors: 李剑锋; 王文昆; 季育文
Original assignee: Zhuzhou CSR Times Electric Co Ltd
Current assignee: Zhuzhou CRRC Times Electric Co Ltd
Priority date: 2022-01-26
Filing date: 2022-01-26
Publication date: 2023-06-30
Anticipated expiration: 2042-01-26
Also published as: CN114379598A

Abstract

The invention discloses a railway comprehensive inspection system, wherein a gesture determination positioning system acquires GNSS (Global navigation satellite System), IMU (inertial measurement Unit) and DMI (digital mobile unit) data of a comprehensive inspection vehicle, a laser radar acquires three-dimensional coordinates of an object to be detected, and a panoramic camera acquires panoramic image data. The space-time synchronization system provides clock synchronization for three-dimensional coordinates, IMU, DMI and panoramic image data according to GNSS time service. The storage/control system stores the three-dimensional coordinates, the IMU, the DMI and the panoramic image data after clock synchronization in real time. And the point cloud processing system firstly uses GNSS, IMU and DMI data to fuse and calculate POS data, then uses the POS data and laser radar data to jointly calculate point cloud data, and fuses the point cloud data and panoramic image data to obtain the real scenic spot cloud. And the comprehensive inspection analysis system takes the point cloud and the panoramic image data as data sources to carry out railway inspection analysis. The invention can solve the technical problems of high complexity, high labor cost and high data processing difficulty of the existing system.

Description

Railway comprehensive inspection system

Technical Field

The invention relates to the technical field of railway engineering machinery, in particular to a railway comprehensive inspection system for maintaining railway lines.

Background

With the rapid increase of railway mileage in China, the detection tasks of railway inspection, railway infrastructure and parts thereof are becoming more and more heavy. In the past, inspection of railways and detection of infrastructure has been accomplished by manual inspection or by dedicated equipment. The manual inspection has high requirements on the quality of workers, high working strength and missed inspection caused by easy dispersion of the attention of people. With the advancement of technology, the inspection of railway infrastructure and its components is gradually upgraded to specialized inspection tools or work vehicles. The variety of railway infrastructure and parts thereof is high, which directly results in a large number of special detection tools or systems, and each type of tools needs to be provided with special technical staff for operation, which is time-consuming and labor-consuming for railway working departments. In recent years, a railway comprehensive inspection vehicle gradually pushes to a railway market, the comprehensive inspection vehicle integrates various detection tools and detection systems on the same inspection vehicle, and the inspection work of a plurality of parts of the railway is completed by using one operation skylight.

The detection item points of the comprehensive inspection vehicle generally comprise steel rails, rail side facilities, the size of a contact net and defect detection, and can be applied to departments of work, electric service, power supply and the like. The existing comprehensive inspection vehicle is provided with different detection devices aiming at different detection item points, most of the detection devices are cameras, the used technology is acquisition and processing of two-dimensional images, and the detection device comprises a small amount of small-scale short-distance three-dimensional detection. Before operation, all detection systems are required to be started, during operation, the system collects original photos (part of the systems can detect in real time), and after operation is completed, data are transferred to a ground processing system for storage and processing. The operation mode needs more personnel, long time, high labor intensity and low efficiency, and the checking effect is directly related to the experience and responsibility of post-treatment personnel. According to the operation method, the existing comprehensive inspection system is complex in configuration, low in utilization rate of hardware equipment and high in economic cost, and operators are needed more during operation because of more system sets. Meanwhile, the existing comprehensive inspection system is mainly based on image processing, the number of the configured cameras ranges from millions of pixels to tens of millions of pixels, the amount of image data generated by one operation is huge, and the off-line processing is also difficult.

In the prior art, a baby chicken middle car-carrying engineering machinery limited company applies for 2017 in 09 month 15 and discloses a Chinese invention application with publication number of CN107472269A in 2017 in 12 month 15, a car body is arranged on a power bogie through a car frame, living facilities and maintenance detection units are arranged at the top and the inside of the car body, and electric interfaces and external sockets which are convenient to connect with external vehicles and electric equipment are arranged at the front end and the rear end of the car body. The bottom of the vehicle body is provided with a power system I and a power system II, the power system I is connected with a power bogie and a driving control unit for controlling the running of the whole vehicle, and the power system II is connected with a living facility and maintenance detection unit to realize a single power supply mode of the running of the whole vehicle, the living facility and the maintenance detection unit. The power system I and the power system II arranged at the bottom of the vehicle body respectively supply power to the whole vehicle running and living facilities and the maintenance and repair detection unit, so that the safety performance of the whole vehicle running is improved, the comprehensive inspection vehicle integrating the detection functions of the power supply, the work and the electric service system is realized, the speed of the inspection vehicle is up to 160km/h, the working efficiency is greatly improved, and the detection requirement of high-speed rail is met.

In addition, apply for 18 on 2013 04 month by the south car-to-south car-generation engineering machinery limited company of baby, and disclose on 2016 02 month 03 day, china's invention application of publication No. CN103231719A discloses a double-power railway track comprehensive inspection vehicle, including automobile body, frame, front power unit and back power unit, the automobile body sets up on the frame, and the automobile body both ends are equipped with main cab and auxiliary cab respectively, are equipped with detection room and control room between main cab and auxiliary cab, and the frame lower extreme both sides are equipped with front power bogie and back power bogie respectively, and front power unit provides driving power for front power bogie, and back power unit provides driving power for back power bogie, is equipped with diesel tank and diesel tank is fixed in the frame lower extreme between front power unit and back power unit, installs a plurality of detection device on the frame. The invention provides self-powered, and can be used for carrying out quick dynamic scanning, intelligent state judgment and alarm on the states of structural components on the surface of the high-speed railway, the contours of the steel rails and the line limits, timely grasping the appearance state of the high-speed railway working equipment, guiding daily maintenance and repair, and guaranteeing safe and reliable operation of the high-speed railway.

The application of the comprehensive inspection vehicle improves the service efficiency of the operation skylight, but has the following obvious technical defects:

1) The number of detection systems integrated by the comprehensive inspection vehicles produced by different manufacturers is not completely the same, generally between 12 and 16, but the data compatibility among a plurality of detection devices becomes a problem, and the obtained detection data lack intercommunication and interconnection, and the multiplexing of the devices is not realized;

2) Because only single detection equipment is intensively installed, the number of detection equipment is large because the number of detection item points is large; meanwhile, each device is required to be provided with a corresponding control system and a corresponding post-processing system, so that the comprehensive inspection vehicle is complex to operate, and the data post-processing difficulty and the task weight are high;

3) The number of the detection devices is not reduced, so that the number of the operators of the devices is not reduced, and the detection systems of the comprehensive inspection vehicle are numerous, so that the economic, personnel and use costs of the comprehensive inspection application are high, the operation process is complex, and the comprehensive inspection operation requirements of the railway are not adapted to and met.

Disclosure of Invention

In view of the above, the invention aims to provide a comprehensive railway inspection system, which solves the technical problems of high complexity, high labor cost and high data processing difficulty of the existing system, and cannot adapt to and meet the increasing requirements of comprehensive railway inspection operation.

In order to achieve the above purpose, the present invention specifically provides a technical implementation scheme of a comprehensive inspection system for railways, including: the system comprises a three-dimensional scanning system, a space-time synchronization system, a storage/control system and a patrol application system, wherein the three-dimensional scanning system comprises a gesture determining and positioning system, a laser radar and a panoramic camera, and the patrol application system comprises a point cloud processing system and a comprehensive patrol analysis system. The attitude determination positioning system acquires GNSS, IMU and DMI data of the railway comprehensive inspection vehicle, the laser radar acquires three-dimensional coordinate data of an object to be detected, and the panoramic camera acquires panoramic image data. The space-time synchronization system provides clock synchronization for three-dimensional coordinates, IMU, DMI and panoramic image data according to GNSS time service. The storage/control system stores the three-dimensional coordinates, the IMU, the DMI and the panoramic image data after clock synchronization in real time. The point cloud processing system firstly uses GNSS, IMU and DMI data to fuse and calculate POS data, then uses the POS data and laser radar data to jointly calculate point cloud data, and then fuses the point cloud data and panoramic image data to obtain the real scenery point cloud. The comprehensive inspection analysis system uses the point cloud and the panoramic image data as data sources to conduct railway inspection analysis.

Further, the comprehensive inspection analysis system comprises a side slope disaster inspection module, a plurality of laser mark targets are arranged on the railway side slope to be inspected, the three-dimensional scanning system is used for collecting original data in field operation, and the point cloud processing system is used for generating point cloud and panoramic image data in field operation. The slope disaster inspection module accurately extracts laser mark targets from the scanned point cloud data, and sets unique numbers or names for the extracted laser mark targets, wherein the numbers or names of the laser mark targets at the same position in the point cloud obtained by scanning in different periods are the same. And acquiring the current-period point cloud, reading in the to-be-compared previous-period point cloud, performing coarse superposition display on the two-period point cloud by using the mileage and coordinates of the laser mark targets, and calculating accurate point cloud superposition parameters by using the coordinates of the previous-period laser mark targets and the coordinates of the current-period laser mark targets. And carrying out coordinate conversion on the current-period coordinates by using accurate point cloud superposition parameters, carrying out accurate point cloud superposition comparison after conversion, and calculating coordinate differences of the same positions in the two-period point clouds after superposition. And (3) checking the change amount, and taking panoramic images in the current period and the earlier period for the region with large coordinate difference, and manually confirming whether slope change occurs. And (3) storing the coordinate difference of the point cloud superposition contrast, reading in the point cloud contrast coordinate difference in the earlier stage, drawing a coordinate difference change curve, detecting whether the coordinate difference has a trend of gradually amplifying, and carrying out early warning or alarming on an area where the coordinate difference is gradually amplified.

Further, the slope disaster inspection module gridding the point cloud in the current period, performing plane fitting on the point cloud grid C processed first to obtain a plane P, and solving the center coordinate O of the point cloud grid C. And then, obtaining the normal line of the plane P passing through the point O, obtaining the normal line L, obtaining the intersection point J of the normal line L and the earlier point cloud, and calculating the distance of the line segment OJ, namely the variation of the same position in the two-stage point cloud. And traversing each grid in the current-period point cloud, and calculating and obtaining the front and rear period variation of each grid point cloud, thereby obtaining the coordinate difference of the same position in the two-period point cloud.

Further, the comprehensive inspection analysis system comprises a slope disaster evaluation module, a plurality of laser mark targets are arranged on the railway slope to be inspected, the three-dimensional scanning system is used for collecting original data in field operation, and the point cloud processing system is used for generating point cloud and panoramic image data in field operation. The slope disaster evaluation module extracts laser mark targets from the scanned point cloud data, and sets unique numbers or names for the extracted laser mark targets, wherein the numbers or names of the laser mark targets at the same position in the point clouds at different periods are the same. And acquiring the current-period point cloud, reading in the to-be-compared previous-period point cloud, performing coarse superposition display of the two-period point cloud by using the mileage and coordinates of the laser mark targets, and calculating accurate point cloud superposition parameters by using the coordinates of the previous-period laser mark targets and the current-period laser mark targets. And carrying out coordinate conversion on the current-period coordinates by using accurate superposition parameters, carrying out accurate point cloud superposition comparison after conversion, calculating the point cloud coordinate difference of the same position in the two-period point clouds after superposition, and focusing on the difference value of the elevation coordinate z. And calculating the volume of the part clamped by the two point clouds by utilizing the coordinate difference of the two point clouds, namely, the earth and stone side of the landslide, and calculating the area of the landslide part. And (3) taking panoramic image data scanned in the previous period, checking actual topography and topography of a disaster point, intercepting and storing the three-dimensional point cloud according to mileage, positioning the three-dimensional point cloud to the mileage of an emergency event, and taking the three-dimensional point cloud and the panoramic image at the position as a base map of emergency treatment to provide map reference for the emergency treatment.

Further, the comprehensive inspection analysis system comprises a line virtual inspection/multiplication module, the three-dimensional scanning system is used for collecting original data during field operation, and the point cloud processing system is used for generating point cloud, panoramic images and POS track files during field operation. The virtual line inspection/multiplication module utilizes the calibrated panoramic image as a true color value of the gray point cloud to generate a true color point cloud, divides the true color point cloud into point cloud segments, and simultaneously divides the POS track file into POS track file segments corresponding to the point cloud segments. And reading in a point cloud segment of the mileage to be inspected and a corresponding POS track file. The default viewpoint of the line virtual inspection/multiplication module display interface is positioned at the starting point of the POS track file, the view field direction is consistent with the POS track file, and after the user clicks, the point cloud roaming can be performed in various modes, so that the visual effect of virtual inspection/multiplication is realized. In the point cloud roaming process, a user can change the view field direction at any time and pause the movement of the point cloud, after the movement of the point cloud is paused, the user can check a specific part through zooming, and the image information of the panoramic image checking part corresponding to the point cloud is called out. And if a certain component has a defect or a fault, the virtual line inspection/multiplication module outputs defect information by creating a defect report. And when the inspection of the point cloud of one section is finished, automatically reading the point cloud of the next section for inspection until the inspection of the line is finished.

Further, the comprehensive inspection analysis system comprises an acoustic-wind barrier inspection module, a plurality of laser mark targets are arranged on a railway barrier to be inspected, the three-dimensional scanning system is used for collecting original data during field operation, and the point cloud processing system is used for generating point cloud, POS track files and panoramic image data during field operation. The sound-wind barrier inspection module manually intercepts the point cloud of the road section only comprising the barrier, if the road section is long, the point cloud is divided into point cloud sections, meanwhile, the POS track file is divided into POS track file sections corresponding to the point cloud sections, and the POS track file is matched with the point cloud. Reading a point cloud segment of a mileage to be inspected and a corresponding POS track file, separating the point cloud of an aerial part to obtain a ground point cloud, and dividing the ground point cloud to separate the point cloud of a barrier part. And extracting point clouds C taking the POS track line as a central line, setting the width range at two sides of the point cloud C, performing high-level filtering processing on the point clouds C, removing the point clouds with the elevation coordinate z value smaller than the set value to obtain residual point clouds C ', calculating the point cloud normal line of the C', removing the point clouds which are not perpendicular to the POS track line in the point clouds, and finally obtaining the residual point clouds CP only comprising the barrier part. Extracting laser mark targets from the point cloud CP, and setting unique numbers or names for the extracted laser mark targets, wherein the numbers or names of the laser mark targets at the same position in the point cloud at different periods are the same. And reading in the early-stage point clouds to be compared, performing coarse superposition display of the two-stage point clouds by using the mileage and coordinates of the laser marking targets, and calculating accurate point cloud superposition parameters by using the coordinates of the early-stage laser marking targets and the coordinates of the laser marking targets in the current stage. And carrying out coordinate conversion on the current-period coordinates by using accurate superposition parameters, carrying out accurate point cloud superposition comparison after conversion, traversing the current-period point clouds after superposition, counting the number of the point clouds of each point cloud within a set radius, and regarding the point clouds of which the number is not 0 in the previous-period point clouds and the number is 0 in the current-period point clouds, namely, considering that the barrier in the current-period point clouds is damaged. And (3) calculating the area of the damaged part by statistics, and storing the panoramic image and the three-dimensional point cloud scanned in the current period at the damaged mileage so as to be convenient for the staff to confirm afterwards.

Further, the comprehensive inspection analysis system comprises a contact net wire pole verticality inspection module, wherein the contact net wire pole verticality inspection module is used for carrying out point cloud segmentation after collecting holographic point clouds around a railway to obtain point clouds of a wire pole part, solving the point cloud center of the wire pole part, carrying out linear fitting to obtain a vector equation of the contact net wire pole center, and obtaining the verticality of the wire pole.

Further, the three-dimensional scanning system collects original data during field operation, and the point cloud processing system generates point cloud, POS track files and panoramic image data during field operation. The overhead line pole verticality inspection module divides the point cloud into point cloud segments, divides the POS track file into POS track file segments corresponding to the point cloud segments, and matches the POS track file with the point cloud file. And reading in a point cloud segment of the mileage to be inspected and a corresponding POS track file, deleting redundant point clouds by using the POS track, searching all telegraph poles by using the residual point clouds as C, calculating the position and the slope of each telegraph pole, and storing and outputting a calculation result.

Further, the overhead line pole verticality inspection module obtains a cross section from point cloud at the initial position of the POS track, the height of the cross section is the same as the POS track, the cross section is horizontal, and the intersection point of the cross section and the point cloud C is obtained. And respectively calculating the diameter or the size of the point cloud at each intersection point, and calculating the distance between the centers of the point clouds at two adjacent intersection points. The diameter of the telegraph pole is a fixed value, the distance value between the telegraph poles is also a fixed value, interference items are eliminated, the remaining point cloud is obtained, namely the telegraph pole is obtained, the obtained telegraph pole is marked as D, and the coordinates of the center of the telegraph pole are x and y. And (3) starting from the starting point of the POS track, obtaining a cross section at intervals of a set distance, merging the obtained telegraph poles into the D, and circularly calculating to obtain all the telegraph poles D.

Further, the overhead line pole verticality inspection module takes out one telegraph pole Di (xi, yi) in the D, extracts the point cloud C-Di taking Di (xi, yi) as the center of a circle in the point cloud C, sets the point cloud C-Di in a radius range, and calculates the maximum value zmax and the minimum value zmin of the elevation coordinates zi of all points in the point cloud C-Di. And setting a horizontal cross section between the maximum value zmax and the minimum value zmin at intervals of a set distance, solving an intersecting surface of the cross section and the point cloud C-Di, and solving a center coordinate oi of the point cloud of the intersecting surface. After the centers of the intersecting surfaces of all cross sections and the point clouds C-Di are obtained, all the center points are fitted into a straight line, the slope of the straight line is obtained, and the position and the slope of the telegraph pole are recorded.

By implementing the technical scheme of the railway comprehensive inspection system provided by the invention, the railway comprehensive inspection system has the following beneficial effects:

(1) According to the railway comprehensive inspection system, the existing multiple sets of systems on the comprehensive inspection vehicle are replaced by data acquired by a single set of equipment, so that the integration of the detection equipment is greatly improved, and the complexity of the secondary system and the operation and the economic cost are far lower than those of the existing comprehensive inspection system;

(2) According to the railway comprehensive inspection system, the data collected by a single set of equipment can be used by a plurality of subsequent application systems, so that the compatibility of the data among a plurality of detection equipment is improved, the multiplexing of the plurality of equipment and the intercommunication and interconnection of the detection data are realized, and the development complexity of the application system is reduced;

(3) Compared with the traditional detection application based on the plane image, the comprehensive inspection system for the railway can perform qualitative analysis and accurate quantitative analysis, can realize a large number of infrastructure detection functions, can realize disaster analysis and evaluation, and has more abundant functions;

(4) The railway comprehensive inspection system adopts three-dimensional live-action inspection, has true color point cloud and panoramic images, can accurately measure, has great progress compared with the traditional inspection based on images, simultaneously provides a virtual inspection function, greatly reduces the working intensity of inspection workers, and greatly improves the accuracy and automation degree of inspection.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. It is evident that the drawings in the following description are only some embodiments of the invention, from which other embodiments can be obtained for a person skilled in the art without inventive effort.

FIG. 1 is a schematic diagram of an installation structure of a railway integrated inspection system according to an embodiment of the present invention on an integrated inspection vehicle;

FIG. 2 is a front view of a mounting structure of one embodiment of the integrated railroad inspection system of the present invention;

FIG. 3 is a side view of a mounting structure of one embodiment of the integrated railroad inspection system of the present invention;

FIG. 4 is an isometric view of a mounting structure of one embodiment of a railway integrated inspection system of the present invention;

FIG. 5 is a block diagram of the system architecture of one embodiment of the integrated inspection system of the present invention;

FIG. 6 is a flow chart of the operational principle of one embodiment of the integrated railroad inspection system of the present invention;

FIG. 7 is a schematic diagram of a side slope geological disaster inspection flow in accordance with one embodiment of the present invention;

FIG. 8 is a schematic diagram of a slope landslide hazard assessment flow in one embodiment of the railway comprehensive inspection system of the invention;

FIG. 9 is a schematic diagram of a circuit virtual inspection/multiplication process of one embodiment of the integrated inspection system of the present invention;

FIG. 10 is a schematic diagram of an acoustic-wind barrier inspection flow for one embodiment of the integrated railroad inspection system of the present invention;

FIG. 11 is a schematic diagram of a pole verticality inspection flow in one embodiment of a railway integrated inspection system of the present invention;

FIG. 12 is a schematic diagram of a track geometry detection process for one embodiment of the integrated inspection system of the present invention;

FIG. 13 is a schematic view of a track demarcation and profile inspection process for one embodiment of the integrated railway inspection system of the present invention;

fig. 14 is a schematic diagram of a geometric parameter inspection flow of a catenary in a specific embodiment of the railway comprehensive inspection system of the present invention;

FIG. 15 is a schematic diagram of a line spacing inspection process for one embodiment of the railway integrated inspection system of the present invention;

FIG. 16 is a schematic view of a ballast fullness inspection flow of one embodiment of a railway integrated inspection system of the present invention;

FIG. 17 is a flow chart of an operational procedure of one embodiment of a method for integrated inspection of railways based on the system of the present invention;

in the figure: 1-three-dimensional scanning system, 10-attitude positioning system, 11-laser radar, 12-panoramic camera, 13-GNSS equipment, 14-IMU equipment, 15-DMI equipment, 2-space-time synchronization system, 3-storage/control system, 4-inspection application system, 40-point cloud processing system, 41-comprehensive inspection system, 42-side slope disaster inspection module, 43-side slope disaster assessment module, 44-line virtual inspection/multiplication module, 45-track geometry detection module, 46-track limit and contour inspection module, 47-sound wind barrier inspection module, 48-overhead line geometry inspection module, 49-line interval inspection module, 410-track plumpness inspection module, 411-wire pole verticality inspection module, 100-railway comprehensive inspection vehicle, 101-vehicle body, 102-roof platform, 103-host cable, 104-control box, 105-encoder signal input, 106-power line, 107-power supply, 108-air switch, 109-alternating current input, 110-network line, 111-control terminal.

Detailed Description

For purposes of reference and clarity, technical terms, abbreviations or abbreviations used hereinafter are described as follows:

and (3) point cloud data: the scan data is recorded in the form of dots, each dot comprising three-dimensional coordinates, some possibly containing color information (RGB) or reflectance Intensity information (Intensity); in addition to the geometrical positions, the point cloud data has color information, wherein the color information usually obtains a color image through a camera, and then the color information (RGB) of the pixels at the corresponding positions is endowed to the corresponding points in the point cloud; the acquisition of intensity information is the echo intensity acquired by a laser scanner receiving device, and the intensity information is usually related to the surface material, roughness, incident angle direction of a target and the emission energy of the instrument, and the laser wavelength;

distribution algorithm: a point cloud ground point filtering (Cloth Simulation Filter, CSF) algorithm;

and (3) GNSS: global Navigation Satellite System, short for global navigation satellite system;

IMU: inertial Measurement Unit inertial measurement unit abbreviation;

DMI: distance Measuring Instrument, short for distance measuring instrument;

POS: position and Orientation System, a positioning and attitude determination system for short, is an IMU/GPS combined high-precision position and attitude measurement system;

Internal work: mapping terminology, meaning working indoors;

field operation: mapping terminology refers to working outdoors.

In order to make the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. It will be apparent that the described embodiments are only some, but not all, embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

Referring now to fig. 1 to 17, there is shown an embodiment of the integrated railway inspection system according to the present invention, and the present invention will be further described with reference to the accompanying drawings and the embodiment.

As shown in fig. 1 to 4, in an embodiment of a railway integrated patrol car 100 on which the railway integrated patrol system of the present invention is based, a three-dimensional scanning system 1 is installed as an important device on a car body 101 of the railway integrated patrol car 100. The embodiment of the invention provides a railway comprehensive inspection system based on a three-dimensional (mobile) scanning system 1, which takes the three-dimensional (mobile) scanning system 1 as a data acquisition sensor, carries out subsequent railway infrastructure detection work on rich data acquired by the three-dimensional scanning system 1, can replace most of detection systems in the existing railway comprehensive inspection vehicle, provides detection or other functions which are not possessed by the original detection systems, improves the operation efficiency of the comprehensive inspection system, reduces the complexity of the system and greatly saves the labor cost. The three-dimensional (mobile) scanning system 1 integrates a high-precision laser radar 11, a high-resolution panoramic camera 12, a GNSS device 13, an inertial navigation device (i.e. an IMU device 14) and a mileage encoder (i.e. a DMI device 15), can completely replace the detection device of the traditional railway comprehensive inspection vehicle, can acquire complete three-dimensional point cloud data, panoramic image data, mileage data, GNSS data and inertial navigation data of a railway track by singly using the three-dimensional scanning system 1 for data acquisition, and the point cloud processing system 40 manages the acquired data, automatically analyzes and detects defects and manually checks the defects.

As shown in fig. 5, an embodiment of the integrated railway inspection system of the present invention specifically includes: the three-dimensional scanning system 1, the space-time synchronization system 2, the storage/control system 3 and the inspection application system 4, wherein the three-dimensional scanning system 1 comprises a gesture positioning system 10, a laser radar 11 and a panoramic camera 12, and the inspection application system 4 comprises a point cloud processing system 40 and a comprehensive inspection analysis system 41. The attitude determination positioning system 10 acquires GNSS, IMU and DMI data of the railway comprehensive inspection vehicle, the laser radar 11 acquires three-dimensional coordinate data of an object to be inspected, and the panoramic camera 12 acquires panoramic image data (specifically, panoramic photograph and panoramic image data). The space-time synchronization system 2 provides clock synchronization for three-dimensional coordinates, IMU, DMI and panoramic image data according to GNSS time service. The storage/control system 3 stores the three-dimensional coordinates, IMU, DMI and panoramic image data after clock synchronization in real time. The point cloud processing system 40 first uses GNSS, IMU and DMI data to fuse and calculate POS data, then uses POS data and lidar data to jointly calculate point cloud data, and then fuses the point cloud data with panoramic image data to obtain a real-scene point cloud. The comprehensive inspection analysis system 41 performs railway inspection analysis and railway infrastructure detection analysis by using the point cloud and panoramic image data as data sources, as shown in fig. 6. The functions of each device in the railway comprehensive inspection system are described in detail below.

The hardware of the whole railway comprehensive inspection system is mainly divided into three parts: the system comprises an off-board three-dimensional (mobile) scanning system 1, an on-board storage/control system 2, a ground point cloud processing system 40 and a comprehensive inspection analysis system 41, wherein the ground comprehensive inspection analysis system 41 takes data acquired by the three-dimensional scanning system 1 as input, and is provided with different detection (software) functional modules aiming at different detection item points. The three-dimensional (mobile) scanning system 1 integrates two laser scanners (i.e. lidar 11), a panoramic camera 12, a GNSS device 13, a high-precision IMU device 14, a DMI device 15 and other electrical devices.

Laser radar 11: the sensor is used for acquiring the local three-dimensional coordinates of the surface of the scanned object at a high speed, can perform laser scanning in 360-degree rotation, and acquires the three-dimensional coordinates with the laser radar phase center as the origin of coordinates at a speed of 100 ten thousand points per second with high precision. In this embodiment, as a z+f9012 laser radar may be used, the point frequency of the radar reaches 101.6 ten thousand points/second, the rotation speed reaches 200 circles/second, the ranging range reaches 190 meters, and the ranging precision is less than 2mm, which is the laser radar with the highest ranging precision under the same point frequency condition in the present stage. Because the railway detection has higher requirements on various indexes of data, two Z+F9012 laser radars can be simultaneously carried in the embodiment, and laser planes are crossed at 90 degrees, so that higher point cloud density is realized, and shielding of cylindrical objects is reduced.

Panoramic camera 12: panoramic photos are acquired at a fixed frequency, and a ladybug5Plus camera can be specifically adopted, the resolution of the panoramic photos output by the camera is 3000 ten thousand pixels, and the frame rate can reach 5 frames/second.

Positioning and attitude determination system 10: further comprising a GNSS device 13, an IMU device 14 (i.e. inertial navigation device) and a DMI device 15 (i.e. mileage encoder). The GNSS device 13 is a global navigation satellite system, and obtains absolute geodetic coordinate information and accurate clock information at a frequency of 20Hz, and is configured to provide real-time geodetic coordinate data and accurate time-service clock source for the system. The IMU device 14 is an inertial navigation unit, calculates triaxial acceleration and pose data of the device at a frequency of 200Hz, and is used for providing real-time pose data for the system, and is combined with GNSS to form a positioning and pose-determining system, so as to provide real-time position and pose data for the system. The DMI device 15 is a mileage encoder, and is installed on a rotating shaft of the wheel set, for providing mileage information.

The sensor is installed on the railway comprehensive inspection vehicle 100 according to a designed rigid structure, and high-precision three-dimensional geometric information and image texture information within the range of 50 meters of the central line of the railway line can be synchronously acquired along with the advancing of the vehicle body 101. The GNSS device 13, the IMU device 14 and the DMI device 15 together form the positioning and attitude determination system 10 of the railway comprehensive inspection system, and the IE (Inertial Explorer) post-processing software is adopted to jointly calculate three navigation data, so that the corresponding attitude and position can be obtained. The laser scanner (i.e. the laser radar 11) with the help of these three positioning and attitude determination sensors enables the system to acquire absolute three-dimensional geometric information, while the panoramic camera 12 can acquire image texture information. And fusing the acquired laser scanner data with the panoramic image data to obtain three-dimensional point cloud data comprising color information.

Space-time synchronization system 2: the method is used for receiving the time service of the GNSS, marking uniform high-precision time stamps on all the sensor data in real time, and providing a high-precision clock synchronization basis for the subsequent internal processing data.

Storage/control system 3: the system comprises a plurality of sensors with very large data volume, so that a special storage system is provided for storing the original data of each sensor after clock synchronization in real time. Meanwhile, the system can also control the operation of the whole railway comprehensive inspection system.

Point cloud processing system 40: the system takes the original data in the storage/control system 3 as input, outputs the whole three-dimensional point cloud based on the geodetic coordinates after internal processing, and realizes the registration of panoramic images (photos) and the point cloud. The three-dimensional point cloud comprises high-precision absolute coordinate information and relative coordinate information, a gray level image is generated by utilizing the reflection intensity of the laser point, and a true color image is generated by utilizing the panoramic camera. Therefore, the data acquired by the three-dimensional mobile measurement system (namely the three-dimensional scanning system 1) not only contains accurate size data, but also contains rich image data, and can be used for comprehensive inspection of railways.

The point cloud processing system 40 first calculates POS data using GNSS, IMU, and DMI data fusion, and then jointly calculates point cloud data using POS data and lidar data. After POS data are obtained, pose data are calculated for each panoramic photo, and after pose information of the panoramic photo is obtained, the panoramic photo can be sleeved with the point cloud for joint use.

In the embodiment of the present invention, the data output by the lidar 11 and the panoramic camera 12 in unit time are very large, so the data collected in real time is first stored in the storage/control system 3 in the vehicle, and then dumped off-line to the comprehensive inspection (processing) system 41 at the ground end by means of a mobile hard disk and the like for unified processing. The three-dimensional point cloud includes high-precision absolute coordinate information and relative coordinate information, and generates a gray scale map by using the reflection intensity of the laser point and a true color map by using the panoramic camera 12. Therefore, the data acquired by the three-dimensional mobile measurement system (namely the three-dimensional scanning system 1) not only contains accurate size data, but also contains rich image data, and can be used for realizing the comprehensive inspection function of various railway lines, and specific functional item points are shown in fig. 5.

The three-dimensional scanning system 1 comprises a gesture positioning system 10, a laser radar 11 and a panoramic camera 12, wherein the gesture positioning system 10 further comprises a GNSS device 13, an IMU device 14 and a DMI device 15. In order to ensure the non-shielding performance of the laser radar and the image acquisition performance of the panoramic camera, the laser radar 11, the panoramic camera 12, the GNSS device 13 and the IMU device 14 of the three-dimensional scanning system 1 are arranged on a roof platform 102 at the head or tail of a railway comprehensive inspection vehicle body 101, and a mileage encoder (namely, a DMI device 15) is arranged on a wheel shaft. The inside of the vehicle body 101 is provided with a control box 104 (in which the storage/control system 3 is installed), a power supply 107, a control terminal 111, and the like, the three-dimensional scanning system 1 is connected to the control box 104 through a host cable 103, and the control box 104 is connected to the control terminal 111 through a network cable 110. The encoder signal input 105 is connected to the control box 104, and the ac input 109 is connected to a power supply 107 (a UPS uninterruptible power supply may be specifically used) through an air switch 108, and then supplies power to the control box 104 through a power line 106.

The operation process of comprehensive inspection of railway lines by using the system described by the embodiment of the invention is divided into field original data acquisition and field data processing, and the specific flow is shown in figure 6. The system firstly collects the original data of sensors such as GNSS, IMU, laser scanner, panoramic camera, DMI and the like through field operation, saves the data and carries out simple data check, and when in field operation, the original data collected by field operation is transferred to a point cloud processing system 40 at the ground end through means such as mobile hard disk transfer or network and the like to generate a three-dimensional point cloud and panoramic image. And the civil engineering data platform firstly uses GNSS, IMU and DMI data fusion to calculate POS data, and then uses the POS data and the laser radar data to jointly calculate point cloud data. And fusing the point cloud data and the panoramic photo to obtain true color real scenery point cloud. Finally, the comprehensive inspection analysis system 41 performs various railway inspection analysis and infrastructure detection analysis by using the point cloud and the panoramic photo as data sources.

The comprehensive inspection analysis system 41 includes a slope disaster inspection module 42, a plurality of laser mark targets (mark targets for short) are arranged on the railway slope to be inspected, the three-dimensional scanning system 1 performs raw data acquisition during field operation, and the point cloud processing system 40 generates point cloud and panoramic image data during field operation. As shown in fig. 7, the slope disaster inspection module 42 precisely extracts laser mark targets from the scanned point cloud data, and sets unique numbers or names for the extracted laser mark targets, wherein the numbers or names of the laser mark targets at the same position in the point cloud obtained by different period scanning are the same. And acquiring the current-period point cloud, reading in the to-be-compared previous-period point cloud, performing coarse superposition display on the two-period point cloud by using the mileage and coordinates of the laser mark target, and calculating accurate point cloud superposition parameters by using the coordinates of the previous-period laser mark target and the coordinates of the current-period laser mark target. And carrying out coordinate conversion on the current-period coordinates by using accurate point cloud superposition parameters, carrying out accurate point cloud superposition comparison after conversion, and calculating coordinate differences of the same positions in the two-period point clouds after superposition. And (3) checking the change amount, and taking panoramic images in the current period and the earlier period for the region with large coordinate difference, and manually confirming whether slope change occurs. And (3) storing the coordinate difference of the point cloud superposition contrast, reading in the point cloud contrast coordinate difference in the earlier stage, drawing a coordinate difference change curve, detecting whether the coordinate difference has a trend of gradually amplifying, and carrying out early warning or alarming on an area where the coordinate difference is gradually amplified. The slope disaster inspection module 42 performs plane fitting on the point cloud grid C processed first to obtain a plane P, and obtains a center coordinate O of the point cloud grid C. And then, obtaining the normal line of the plane P passing through the point O, obtaining the normal line L, obtaining the intersection point J of the normal line L and the earlier point cloud, and calculating the distance of the line segment OJ, namely the variation of the same position in the two-stage point cloud. And traversing each grid in the current-period point cloud, and calculating and obtaining the front and rear period variation of each grid point cloud, thereby obtaining the coordinate difference of the same position in the two-period point cloud.

The comprehensive inspection analysis system 41 comprises a slope disaster evaluation module 43, a plurality of laser marking targets are arranged on the railway slope to be inspected, the three-dimensional scanning system 1 collects original data during field operation, and the point cloud processing system 40 generates point cloud and panoramic image data during field operation. The side slope disaster evaluation module 43 extracts laser mark targets from the scanned point cloud data, and sets unique numbers or names for the extracted laser mark targets, wherein the numbers or names of the laser mark targets at the same position in the point clouds at different periods are the same. And acquiring the current-period point cloud, reading in the to-be-compared previous-period point cloud, performing coarse superposition display of the two-period point cloud by using the mileage and coordinates of the laser mark targets, and calculating accurate point cloud superposition parameters by using the coordinates of the previous-period laser mark targets and the current-period laser mark targets. And carrying out coordinate conversion on the current-period coordinates by using accurate superposition parameters, carrying out accurate point cloud superposition comparison after conversion, calculating the point cloud coordinate difference of the same position in the two-period point clouds after superposition, and focusing on the difference value of the elevation coordinate z. And calculating the volume of the part clamped by the two point clouds by utilizing the coordinate difference of the two point clouds, namely, the earth and stone side of the landslide, and calculating the area of the landslide part. And (3) taking panoramic image data scanned in the previous period, checking actual topography and topography of a disaster point, intercepting and storing the three-dimensional point cloud according to mileage, positioning the three-dimensional point cloud to the mileage of an emergency event, and taking the three-dimensional point cloud and the panoramic image at the position as a base map of emergency treatment to provide map reference for the emergency treatment.

The comprehensive inspection analysis system 41 comprises a line virtual inspection/multiplication module 44, the three-dimensional scanning system 1 performs raw data acquisition during field operation, and the point cloud processing system 40 generates point cloud, panoramic image and POS track file during field operation. The line virtual inspection/multiplication module 44 uses the calibrated panoramic image to assign a true color value to the gray point cloud, generates a true color point cloud, divides the true color point cloud into point cloud segments, and simultaneously divides the POS track file into POS track file segments corresponding to the point cloud segments. And reading in a point cloud segment of the mileage to be inspected and a corresponding POS track file. The line virtual inspection/multiplication module 44 displays that the default viewpoint of the interface is positioned at the starting point of the POS track file, the view field direction is consistent with the POS track file, and the user can perform point cloud roaming in various modes after clicking to start, so that the visual effect of virtual inspection/multiplication is realized. In the point cloud roaming process, a user can change the view field direction at any time and pause the movement of the point cloud, after the movement of the point cloud is paused, the user can check a specific part through zooming, and the image information of the panoramic image checking part corresponding to the point cloud is called out. If a component has a defect or failure, the line virtual inspection/multiplication module 44 outputs defect information by creating a defect report. And when the inspection of the point cloud of one section is finished, automatically reading the point cloud of the next section for inspection until the inspection of the line is finished.

The comprehensive inspection analysis system 41 comprises an acoustic-wind barrier inspection module 47, a plurality of laser marking targets are arranged on a railway barrier to be inspected, the three-dimensional scanning system 1 performs original data acquisition during field operation, and the point cloud processing system 40 generates point cloud, POS track files and panoramic image data during field operation. The sound-wind barrier inspection module 47 manually intercepts the point cloud of the road section only including the barrier, if the road section is long, the point cloud is divided into point cloud sections, and meanwhile, the POS track file is divided into POS track file sections corresponding to the point cloud sections, and the POS track file is matched with the point cloud. Reading a point cloud segment of a mileage to be inspected and a corresponding POS track file, separating the point cloud of an aerial part to obtain a ground point cloud, and dividing the ground point cloud to separate the point cloud of a barrier part. And extracting point clouds C taking the POS track line as a central line, setting the width range at two sides of the point cloud C, performing high-level filtering processing on the point clouds C, removing the point clouds with the elevation coordinate z value smaller than the set value to obtain residual point clouds C ', calculating the point cloud normal line of the C', removing the point clouds which are not perpendicular to the POS track line in the point clouds, and finally obtaining the residual point clouds CP only comprising the barrier part. Extracting laser mark targets from the point cloud CP, and setting unique numbers or names for the extracted laser mark targets, wherein the numbers or names of the laser mark targets at the same position in the point cloud at different periods are the same. And reading in the early-stage point clouds to be compared, performing coarse superposition display of the two-stage point clouds by using the mileage and coordinates of the laser marking targets, and calculating accurate point cloud superposition parameters by using the coordinates of the early-stage laser marking targets and the coordinates of the laser marking targets in the current stage. And carrying out coordinate conversion on the current-period coordinates by using accurate superposition parameters, carrying out accurate point cloud superposition comparison after conversion, traversing the current-period point clouds after superposition, counting the number of the point clouds of each point cloud within a set radius, and regarding the point clouds of which the number is not 0 in the previous-period point clouds and the number is 0 in the current-period point clouds, namely, considering that the barrier in the current-period point clouds is damaged. And (3) calculating the area of the damaged part by statistics, and storing the panoramic image and the three-dimensional point cloud scanned in the current period at the damaged mileage so as to be convenient for the staff to confirm afterwards.

The comprehensive inspection analysis system 41 comprises a contact net wire pole verticality inspection module 411, wherein the contact net wire pole verticality inspection module 411 performs point cloud segmentation after collecting holographic point clouds around a railway to obtain point clouds of a wire pole part, obtains the point cloud center of the wire pole part, performs linear fitting to obtain a vector equation of the contact net wire pole center, and obtains the verticality of the wire pole. The three-dimensional scanning system 1 performs raw data acquisition during field operations, and the point cloud processing system 40 generates point cloud, POS track file, and panoramic image data during field operations. The overhead line pole verticality inspection module 411 divides the point cloud into point cloud segments, simultaneously divides the POS track file into POS track file segments corresponding to the point cloud segments, and matches the POS track file with the point cloud file. And reading in a point cloud segment of the mileage to be inspected and a corresponding POS track file, deleting redundant point clouds by using the POS track, searching all telegraph poles by using the residual point clouds as C, calculating the position and the slope of each telegraph pole, and storing and outputting a calculation result. The contact net wire pole verticality inspection module 411 obtains a cross section from a point cloud at the beginning of a POS track, and the cross section is the same as the POS track in height and horizontal to obtain an intersection point of the cross section and the point cloud C. And respectively calculating the diameter or the size of the point cloud at each intersection point, and calculating the distance between the centers of the point clouds at two adjacent intersection points. The diameter of the telegraph pole is a fixed value, the distance value between the telegraph poles is also a fixed value, interference items are eliminated, the remaining point cloud is obtained, namely the telegraph pole is obtained, the obtained telegraph pole is marked as D, and the coordinates of the center of the telegraph pole are x and y. And (3) starting from the starting point of the POS track, obtaining a cross section at intervals of a set distance, merging the obtained telegraph poles into the D, and circularly calculating to obtain all the telegraph poles D. The overhead line pole verticality inspection module 411 takes out one wire pole Di (xi, yi) in the D, extracts the point cloud C-Di taking Di (xi, yi) as the center of a circle in the point cloud C, sets the point cloud C-Di in the radius range, and calculates the maximum value zmax and the minimum value zmin of the elevation coordinates zi of all points in the point cloud C-Di. And setting a horizontal cross section between the maximum value zmax and the minimum value zmin at intervals of a set distance, solving an intersecting surface of the cross section and the point cloud C-Di, and solving a center coordinate oi of the point cloud of the intersecting surface. After the centers of the intersecting surfaces of all cross sections and the point clouds C-Di are obtained, all the center points are fitted into a straight line, the slope of the straight line is obtained, and the position and the slope of the telegraph pole are recorded.

The comprehensive inspection analysis system 41 comprises a track geometric parameter detection module 45, the three-dimensional scanning system 1 collects point cloud and panoramic image raw data during field operation, and the point cloud processing system 40 generates point cloud, POS track files and panoramic image data during field operation. As shown in fig. 12, the track geometry parameter detection module 45 segments the generated point cloud and segments the POS track file into file segments corresponding to the point cloud segments. Reading a section of point cloud, separating to obtain a ground point cloud, reading a POS track file corresponding to the section of point cloud, removing the point cloud which takes the POS track as the center and has a distance exceeding a set range, and obtaining the rest point cloud which is the point cloud only comprising the track bed. In the ground point cloud, the point cloud of the steel rail part is obtained by utilizing elevation difference, the point cloud of the steel rail part obtained by separation is utilized to carry out steel rail top surface fitting and extraction, the left and right steel rail top surface vector lines (namely the center line of the steel rail top surface along the rail direction) are obtained, and the reliability of the extraction of the steel rail top surface vector lines is checked by utilizing the track gauge range between the steel rails. And carrying out track center line calculation by using the extracted vector lines of the left steel rail and the right steel rail to obtain track center line vectors, setting a distance between vector nodes of the center line vectors, assigning mileage values to the track center line, and unifying the mileage of the track center line with the railway maintenance mileage by setting starting points and end points. And carrying out refinement treatment on the coordinates of the track center line, and outputting the computed vector line of the track center line after refinement treatment for computing the geometric parameters of the line. Wherein the track geometry parameter detection module 45 calculates the cross section of the steel rail passing through the vector node in the track, and the thickness of the cross section does not exceed the set distance. And projecting the point cloud on the cross section relative to the cross section, obtaining two-dimensional coordinates of the cross section of the steel rail after projection, respectively carrying out accurate matching on the cross sections of the left steel rail and the right steel rail and the standard profile of the steel rail, obtaining top surface coordinates of the left steel rail and the right steel rail after matching, and updating the top surface coordinates of the original steel rail by using the top surface coordinates. And calculating accurate track center line coordinates, updating original track center line vector node coordinates, traversing all the track center line vector nodes which are calculated, and finishing the refinement treatment of the track center line coordinates.

The comprehensive inspection analysis system 41 comprises a track limit and contour inspection module 46, the three-dimensional scanning system 1 collects original data of point clouds and panoramic images during field operation, and the point cloud processing system 40 generates the point clouds and the panoramic image data during field operation. As shown in fig. 13, the track limit and profile inspection module 46 segments the generated point cloud, reads a segment of the point cloud, separates the point cloud to obtain a ground point cloud, and obtains the point cloud of the steel rail part in the ground point cloud by using elevation difference. And fitting and extracting the top surfaces of the steel rails by utilizing the point clouds of the separated steel rail parts to obtain left and right top surface vector lines of the steel rails, and checking the reliability of the extraction of the top surface vector lines of the steel rails by utilizing the track gauge range between the steel rails. And calculating a track center line by using the extracted vector lines of the top surfaces of the left steel rail and the right steel rail, obtaining the track center line, assigning mileage values for the track center line, and unifying the mileage of the track center line with the railway maintenance mileage by setting starting points and end points. And calculating the cross section of the point cloud according to the set mileage interval value, outputting all the cross section point clouds, performing limit calculation for each cross section point cloud, and storing and outputting the limit calculation result. The track limit and profile inspection module 46 takes the intersection point of the cross section and the track central line as the origin of coordinates, the x-axis is parallel to the top surface of the steel rail and points to the central line of the right steel rail top surface along the track direction, and the z-axis is perpendicular to the x-axis to establish a limit reference coordinate system. And projecting the cross section point cloud relative to the cross section, and obtaining the two-dimensional coordinates of the steel rail cross section point cloud after projection. And converting the coordinates of the cross section point Yun Erwei into a limit reference coordinate system, generating a standard building limit model in the coordinate system, performing superposition collision analysis on the cross section point cloud and the standard building limit model, and obtaining and storing a limit calculation result.

The comprehensive inspection analysis system 41 comprises a contact net geometric parameter inspection module 48, the three-dimensional scanning system 1 collects original data of point clouds and panoramic images during field operation, and the point cloud processing system 40 generates point clouds, POS track files and panoramic image data during field operation. As shown in fig. 14, the catenary geometric parameter inspection module 48 segments the generated point cloud and segments the POS track file into file segments corresponding to the point cloud segments. And reading in a section of point cloud, reading in a POS track file corresponding to the point cloud section, removing the point cloud which takes the POS track as the center and is more than a set range from the POS track, wherein the rest point cloud only comprises a track bed and the point cloud Ce above the track bed. And extracting a point cloud Cf-d of the ground part in the point cloud Ce, extracting the point cloud Cf-d from the point cloud Ce to obtain a point cloud Cf-k of the air part, and separating the point cloud Cf-d of the ground part by utilizing elevation difference to obtain a point cloud of the steel rail part. And fitting and extracting the top surfaces of the steel rails by utilizing the point clouds of the separated steel rail parts to obtain left and right top surface vector lines of the steel rails, and checking the reliability of the extraction of the top surface vector lines of the steel rails by utilizing the track gauge range between the steel rails. And carrying out track center line calculation by using the extracted left and right steel rail top surface vector lines, and obtaining track center line vectors, wherein the vector node interval of the center line vectors is a set distance. And (3) giving a mileage value to the track center line, and unifying the mileage of the track center line with the railway maintenance mileage by setting a starting point mileage and an ending point mileage. And identifying and extracting contact network lines in the aerial point cloud Cf-k, and performing vector line fitting, wherein the distance between vector nodes is a set value. And calculating the geometric parameters of the adjacent contact network wires, and storing and outputting the geometric parameters of the contact network wires. The geometric parameter inspection module 48 of the overhead line system traverses all vector nodes of the central line of the top surface of the left steel rail in the railway steel rail, calculates the direction vector of each vector node A, and generates a normal plane F perpendicular to the direction vector. Traversing vector nodes of the middle line of the top surface of the right steel rail, searching two adjacent vector nodes C and D which are positioned at two sides of a normal plane F, and calculating an intersection point B of the normal plane F and a CD connecting line. Traversing vector nodes contacting with the network wire, searching two adjacent vector nodes E and F positioned on two sides of the normal plane F, and calculating an intersection point G of the normal plane F and EF connecting lines. The perpendicular line of the line segment AB is made through the point G, the drop foot is marked as a point H, the middle point of the line segment AB is marked as an I, the length of the HI connecting line is the contact line pulling-out value of the mileage corresponding to the point I, and the length of the GH is the contact line height guiding value of the mileage corresponding to the point I, so that the geometric parameters of the adjacent contact network lines are obtained.

The comprehensive inspection analysis system 41 comprises a line interval inspection module 49, the three-dimensional scanning system 1 collects original data of point clouds and panoramic images during field operation, and the point cloud processing system 40 generates the point clouds and the panoramic image data during field operation. As shown in fig. 15, the line-to-line inspection module 49 segments the generated point cloud, reads a segment of the point cloud, separates the point cloud to obtain a ground point cloud, and uses elevation difference separation to obtain a point cloud of the rail portion in the ground point cloud. And fitting and extracting the top surfaces of the steel rails by utilizing the point clouds of the separated steel rail parts to obtain left and right top surface vector lines of the steel rails, and checking the reliability of the extraction of the top surface vector lines of the steel rails by utilizing the track gauge range between the steel rails. And calculating a track center line by using the extracted vector lines of the top surfaces of the left steel rail and the right steel rail, obtaining the track center line, assigning mileage values for the track center line, and unifying the mileage of the track center line with the railway maintenance mileage by setting starting points and end points. And calculating the line spacing of the lines in two adjacent tracks, and storing and outputting the line spacing. The line-to-line inspection module 49 traverses all vector nodes in a line in one of the tracks, calculates a direction vector of each vector node a, and generates a normal plane perpendicular to the direction vector. Traversing vector nodes of the central lines of adjacent tracks, searching two adjacent vector nodes C and D which are positioned on two sides of the normal plane, and calculating an intersection point B of the normal plane and a CD connecting line. Calculating the horizontal distance of the AB, wherein the horizontal distance of the AB is the line spacing at the point, and recording the mileage of the point A, wherein the mileage of the point A is the mileage value of the line spacing.

The comprehensive inspection analysis system 41 comprises a ballast plumpness inspection module 410, the three-dimensional scanning system 1 performs original data acquisition during field operation, and the point cloud processing system 40 generates point cloud, POS track files and panoramic image data during field operation. As shown in fig. 16, the ballast plumpness inspection module 410 divides the point cloud into point cloud segments, divides the POS track file into POS track file segments corresponding to the point cloud segments, and matches the POS track file with the point cloud file. Reading in a point cloud segment of mileage to be patrolled and examined and a corresponding POS track file, separating to obtain a ground point cloud, removing the point cloud of an aerial part, extracting the point cloud within a range of width by taking a POS track line as a central line and setting the two sides respectively, wherein the extracted point cloud is the point cloud only comprising a ballast bed part. In the point cloud only comprising the track bed part, the point cloud of the steel rail part is obtained by utilizing elevation difference, the point cloud of the separated steel rail part is utilized to carry out steel rail top surface fitting and extraction, the left and right steel rail top surface vector lines are obtained, and the reliability of the extraction of the steel rail top surface vector lines can be checked by utilizing the track gauge range between the steel rails. And calculating a track center line by using the extracted vector lines of the top surfaces of the left steel rail and the right steel rail to obtain the track center line, assigning mileage values to the track center line, and unifying the mileage of the track center line with the railway maintenance mileage by setting starting points and end points. Calculating the cross section of the point cloud according to the set mileage interval value, outputting all cross section point clouds, calculating the ballasting plumpness aiming at each cross section point cloud, accumulating after calculating the ballasting plumpness aiming at each frame point cloud of a continuous section of line, obtaining the ballasting plumpness of the whole section of line, and storing and outputting the result of limit calculation. The ballast plumpness inspection module 410 projects the point cloud in the cross section to form a two-dimensional point cloud C. The rail is positioned in the cross section by calculating the intersection point of the vector line of the top surface of the rail and the cross section, and the position of the ballast slope is positioned in the cross section by utilizing the relative position relation between the ballast slope and the rail. And establishing a standard coordinate system by taking the connecting line of the central lines of the top surfaces of the left steel rail and the right steel rail as an x axis, taking the midpoint of the connecting line as a coordinate origin, taking the vertical connecting line upwards as a y axis, and rotationally translating the two-dimensional point cloud C to the standard coordinate system. Generating a ballast bed standard cross section model in a standard system, calculating the areas of the point cloud and the graph enclosed by the standard model one by one, wherein the unnecessary ballasts above the standard model are the unnecessary ballasts below the standard model, and the ballasts are missing, so that the ballasts plumpness of the section can be obtained after calculation is finished.

As shown in fig. 17, an embodiment of a comprehensive railway inspection method based on the system of the invention specifically comprises the following steps:

s1) acquiring GNSS (global navigation satellite system), IMU (inertial measurement unit) and DMI (digital mobile camera) data of a railway comprehensive inspection vehicle by using a gesture positioning system 10, acquiring three-dimensional coordinate data of an object to be detected by using a laser radar 11, and acquiring panoramic image data (specifically, panoramic photo and panoramic image data) by using a panoramic camera 12;

s2) a space-time synchronization system 2 provides clock synchronization for three-dimensional coordinates, IMU, DMI and panoramic image data according to GNSS time service;

s3) the storage/control system 3 stores the three-dimensional coordinates, the IMU, the DMI and the panoramic image data after clock synchronization in real time;

s4) the point cloud processing system 40 firstly uses GNSS, IMU and DMI data to fuse and calculate POS data, then uses the POS data and laser radar data to jointly calculate point cloud data, and then fuses and processes the point cloud data and panoramic image data to obtain a real-scene point cloud;

s5), the comprehensive inspection system 41 uses the point cloud and the panoramic image data as data sources to conduct railway inspection analysis and railway infrastructure detection analysis.

The railway inspection analysis specifically comprises geological disaster inspection (namely side slope disaster inspection) of side slope landslide and the like, side slope (geological) disaster evaluation, virtual line inspection/multiplication, sound and wind barrier inspection and overhead line pole perpendicularity inspection. The railway infrastructure detection analysis specifically comprises track geometric parameter detection, track limit and contour inspection, contact net geometric parameter inspection, line interval inspection and railway ballast plumpness inspection. The steps are described in detail below.

1) Geological disaster inspection for slope landslide and the like

The side slope of the railway is sometimes subjected to geological disasters such as side slope landslide and the like due to rain wash or other factors, the occurrence of the geological disasters has very great influence on the operation of the railway, the railway is interrupted if the light side slope is used, and the casualties are caused if the heavy side slope is used, so that the inspection and early warning of the geological disasters such as the side slope landslide of the railway are very important. The railway comprehensive inspection system uses a three-dimensional (mobile) scanning system 1 to periodically scan a side slope area along a railway, uniformly manages point clouds obtained each time, and can finely compare the point clouds because the point clouds obtained in multiple periods have the same mileage and coordinate system, find the changing place and the changing quantity of the side slope according to the comparison result of the point clouds, and early warn the landslide according to the change place and the changing quantity. As shown in fig. 7, the specific implementation steps are as follows:

s100) in order to enable the positioning effect of point cloud comparison to be better, a plurality of laser marking targets with very stable positions are arranged on a railway slope to be inspected before three-dimensional scanning of the slope is carried out, the adjacent laser marking targets are arranged in a Z shape, the horizontal distance between every two laser marking targets is about 100 meters, and the vertical distance is more than or equal to 10 meters; during field operation, acquiring raw data of sensors such as a laser radar 11, a panoramic camera 12, a GNSS device 13, an IMU device 14, a DMI device 15 and the like; during the field operation, generating point cloud and panoramic image data (namely panoramic photo and panoramic image); precisely extracting laser mark targets from the point cloud obtained by scanning, setting unique numbers or names for the extracted laser mark targets, wherein the numbers or names of the laser mark targets at the same positions in the point cloud obtained by scanning in different periods are the same;

S101) acquiring the point cloud of the current period, and then reading in the earlier period point cloud to be compared;

s102) performing coarse superposition display on two-stage point clouds by using mileage and coordinates of a laser mark target;

s103), calculating accurate point cloud superposition parameters, namely seven parameters of point cloud coordinate conversion, by utilizing the earlier-stage laser mark target coordinates and the current-stage laser mark target coordinates;

s104) carrying out coordinate conversion on the current coordinate by utilizing the accurate point cloud superposition parameters, and carrying out accurate point cloud superposition comparison after conversion;

s105), after superposition, calculating the point cloud coordinate difference, namely the variation, of the same position in the two-stage point clouds, wherein the specific steps are as follows:

s1051) first meshing the current point cloud into a mesh of about 10cm x 10 cm;

s1052) traversing each grid in the current-period point cloud, assuming that the first-processed point cloud grid is C;

s1053) performing plane fitting on the point cloud C to obtain a plane P, and solving the center coordinate O of the point cloud grid C;

s1054) obtaining a normal line of a plane P passing through the point O, and obtaining a normal line L;

s1055), obtaining an intersection point J of a normal L and the previous point cloud, and calculating the distance of a line segment OJ, namely the variation of the same position in the two-stage point cloud;

s1056) repeating the steps for each grid in the grids to obtain the front-and-back period variation of each grid point cloud;

S106) checking the result of the variation, and taking panoramic image data of the current period and the earlier period for the region with larger coordinate difference to manually confirm whether the slope change is actually generated;

s107) saving the coordinate difference of the point cloud superposition contrast;

s108) reading in the point cloud contrast coordinate difference in the earlier stage, drawing a coordinate difference change curve, and detecting whether the coordinate difference has a trend of gradual amplification;

s109) early warning or alarming is performed on the region where the coordinate difference appears to be gradually enlarged.

2) Disaster evaluation such as slope landslide

Railway side slopes have the phenomenon of geological disasters such as side slope landslide in rainy seasons. When geological disasters such as landslide occur, the size of the disasters needs to be evaluated in the first time, and a basis is provided for the processing of the disasters. After an emergency occurs, the three-dimensional scanning system 1 can be used for rapidly scanning the line again, and compared with the data in the earlier stage, the landslide earth and stone can be immediately judged, the size of the disaster is estimated, and accurate and quantitative data basis is provided for emergency treatment. And the three-dimensional point cloud and the panoramic image obtained by scanning can be used as a base map during disaster treatment, so that powerful map support is provided for disaster treatment. As shown in fig. 8, the specific implementation steps are as follows:

S200) in order to enable the positioning effect of point cloud comparison to be better, a plurality of laser marking targets with very stable positions are arranged on a railway slope to be inspected before three-dimensional scanning of the slope is carried out, adjacent targets are arranged in a Z shape, the horizontal distance between every two marking targets is about 100 meters, and the vertical distance is more than or equal to 10 meters; when in field operation, acquiring sensor raw data of point cloud and panoramic photo; when in the field operation, generating point cloud and panoramic photos; extracting laser mark targets from the scanned point clouds, and setting unique numbers or names for the extracted laser mark targets, wherein the numbers or names of the laser mark targets at the same position in the point clouds at different periods are the same;

s201), acquiring the point cloud of the current period, and then reading in the earlier period point cloud to be compared;

s202) performing coarse superposition display of two-stage point clouds by using mileage and coordinates of a laser mark target;

s203), calculating accurate point cloud superposition parameters by utilizing the earlier-stage laser mark target coordinates and the current-stage laser mark target coordinates;

s204) carrying out coordinate conversion on the current coordinate by utilizing the accurate superposition parameters, and carrying out accurate point cloud superposition comparison after conversion;

s205) after superposition, calculating the coordinate difference of the point clouds at the same position in the two-stage point clouds, and emphasizing the difference value of the elevation coordinate z;

S206) calculating the volume of the part contained by the two point clouds by utilizing the coordinate difference of the two point clouds, namely the landslide earth and stone;

s207) statistically calculating an area of the landslide section;

s208) the panoramic image data scanned in the previous period is called, the actual topography and the relief of the disaster point are checked, and the three-dimensional point cloud is intercepted and stored according to mileage;

s209) simultaneously locating the mileage of the emergency event, and using the three-dimensional point cloud and the panoramic image of the position as the base map of the emergency treatment, thereby providing the latest and truest map reference for the emergency treatment.

3) Virtual line inspection/addition

For railway inspection, the current common practice in the industry is to manually inspect the railway at regular intervals, and the quality of manual inspection depends on the working attitude and professional skills of workers. When an experienced patrol worker patrols, the quality and the efficiency of the patrol are guaranteed, but if the professional quality of the patrol worker is not high, the problems of low quality, low efficiency and high omission factor exist, the working strength of the manual patrol is very high, and the working mode is very backward.

The three-dimensional scanning system 1 can acquire high-precision point cloud and panoramic photo data within the range of 50 meters of a railway, and the point cloud and the image after nesting are very suitable for operators to watch, and also have the characteristic of measurement, so that the three-dimensional scanning system is very suitable for virtual line inspection, and can greatly reduce the workload of field industry. Meanwhile, at present, the requirement of regular multiplication is met in each office, the three-dimensional point cloud and the panoramic image generated by the three-dimensional scanning system 1 are very detailed, virtual reality VR display can be carried out after the point cloud and the panoramic photo which are sleeved are subjected to virtual reality processing, and virtual multiplication is carried out by simulating the visual angle of a locomotive driver. As shown in fig. 9, the specific implementation steps are as follows:

S300) acquiring original data of each sensor by field operation, and generating point cloud, panoramic image, POS track file and the like by the inner industry;

s301) utilizing the calibrated panoramic image to endow a true color value for the gray point cloud to generate a true color point cloud (the true color point cloud used by railway inspection/virtual multiplication can increase the presence);

s302) because the point cloud is in mass level, in order to inspect the fluency of browsing the point cloud, the point cloud needs to be segmented, and the point cloud is best segmented into point cloud segments with the size of 1GB-3GB according to the general performance of the current processor;

s303) synchronously dividing the POS track file into POS track file segments corresponding to the point cloud segments;

s304) reading in a point cloud segment of mileage to be inspected and a corresponding POS track file;

s305) positioning a default viewpoint of the comprehensive inspection system interface at the starting point of the POS track file, wherein the view field direction is consistent with the POS track file, and after the user clicks, performing point cloud roaming (automatic roaming or free roaming can be adopted) in various modes to realize the visual effect of addition inspection/virtual;

wherein, automatic roaming: after the user sets the roaming direction (direction: from large mileage to small mileage or from small mileage to large mileage), the system automatically performs point cloud browsing at a certain speed according to the set direction, and the automatic roaming speed is adjustable; free roaming: free roaming is divided into two types, (1) full free roaming, the roaming direction and speed are all determined by a user through a keyboard and a mouse, the viewpoint and the view field are irrelevant to POS track files during full free roaming, and the viewpoint and the view field are determined by the user; (2) Step roaming is similar to automatic roaming, but the inspection system does not automatically move the point cloud during step roaming, but the user browses forwards or backwards through a direction control button, and the point cloud moves a distance when clicking the direction control button once;

S306) in the process of point cloud roaming, a user can change the view field direction at any time and pause the point cloud movement;

s307) after suspending the point cloud movement, the user can carefully check the specific component by zooming, and can adjust out the image information of the panoramic photo checking component corresponding to the point cloud;

s308), if a certain part has a defect or a fault, the comprehensive inspection system outputs defect information by creating a defect report; when creating the defect report, the information that must be saved includes: mileage of a defective component, POS track coordinates, corresponding point cloud information, corresponding influence information, types and levels of faults/absences;

s309) after the inspection of one section of point cloud is finished, automatically reading the point cloud of the next section, and repeating the above steps S301) to S308) to perform inspection until the line inspection is finished.

4) Sound-wind barrier inspection

In downtown areas, residential areas, and areas such as bridges or gobi where wind forces are large, sound/wind barriers are provided on both sides of the railway line. These barriers play a key role in reducing noise, especially noise injuries in nearby populated areas, and in areas where wind forces are high, sound/wind barriers also play a key role in protecting railway and operational safety. As the operation time increases, these barriers are broken for various reasons, and therefore, it is necessary to periodically perform inspection, find the broken condition, and detect the severity of the broken condition. The detected sound/wind barrier damage information can be used for guiding maintenance personnel to maintain and replace the damaged sound barrier guard rail.

The three-dimensional point cloud data and panoramic image data of the acquisition line are scanned by the three-dimensional scanning system 1, and the method can be used for inspection of the sound and wind barrier. Based on the geometric characteristics of the sound/wind barrier, a corresponding algorithm can be designed to extract sound/wind barrier information from the acquired three-dimensional point cloud data, point cloud data of a sound/wind barrier area are extracted and analyzed through comparison of multi-period point cloud data, and whether the sound/wind barrier is damaged or not and the damage shape and area are detected. As shown in fig. 10, the specific implementation steps are as follows:

s400) in order to make the positioning effect when the point cloud compares better, before carrying out the three-dimensional scanning of railway line, need set up a plurality of very stable laser mark targets in position on the railway barrier of inspection, adjacent target is "zigzag" and arranges, every two mark targets horizontal distance about 100 meters to adopt the principle of laying in the vertical direction: i.e. the high-order marker targets are positioned at the top end of the barrier, and the low-order marker targets are positioned at the bottom of the barrier; when in field operation, acquiring sensor raw data of point cloud and panoramic photo; during the inner industry, generating point cloud, POS track file and panoramic photo;

S401) because the point cloud is in mass level, in order to ensure the smoothness of the inspection calculation, the method is suggested to manually intercept the point cloud of the road section only comprising the barrier, if the length of the road section is still longer, the point cloud is required to be segmented, and the point cloud is preferably divided into the point cloud segments with the size of 1GB-3GB according to the general performance of the current processor;

s402) synchronously dividing the POS track file into POS track file segments corresponding to the point cloud segments, and matching the POS track file with the point cloud file;

s403) reading a point cloud segment of mileage to be inspected and a corresponding POS track file;

s404) separating by using a cloth algorithm to obtain a ground point cloud, and removing the point cloud of the aerial part;

s405) since the inspection is performed on the barrier, the point cloud acquired in step S404) needs to be first divided to separate the point cloud of the barrier portion: because the inertial navigation device is fixedly installed during field operation data acquisition, the relative position relation between the inertial navigation device and the ballast bed is relatively stable, the ballast bed can be segmented and extracted by utilizing a POS track file, point clouds of a barrier part are all point clouds in the vertical direction, and the normal line of the point clouds of the part is perpendicular to the POS track;

s406) extracting point clouds (the point clouds of the barrier part are contained in the point clouds of the part) with the POS track line as a central line and the width range of 2.5-3.5 meters at both sides, wherein the point clouds of the part are C, performing high-level filtering on the point clouds C, removing the point clouds with the height coordinate value Z smaller than 0.7 meter to obtain residual point clouds C ', calculating the point cloud normal line of the C', removing the point clouds of the point clouds which are not perpendicular to the POS track line, and finally obtaining the point clouds CP only containing the barrier part;

S407) extracting laser mark targets from the point cloud CP, and setting unique numbers or names for the extracted laser mark targets, wherein the numbers or names of the laser mark targets at the same position in the point cloud at different periods are the same;

s408) reading in a front-stage point cloud to be compared;

s409) performing coarse superposition display of two-stage point clouds by using mileage and coordinates of a laser mark target;

s410) calculating accurate point cloud superposition parameters by using the earlier laser mark target coordinates and the current laser mark target coordinates;

s411) carrying out coordinate conversion on the current coordinate by utilizing the accurate superposition parameters, and carrying out accurate point cloud superposition comparison after conversion;

s412) after superposition, traversing the current-period point clouds, and counting the number of the point clouds within the radius range of 2cm of each point cloud, wherein for the point clouds with the number of not 0 in the previous-period point clouds and the number of 0 in the current-period point clouds, the barrier in the current-period point clouds is considered to be damaged;

s413) statistically calculating an area of the damaged portion;

s414) storing the panoramic image and the three-dimensional point cloud scanned at the current period of the damaged mileage, and facilitating the post-hoc staff to confirm manually.

5) Wire pole verticality inspection

The utility pole is a load-bearing body that contacts the wire, and whether the utility pole is tilted or not has a significant impact on the power supply to the railway. The utility pole used for a long time has the risk of tilting or even dumping under the pulling action of the contact net wires, especially in Yun Guigao original and two wide areas, due to more geological disasters, heavy rain and typhoons, the probability of occurrence of the faults is higher than that of other areas, so that the perpendicularity of the utility pole is very necessary to be detected periodically. After the holographic point cloud around the railway is acquired, the comprehensive railway inspection system can divide the point cloud to obtain the point cloud of the telegraph pole part, calculate the point cloud center of the telegraph pole part, perform straight line fitting to obtain a vector equation of the center of the telegraph pole of the overhead line system, and further obtain the perpendicularity of the telegraph pole. As shown in fig. 11, the specific implementation steps are as follows:

S500), during field operation, acquiring sensor raw data of point cloud and panoramic photos; when in internal operation, generating point cloud, POS track file and panoramic photo;

s501) because the point cloud is in mass level, in order to check the smoothness of calculation, the point cloud is preferably divided into point cloud segments with the size of 1GB-3GB according to the general performance of the current processor;

s502) synchronously dividing the POS track file into POS track file segments corresponding to the point cloud segments, and matching the POS track file with the point cloud file;

s503) reading in a point cloud segment of mileage to be inspected and a corresponding POS track file;

s504) because the telegraph pole is positioned at the distance of two sides of the railway and the distance between the telegraph pole and the center line of the rail (the center line of the left and right steel rails of the rail along the rail direction) is 3.5 meters to 6 meters, redundant point clouds can be deleted by utilizing the POS track, the calculated amount is reduced, and the rest point clouds are marked as C;

s505) first find all poles, the method of finding is as follows:

s5051) calculating a cross section of the point cloud at the beginning of the POS track, wherein the height of the cross section is the same as that of the POS track, and the cross section is horizontal;

s5052) obtaining an intersection point of the cross section and the point cloud C, wherein the cross section is necessary to intersect with a nearby telegraph pole because the height of the POS track is positioned between the contact net line and the steel rail;

S5053) respectively calculating the diameter or size of the point cloud at each intersection point, and calculating the distance between the centers of the point clouds at two adjacent intersection points;

s5054) since the diameter of the utility poles is a fixed value and the distance values between utility poles are also relatively fixed, some interference items can be eliminated with these two conditions;

s5055) removing the interference item, obtaining a remaining point cloud, namely, considering the remaining point cloud as a telegraph pole, and recording the obtained telegraph pole as D (recording x and y coordinates of the center of the telegraph pole);

s5056) since the railway fluctuates, the cross section at the POS track start point does not necessarily find all poles, so the cross section is obtained every 200 meters from the POS track start point (the calculation process of each cross section repeats steps S5052) to S5055), and the obtained poles are incorporated into D;

s5057) performing cyclic calculation to obtain all the poles D;

s506) for each pole in D, the following operation is performed:

s5061) taking out one of the utility poles Di (xi, yi) in D;

s5062) extracting point clouds C-Di with Di (xi, yi) as a circle center and a radius of about 1 meter from the point clouds C, and calculating the maximum value zmax and the minimum value zmin of the elevation coordinates zi of all points in the C-Di;

s5063) setting a horizontal cross section at intervals of 100 mm between a maximum value zmax and a minimum value zmin, solving an intersecting surface of the cross section and C-Di, and solving a center oi of point cloud of the intersecting surface;

S5064) after obtaining the centers of the intersecting surfaces of all cross sections and C-Di, fitting all center points into a straight line, and obtaining the slope of the straight line;

s5065) records the position and slope of the utility pole.

S507) saves and outputs the result of the calculation.

6) Track geometry detection

The feeling of visually feeding back the smoothness of the track to passengers is the smoothness of the train, and the better the smoothness is, the more stable the train operates. The railway after long-time operation generally generates linear change and offset, and the offset must be obtained through regular track geometric parameter measurement and then subjected to linear correction. The smoothness of the track is generally measured by adopting geometric parameters of the track, and particularly comprises track gauge, superelevation, horizontal offset, vertical offset and the like of the line.

The three-dimensional scanning system 1 scans to obtain a three-dimensional point cloud model of the line environment, and high-precision steel rail extraction is performed on the basis. After the rail coordinates are obtained, the rail center line coordinates can be obtained, so that line plane and elevation measurement and track linear measurement can be performed. As shown in fig. 12, the specific implementation steps are as follows:

s600) during field operation, acquiring sensor raw data of point cloud and panoramic photos; during the internal operation, the point cloud, POS track files and panoramic photo are produced;

S601) segmenting the generated point cloud, and in order to reduce the burden of the point cloud processing system 40, preferably, segmenting the original point cloud into point cloud segments according to 1000 meters of each segment and simultaneously segmenting the POS track file into file segments corresponding to the point cloud segments because the volume of the point cloud is very large;

s602) reading a section of point cloud, and firstly separating by using a cloth algorithm to obtain a ground point cloud;

s603) reading in a POS track file corresponding to the section of point cloud;

s604) removing point clouds which are centered on the POS track and are more than 2 meters away from the POS track, wherein the rest point clouds are point clouds only comprising ballast beds;

s605) in the ground point cloud, the elevation of the steel rail is higher than the elevation of the railway ballast, so that the point cloud of the steel rail part can be obtained by separation by utilizing the elevation difference;

s606), carrying out rail top surface fitting and extraction by utilizing the point cloud of the separated rail parts to obtain left and right rail top surface vector lines, and checking the reliability of the extraction of the rail top surface vector lines (namely the center line of the rail top surface along the rail direction) by utilizing the track gauge range of 1435+/-3 mm between the rails;

s607) carrying out track center line calculation by using the extracted left and right steel rail top surface vector lines, and obtaining track center line vectors, wherein the vector node interval of the center line vectors is 20cm;

S608), a mileage value is given to the track center line, and the mileage of the track center line is unified with the railway maintenance mileage by a method of setting a starting point mileage and a finishing point mileage;

s609) since the requirement of the track geometric parameter measurement on the precision is very high, the track centerline coordinates also need to be refined, and the specific processing steps are as follows:

traversing the vector nodes in the track which are already obtained, and carrying out the following operation on each vector node:

s6091) calculating the cross section of the steel rail passing through the line vector node in the track, wherein the thickness of the cross section is not more than 15cm;

s6092) projecting the point cloud on the cross section relative to the cross section, and obtaining the two-dimensional coordinates of the cross section of the steel rail after projection;

s6093) respectively carrying out accurate matching on the cross sections of the left steel rail and the right steel rail with the standard profile of the steel rail;

s6094) obtaining top surface coordinates of the left steel rail and the right steel rail after matching, and updating the top surface coordinates of the original steel rail by using the top surface coordinates;

s6095) calculating the center line coordinate of the accurate track, and updating the center line vector node coordinate of the original track;

s610) outputting the calculated and refined orbit centerline vector line for calculating the geometric parameters of the orbit.

7) Track limit and contour inspection

The three-dimensional scanning system 1 is utilized to scan and collect three-dimensional point cloud data of a track, the point cloud is firstly divided into segments along the direction of a steel rail, then the point cloud of a steel rail area is extracted from each segment of tunnel point cloud, a reference coordinate system for tunnel limit analysis is established through the point cloud of the steel rail, high-precision section data of the track is obtained, and finally whether foreign matter invasion exists is judged according to limit regulations of the railway track.

As shown in fig. 13, the specific implementation steps are as follows:

s700), during field operation, acquiring sensor raw data of point cloud and panoramic photos; when in the field operation, generating point cloud and panoramic photos;

s701) segmenting the generated point cloud, and preferably segmenting the original point cloud into segments according to 1000 meters of each segment in order to reduce the burden of a point cloud processing system because the volume of the point cloud is very large;

s702) reading a section of point cloud, and firstly separating by using a cloth algorithm to obtain a ground point cloud;

s703) in the ground point cloud, the elevation of the steel rail is higher than the elevation of the railway ballast, so that the point cloud of the steel rail part can be separated by utilizing the elevation difference;

s704) carrying out rail top surface fitting and extraction by utilizing the point cloud of the separated rail parts to obtain left and right rail top surface vector lines, and checking the reliability of rail top surface vector line extraction by utilizing the track gauge range of 1435+/-3 mm between the rails;

s705), calculating a track center line by using the extracted left and right steel rail top surface vector lines, and obtaining the track center line;

s706), a mileage value is given to the track center line, and the mileage of the track center line is unified with the railway maintenance mileage by a method of setting a starting point mileage and a finishing point mileage;

S707) calculating the cross section of the point cloud according to the set mileage interval value, and outputting all the point clouds of the cross section;

s708) for each cross-sectional point cloud, performing the following operations:

s7081) taking the intersection point of the cross section and the track center line as the origin of coordinates, wherein the x-axis is parallel to the top surface of the steel rail and points to the center line of the right steel rail top surface along the track direction, and the z-axis is perpendicular to the x-axis to establish a limiting reference coordinate system;

s7082) projecting the cross-section point cloud with respect to the cross section, and obtaining two-dimensional coordinates of the rail cross-section point cloud after projection;

s7083) coordinate-converting the cross-section point Yun Erwei to the limit reference coordinate system;

s7084) generating a standard building bounding model in a bounding reference coordinate system;

s7085) performing a superposition collision analysis on the cross-section point cloud and the standard building bounding model to obtain a bounding calculation result;

s709) saves and outputs the result of the limit calculation.

8) Contact net geometric parameters inspection

The geometrical parameters of the contact net represent the relative position relation between the contact net wire and the railway track central line, the transverse distance between the contact net wire and the railway track central line is called a pull-out value, and the longitudinal distance between the contact net wire and the railway track central line is called a guide height. In order to ensure normal current receiving of the pantograph, the geometrical parameters of the overhead contact system need to be in a reasonable interval, and the geometrical parameters of the overhead contact system exceed the normal interval, so that the pantograph can be separated from the pantograph, and train operation safety accidents are caused. The contact net geometry thus belongs to data that must be measured regularly.

And scanning and collecting point clouds of the railway line by using the three-dimensional scanning system 1, then performing cross section processing, extracting steel rails and contact wires from the cross section, and finally calculating the relative position relation between the contact wires and the track center line to obtain the geometrical parameters of the contact net. As shown in fig. 14, the specific implementation steps are as follows:

s800), during field operation, acquiring sensor raw data of point cloud and panoramic photos; when in internal operation, generating point cloud, POS track file and panoramic photo;

s801) segmenting the generated point cloud, since the volume of the point cloud is very large, in order to reduce the burden of the point cloud processing system 40, it is preferable to segment the original point cloud into segments according to 1000 meters per segment, and segment the POS track file into file segments corresponding to the point cloud segments;

s802) reading a section of point cloud and reading a POS track file corresponding to the section of the point cloud;

s803) removing point clouds which are centered on the POS track and are more than 2 meters away from the POS track, wherein the rest point clouds only comprise a track bed and point clouds Ce above the track bed;

s804) extracting point clouds Cf-d of the ground part in the point clouds Ce by using a distribution algorithm, and obtaining point clouds Cf-k of the air part by taking out the point clouds Cf-d from the point clouds Ce;

S805) in the ground point cloud Cf-d, the elevation of the steel rail is higher than the elevation of the railway ballast, so that the point cloud of the steel rail part can be separated by utilizing the elevation difference;

s806) carrying out rail top surface fitting and extraction by utilizing the point cloud of the separated rail parts to obtain left and right rail top surface vector lines, and checking the reliability of rail top surface vector line extraction by utilizing the track gauge range of 1435+/-3 mm between the rails;

s807) carrying out track center line calculation by using the extracted left and right steel rail top surface vector lines, and obtaining track center line vectors, wherein the vector node interval of the track center line vectors is 20cm;

s808) giving a mileage value to the track center line, and unifying the mileage of the track center line with the railway maintenance mileage by setting a starting point mileage method and a finishing point mileage method;

s809) in the aerial point clouds Cf-k, since only the point clouds above the track bed are defined, the interfering point clouds are less, only the point clouds contacting the wire and the carrier cable portion, and the point clouds of the base net support device portion at the utility pole; in order to ensure the accuracy of the extraction of the contact network lines, a method of manual mouse clicking is preferably adopted to designate the contact network lines to be identified, seed point clouds are provided for the extraction of the contact network lines, then a region growing algorithm (namely, pixel points with similar properties are combined together, one seed point is firstly designated for each region to serve as a starting point of growth, then the pixel points in the surrounding field of the seed points are compared with the seed points, the points with similar properties are combined together to continue to grow outwards until pixels which do not meet the conditions are included, thus completing the growth of one region) is utilized to identify and extract the contact network lines, vector line fitting is carried out, and the distance between vector nodes is 20cm;

S810) calculating geometrical parameters of adjacent contact wires, wherein the specific steps are as follows:

s8101) traversing all vector nodes of the center line of the top surface of the left rail in the railway rail;

s8102) calculating a direction vector for each vector node (denoted as point a), generating a normal plane F perpendicular to the direction vector;

s8103) traversing vector nodes of the middle line of the top surface of the right steel rail, and searching two adjacent vector nodes (marked as C and D) positioned at two sides of a normal plane F;

s8104) calculating an intersection point (denoted as point B) of the normal plane F and the CD connection;

s8105) traversing the vector nodes contacting the net wires, and finding two adjacent vector nodes (denoted as E, F) located on both sides of the normal plane F;

s8106) calculating an intersection point (denoted as point G) of the normal plane F and the EF connecting line;

s8107) making a perpendicular line of a line segment AB through a point G, marking the drop foot as a point H, and marking the midpoint of the line segment AB as I;

s8108) the length of HI connecting line is the contact line pull-out value of the mileage corresponding to the point I;

s8109) GH length is the contact line elevation value of the mileage corresponding to the point I;

s811) save and output the result of the contact line geometry calculation.

9) Line spacing inspection

The line spacing characterizes the horizontal distance between the lines of two adjacent lines and is a key index for determining whether the railway line can safely operate. Too small a line spacing may affect locomotive operation on both lines. The index is generally measured after a railway line is overhauled and a track is lifted.

The three-dimensional point cloud data of the line are scanned and collected by the three-dimensional scanning system 1, the point cloud is firstly divided into segments along the direction of the steel rail, then the point cloud of the steel rail area is extracted from each segment of tunnel point cloud, the vector line of the top surface of the steel rail is obtained by utilizing the point cloud of the steel rail, and the track center line can be calculated according to the vector lines of the steel rails on the left side and the right side. When a plurality of tracks exist in the point cloud, a plurality of track center lines can be obtained. As shown in fig. 15, the specific implementation steps are as follows:

s900), acquiring sensor raw data of point cloud and panoramic photos during field operation; when in the field operation, generating point cloud and panoramic photos;

s901) segmenting the generated point cloud, and preferably segmenting the original point cloud into segments according to 1000 meters of each segment in order to reduce the burden of a point cloud processing system because the volume of the point cloud is very large;

s902) reading a section of point cloud, and firstly separating by using a cloth algorithm to obtain a ground point cloud;

s903) in the ground point cloud, the elevation of the steel rail is higher than the elevation of the railway ballast, so that the point cloud of the steel rail part can be separated by utilizing the elevation difference; meanwhile, in order to increase the stability of detection, the position of the steel rail can be determined by a method of clicking the steel rail point cloud by a mouse, and then the steel rail point cloud is extracted by utilizing a region growing algorithm;

S904) carrying out rail top surface fitting and extraction by utilizing the point cloud of the separated rail parts to obtain left and right rail top surface vector lines, and checking the reliability of rail top surface vector line extraction by utilizing the track gauge range of 1435+/-3 mm between the rails;

s905) calculating a track center line by using the extracted vector lines of the top surfaces of the left steel rail and the right steel rail, and obtaining the track center line;

s906) giving a mileage value to the track center line, and unifying the track center line mileage with the railway maintenance mileage by setting a starting point mileage method and a finishing point mileage method;

s907) calculates the line spacing of the lines in two adjacent tracks:

s9071) traversing all vector nodes in one of the track lines,

s9072) calculates a direction vector for each vector point (denoted as point a), generates a normal plane perpendicular to the direction vector,

s9073) traversing vector points of the midlines of adjacent tracks, finding two adjacent vector nodes (denoted C, D) located on either side of the normal plane,

s9074) calculating the intersection point (denoted as point B) of the normal plane and the line segment CD connection;

s9075) calculating the horizontal distance of the line segment AB, wherein the horizontal distance of AB is the line spacing at the point;

s9076) records the mileage of the point a, and the mileage of a is recorded as the mileage value of the line interval.

S908) saves and outputs the result of the pitch calculation.

10 Railway ballast plumpness inspection

The ballast bed is an important component of the track and is the foundation of the track frame. The main function of the track bed is to support the sleeper, uniformly transmit the huge pressure on the upper part of the sleeper to the roadbed, fix the position of the sleeper, prevent the sleeper from moving longitudinally or transversely, greatly reduce the deformation of the roadbed and alleviate the impact of locomotive wheels on the steel rail. However, after long-time running of the locomotive and rail lifting and lining, the defects of irregular track bed shape, incomplete track ballasts and the like can occur. So the shape of the ballast bed needs to be periodically inspected to find diseases in time. The most important inspection index of the ballast is the fullness of the ballast, the fullness of the ballast is judged by using the angle of the ballast side slope, and the standard ballast side slope gradient is 1:1.5.

The comprehensive railway inspection system can be used for rapidly and accurately detecting and analyzing the track bed on the basis of acquiring the panoramic three-dimensional point cloud of the line. Firstly, intercepting a point cloud section of a railway line to obtain high-precision section data of a ballast bed, then identifying and separating point clouds of a ballast bed part, obtaining an angle of a ballast bed slope through linear fitting, and judging whether conditions such as railway ballast exist or not. As shown in fig. 16, the specific implementation steps are as follows:

S1000) collecting the original data of the sensor during field operation; during the internal operation, generating data such as point cloud, panoramic image, POS track file and the like;

s1001) because the point cloud is in mass level, in order to ensure the smoothness of inspection calculation, the point cloud needs to be segmented, and the point cloud is preferably segmented into point cloud segments with the size of 1GB-3GB according to the general performance of the current processor;

s1002) synchronously dividing the POS track file into POS track file segments corresponding to the point cloud segments, and matching the POS track file with the point cloud file;

s1003) reading in a point cloud segment of mileage to be inspected and a corresponding POS track file segment;

s1004) separating by using a cloth algorithm to obtain a ground point cloud, and removing the point cloud of the air part;

s1005) extracting point clouds with the POS track line as a central line and 2.5 m width ranges on two sides, wherein the extracted point clouds are point clouds only comprising a track bed part;

since the inspection is performed on the ballast bed, the point cloud is first divided to separate the point cloud of the ballast bed part: because the inertial navigation device is fixedly installed during field operation data acquisition, the relative position relation between the inertial navigation device and the ballast bed is relatively fixed, so that the POS track file can be utilized to segment and extract the ballast bed;

S1006) in the point cloud obtained in the step S1008), the elevation of the steel rail is higher than the elevation of the railway ballast, so that the point cloud of the steel rail part can be obtained by utilizing the elevation difference;

s1007) fitting and extracting the top surfaces of the steel rails by utilizing the point cloud of the separated steel rail parts to obtain left and right top surface vector lines of the steel rails, and checking the reliability of the extraction of the top surface vector lines of the steel rails by utilizing the track gauge range of 1435+/-3 mm between the steel rails;

s1008) calculating a track center line by using the extracted left and right steel rail top surface vector lines, and obtaining the track center line;

s1009) a mileage value is given to the track center line, and the mileage of the track center line is unified with the railway maintenance mileage by a method of setting the starting point mileage and the ending point mileage;

s1010) calculating the cross section of the point cloud according to the set mileage interval value, and outputting all the point clouds of the cross section;

s1011) for each cross-sectional point cloud, the following operations are performed:

s10111) as the cross section has a certain thickness (generally between 10cm and 30 cm), firstly, the point cloud in the cross section is projected to the cross section to form a two-dimensional point cloud C;

s10112) positioning the rail in cross section: the steel rail can be positioned by calculating the intersection point of the vector line of the top surface of the steel rail and the cross section;

S10113) positioning the position of the ballast bed side slope in the cross section by using the relative positional relationship of the ballast bed side slope and the steel rail, and obtaining according to the standard of the ballast bed cross section: the 850mm of the outer rail going outwards is the high site of the track bed slope, and the 850mm+400mm of the outer rail going outwards is the low site of the track bed slope;

s10114) establishing a standard coordinate system by taking a connecting line of the midpoints of the top surfaces of the left steel rail and the right steel rail as an x axis, taking the midpoint of the connecting line as a coordinate origin, taking a vertical connecting line upwards as a y axis, and rotationally translating the two-dimensional point cloud C to the standard coordinate system;

s10115) generating a ballast bed standard cross section model in a standard system, calculating the areas of the point cloud and the graph enclosed by the standard model one by one, wherein the excess ballast is above the standard model, the ballast is below the standard model, the ballast is missing, and the ballast plumpness of the section can be obtained after calculation is finished;

s10116) saving the calculation result of the plumpness of the railway ballast;

s1012) adding up the calculated points of the steps S10111) to S10116) for each frame point cloud of a continuous section of the line, and obtaining the condition of the ballasts plumpness of the whole section of the line;

s1013) save and output the result of calculation of the ballast fullness.

The railway comprehensive inspection system described in the embodiment of the invention adopts the three-dimensional scanning system 1 to scan to obtain high-precision holographic point cloud and high-resolution panoramic image within 50 meters of the central line of the railway, and the panoramic image and the point cloud are precisely sleeved and combined to obtain true-color holographic point cloud. And then, on the basis of the obtained point cloud and panoramic image, carrying out railway related inspection work, so that most of detection system functions in the conventional railway comprehensive inspection vehicle can be replaced, and detection or other functions which are not possessed by the conventional detection system are provided. Because the integrated performance of the system is high, the collected data are extremely rich, and the precision of the point cloud is higher, more railway infrastructure detection functions, railway asset management, general investigation and other functions can be realized, compared with the prior equipment, the system complexity is greatly reduced, the operation efficiency of the comprehensive inspection system is improved, the labor cost is saved, and the economical and practical performances are improved.

Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and further implementations are included within the scope of the preferred embodiment of the present invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present invention.

It will be appreciated by those of ordinary skill in the art that all or part of the steps carried out in the method of the above-described embodiments may be implemented by a program to instruct related hardware, and the program may be stored in a computer readable medium, where the program when executed includes one or a combination of the steps of the method embodiment.

In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment.

It should be understood that the structures, proportions, sizes, etc. shown in the drawings are for illustration purposes only and should not be construed as limiting the scope of the present disclosure, since any structural modifications, proportional changes, or dimensional adjustments made by those skilled in the art should not be made in the present disclosure without affecting the efficacy or achievement of the present disclosure.

By implementing the technical scheme of the railway comprehensive inspection system described by the specific embodiment of the invention, the following technical effects can be produced:

(1) The railway comprehensive inspection system described in the specific embodiment of the invention replaces the existing multiple sets of systems on the comprehensive inspection vehicle with the data acquired by a single set of equipment, greatly improves the integration of the detection equipment, and is far lower than the existing comprehensive inspection system in terms of complexity and economic cost of the system and operation;

(2) According to the railway comprehensive inspection system described in the embodiment of the invention, the data collected by a single set of equipment can be used by a plurality of subsequent application systems, so that the compatibility of the data among a plurality of detection equipment is increased, the multiplexing of the plurality of equipment and the intercommunication and interconnection of the detection data are realized, and the development complexity of the application system is reduced;

(3) Compared with the conventional detection application based on the plane image, the comprehensive railway inspection system described in the specific embodiment of the invention can perform qualitative analysis and accurate quantitative analysis, can realize a large number of infrastructure detection functions, can realize disaster analysis and evaluation, and has more abundant functions;

(4) The railway comprehensive inspection system described by the embodiment of the invention adopts three-dimensional live-action inspection, has true color point cloud and panoramic image, can accurately measure, has great progress compared with the traditional inspection based on images, simultaneously provides a virtual inspection function, greatly reduces the working intensity of inspection workers, and greatly improves the accuracy and the automation degree of inspection.

In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described by a difference from other embodiments, and identical and similar parts between the embodiments are referred to each other.

The above description is only of the preferred embodiment of the present invention, and is not intended to limit the present invention in any way. While the invention has been described in terms of preferred embodiments, it is not intended to be limiting. Any person skilled in the art can make many possible variations and modifications to the technical solution of the present invention or equivalent embodiments using the method and technical solution disclosed above without departing from the spirit and technical solution of the present invention. Therefore, any simple modification, equivalent substitution, equivalent variation and modification of the above embodiments according to the technical substance of the present invention still fall within the scope of the technical solution of the present invention, unless departing from the technical solution of the present invention.

Claims

1. A comprehensive inspection system for railways, comprising: the three-dimensional scanning system (1), the space-time synchronization system (2), the storage/control system (3) and the inspection application system (4), wherein the three-dimensional scanning system (1) comprises a gesture-determining positioning system (10), a laser radar (11) and a panoramic camera (12), and the inspection application system (4) comprises a point cloud processing system (40) and a comprehensive inspection analysis system (41); the attitude determination positioning system (10) acquires data of a global navigation satellite system, an inertial measurement unit and a distance measurement instrument of the railway comprehensive inspection vehicle, the laser radar (11) acquires three-dimensional coordinate data of an object to be measured, and the panoramic camera (12) acquires panoramic image data; the space-time synchronization system (2) provides clock synchronization for three-dimensional coordinates, an inertial measurement unit, a distance measurement instrument and panoramic image data according to the time service of the global navigation satellite system; the storage/control system (3) stores three-dimensional coordinates, an inertial measurement unit, a distance measurement instrument and panoramic image data after clock synchronization in real time; the point cloud processing system (40) firstly uses a global navigation satellite system, an inertial measurement unit and a distance measuring instrument to perform data fusion and calculation on positioning and attitude determination system data, then uses the positioning and attitude determination system data and laser radar data to jointly calculate point cloud data, and then performs fusion processing on the point cloud data and panoramic image data to obtain real scenic spot cloud; the comprehensive inspection analysis system (41) performs railway inspection analysis by taking point cloud and panoramic image data as data sources; the comprehensive inspection analysis system (41) comprises a side slope disaster inspection module (42), a plurality of laser mark targets are arranged on a railway side slope to be inspected, the three-dimensional scanning system (1) performs original data acquisition during field operation, and the point cloud processing system (40) generates point cloud and panoramic image data during field operation; the slope disaster inspection module (42) accurately extracts laser mark targets from the scanned point cloud data, and sets unique numbers or names for the extracted laser mark targets, wherein the numbers or names of the laser mark targets at the same position in the point cloud obtained by scanning in different periods are the same; acquiring the current-period point cloud, reading in the to-be-compared previous-period point cloud, performing coarse superposition display on the two-period point cloud by using the mileage and coordinates of the laser mark target, and calculating accurate point cloud superposition parameters by using the coordinates of the previous-period laser mark target and the coordinates of the current-period laser mark target; coordinate conversion is carried out on the current-period coordinates by applying accurate point cloud superposition parameters, accurate point cloud superposition comparison is carried out after conversion, and coordinate differences of the same positions in the two-period point clouds are calculated after superposition; checking the result of the variable quantity, and taking panoramic images in the current period and the earlier period for the region with larger coordinate difference, and manually confirming whether slope change occurs or not; and (3) storing the coordinate difference of the point cloud superposition contrast, reading in the point cloud contrast coordinate difference in the earlier stage, drawing a coordinate difference change curve, detecting whether the coordinate difference has a trend of gradually amplifying, and carrying out early warning or alarming on an area where the coordinate difference is gradually amplified.

2. The integrated inspection system of railways of claim 1, wherein: the slope disaster inspection module (42) gridding the point cloud at the current period, performing plane fitting on the point cloud grid C processed first to obtain a plane P, and solving the center coordinate O of the point cloud grid C; then, the normal line of a plane P passing through the point O is obtained, a normal line L is obtained, an intersection point J of the normal line L and the earlier-stage point cloud is obtained, and the distance of the line segment OJ is calculated, namely the variation of the same position in the two-stage point cloud is obtained; and traversing each grid in the current-period point cloud, and calculating and obtaining the front and rear period variation of each grid point cloud, thereby obtaining the coordinate difference of the same position in the two-period point cloud.

3. A comprehensive inspection system for railways, comprising: the three-dimensional scanning system (1), the space-time synchronization system (2), the storage/control system (3) and the inspection application system (4), wherein the three-dimensional scanning system (1) comprises a gesture-determining positioning system (10), a laser radar (11) and a panoramic camera (12), and the inspection application system (4) comprises a point cloud processing system (40) and a comprehensive inspection analysis system (41); the attitude determination positioning system (10) acquires data of a global navigation satellite system, an inertial measurement unit and a distance measurement instrument of the railway comprehensive inspection vehicle, the laser radar (11) acquires three-dimensional coordinate data of an object to be measured, and the panoramic camera (12) acquires panoramic image data; the space-time synchronization system (2) provides clock synchronization for three-dimensional coordinates, an inertial measurement unit, a distance measurement instrument and panoramic image data according to the time service of the global navigation satellite system; the storage/control system (3) stores three-dimensional coordinates, an inertial measurement unit, a distance measurement instrument and panoramic image data after clock synchronization in real time; the point cloud processing system (40) firstly uses a global navigation satellite system, an inertial measurement unit and a distance measuring instrument to perform data fusion and calculation on positioning and attitude determination system data, then uses the positioning and attitude determination system data and laser radar data to jointly calculate point cloud data, and then performs fusion processing on the point cloud data and panoramic image data to obtain real scenic spot cloud; the comprehensive inspection analysis system (41) performs railway inspection analysis by taking point cloud and panoramic image data as data sources; the comprehensive inspection analysis system (41) comprises a side slope disaster assessment module (43), a plurality of laser mark targets are arranged on a railway side slope to be inspected, the three-dimensional scanning system (1) performs original data acquisition during field operation, and the point cloud processing system (40) generates point cloud and panoramic image data during field operation; the slope disaster evaluation module (43) extracts laser mark targets from the scanned point cloud data, and sets unique numbers or names for the extracted laser mark targets, wherein the numbers or names of the laser mark targets at the same position in the point clouds at different periods are the same; acquiring the current-period point cloud, reading the to-be-compared previous-period point cloud, performing coarse superposition display of the two-period point cloud by using the mileage and coordinates of the laser mark target, and calculating accurate point cloud superposition parameters by using the coordinates of the previous-period laser mark target and the current-period laser mark target; performing coordinate conversion on the current-period coordinates by using accurate superposition parameters, performing accurate point cloud superposition comparison after conversion, calculating point cloud coordinate differences of the same positions in two-period point clouds after superposition, and focusing on the difference value of the elevation coordinates z; calculating the volume of the part clamped by the two point clouds by utilizing the coordinate difference of the two point clouds, namely, the earth and stone sides of the landslide, and calculating the area of the landslide part; and (3) taking panoramic image data scanned in the previous period, checking actual topography and topography of a disaster point, intercepting and storing the three-dimensional point cloud according to mileage, positioning the three-dimensional point cloud to the mileage of an emergency event, and taking the three-dimensional point cloud and the panoramic image at the position as a base map of emergency treatment to provide map reference for the emergency treatment.

4. A comprehensive inspection system for railways according to claim 3, wherein: the comprehensive inspection analysis system (41) comprises a line virtual inspection/multiplication module (44), the three-dimensional scanning system (1) is used for collecting original data during field operation, and the point cloud processing system (40) is used for generating a point cloud, a panoramic image and a positioning and attitude determination system track file during field operation; the line virtual inspection/multiplication module (44) utilizes the calibrated panoramic image as a gray point cloud true color value to generate true color point cloud, divides the true color point cloud into point cloud segments, and simultaneously divides the positioning and attitude determination system track file into positioning and attitude determination system track file segments corresponding to the point cloud segments; reading a point cloud segment of a mileage to be inspected and a corresponding positioning and attitude determination system track file; the line virtual inspection/multiplication module (44) displays that the default viewpoint of the interface is positioned at the starting point of the track file of the positioning and attitude determination system, the view field direction is consistent with the track file of the positioning and attitude determination system, and after clicking, a user can perform point cloud roaming in various modes to realize the visual effect of virtual inspection/multiplication; in the point cloud roaming process, a user can change the view field direction at any time and pause the movement of the point cloud, after the movement of the point cloud is paused, the user can check a specific part through zooming, and the image information of the panoramic image checking part corresponding to the point cloud is called out; if a certain component has a defect or a fault, the virtual line inspection/multiplication module (44) outputs defect information by creating a defect report; and when the inspection of the point cloud of one section is finished, automatically reading the point cloud of the next section for inspection until the inspection of the line is finished.

5. A comprehensive inspection system for railways, comprising: the three-dimensional scanning system (1), the space-time synchronization system (2), the storage/control system (3) and the inspection application system (4), wherein the three-dimensional scanning system (1) comprises a gesture-determining positioning system (10), a laser radar (11) and a panoramic camera (12), and the inspection application system (4) comprises a point cloud processing system (40) and a comprehensive inspection analysis system (41); the attitude determination positioning system (10) acquires data of a global navigation satellite system, an inertial measurement unit and a distance measurement instrument of the railway comprehensive inspection vehicle, the laser radar (11) acquires three-dimensional coordinate data of an object to be measured, and the panoramic camera (12) acquires panoramic image data; the space-time synchronization system (2) provides clock synchronization for three-dimensional coordinates, an inertial measurement unit, a distance measurement instrument and panoramic image data according to the time service of the global navigation satellite system; the storage/control system (3) stores three-dimensional coordinates, an inertial measurement unit, a distance measurement instrument and panoramic image data after clock synchronization in real time; the point cloud processing system (40) firstly uses a global navigation satellite system, an inertial measurement unit and a distance measuring instrument to perform data fusion and calculation on positioning and attitude determination system data, then uses the positioning and attitude determination system data and laser radar data to jointly calculate point cloud data, and then performs fusion processing on the point cloud data and panoramic image data to obtain real scenic spot cloud; the comprehensive inspection analysis system (41) performs railway inspection analysis by taking point cloud and panoramic image data as data sources; the comprehensive inspection analysis system (41) comprises an acoustic-wind barrier inspection module (47), a plurality of laser mark targets are arranged on a railway barrier to be inspected, the three-dimensional scanning system (1) performs original data acquisition during field operation, and the point cloud processing system (40) generates point cloud, a positioning and attitude determination system track file and panoramic image data during field operation; the sound-wind barrier inspection module (47) cuts off the point cloud of the road section only comprising the barrier manually, if the road section is long, the point cloud is divided into point cloud sections, meanwhile, the positioning and attitude determination system track file is divided into positioning and attitude determination system track file sections corresponding to the point cloud sections, and the positioning and attitude determination system track file is matched with the point cloud; reading a point cloud segment of mileage to be inspected and a corresponding positioning and attitude determination system track file, separating the point cloud of an aerial part to obtain a ground point cloud, and dividing the ground point cloud to separate the point cloud of a barrier part; extracting point clouds C with a track line of a positioning and attitude determination system as a central line, setting the point clouds C within a width range on two sides, performing height filtering treatment on the point clouds C, removing the point clouds with an elevation coordinate z value smaller than a set value to obtain residual point clouds C ', calculating the point cloud normal line of the C', removing the point clouds of the track line of the non-vertical positioning and attitude determination system in the point clouds, and finally obtaining the residual point clouds CP only comprising a barrier part; extracting laser mark targets from the point cloud CP, and setting unique numbers or names for the extracted laser mark targets, wherein the numbers or names of the laser mark targets at the same position in the point cloud at different periods are the same; reading in early-stage point clouds to be compared, performing coarse superposition display of two-stage point clouds by using mileage and coordinates of a laser marking target, and calculating accurate point cloud superposition parameters by using coordinates of the early-stage laser marking target and coordinates of the laser marking target in the current stage; performing coordinate conversion on the current-period coordinates by using accurate superposition parameters, performing accurate point cloud superposition comparison after conversion, traversing the current-period point clouds after superposition, counting the number of the point clouds of each point cloud in a set radius, and regarding the point clouds of which the number is not 0 in the previous-period point clouds and the number is 0 in the current-period point clouds, namely, considering that a barrier in the current-period point clouds is damaged; and (3) calculating the area of the damaged part by statistics, and storing the panoramic image and the three-dimensional point cloud scanned in the current period at the damaged mileage so as to be convenient for the staff to confirm afterwards.

6. A comprehensive inspection system for railways, comprising: the three-dimensional scanning system (1), the space-time synchronization system (2), the storage/control system (3) and the inspection application system (4), wherein the three-dimensional scanning system (1) comprises a gesture-determining positioning system (10), a laser radar (11) and a panoramic camera (12), and the inspection application system (4) comprises a point cloud processing system (40) and a comprehensive inspection analysis system (41); the attitude determination positioning system (10) acquires data of a global navigation satellite system, an inertial measurement unit and a distance measurement instrument of the railway comprehensive inspection vehicle, the laser radar (11) acquires three-dimensional coordinate data of an object to be measured, and the panoramic camera (12) acquires panoramic image data; the space-time synchronization system (2) provides clock synchronization for three-dimensional coordinates, an inertial measurement unit, a distance measurement instrument and panoramic image data according to the time service of the global navigation satellite system; the storage/control system (3) stores three-dimensional coordinates, an inertial measurement unit, a distance measurement instrument and panoramic image data after clock synchronization in real time; the point cloud processing system (40) firstly uses a global navigation satellite system, an inertial measurement unit and a distance measuring instrument to perform data fusion and calculation on positioning and attitude determination system data, then uses the positioning and attitude determination system data and laser radar data to jointly calculate point cloud data, and then performs fusion processing on the point cloud data and panoramic image data to obtain real scenic spot cloud; the comprehensive inspection analysis system (41) performs railway inspection analysis by taking point cloud and panoramic image data as data sources; the comprehensive inspection analysis system (41) comprises a contact net wire pole verticality inspection module (411), wherein the contact net wire pole verticality inspection module (411) performs point cloud segmentation after collecting holographic point clouds around a railway to obtain point clouds of a wire pole part, calculates the point cloud center of the wire pole part, performs linear fitting to obtain a vector equation of the contact net wire pole center, and obtains the verticality of the wire pole.

7. The integrated inspection system of claim 6, wherein: the three-dimensional scanning system (1) is used for acquiring original data during field operation, and the point cloud processing system (40) is used for generating point cloud, a positioning and attitude determination system track file and panoramic image data during field operation; the overhead line pole verticality inspection module (411) divides the point cloud into point cloud segments, simultaneously divides the positioning and attitude determination system track file into positioning and attitude determination system track file segments corresponding to the point cloud segments, and matches the positioning and attitude determination system track file with the point cloud file; reading in a point cloud segment of mileage to be patrolled and examined and a corresponding positioning and attitude determination system track file, deleting redundant point clouds by utilizing the positioning and attitude determination system track, searching all telegraph poles by using the residual point clouds as C, calculating the position and the slope of each telegraph pole, and storing and outputting a calculation result.

8. The integrated inspection system of claim 7, wherein: the overhead line pole verticality inspection module (411) obtains a cross section from a point cloud at the initial position of the track of the positioning and attitude determination system, and the cross section is the same in height as the track of the positioning and attitude determination system and horizontal to obtain an intersection point of the cross section and the point cloud C; the diameter or the size of the point cloud at each intersection point is calculated respectively, and the distance between the centers of the point clouds at two adjacent intersection points is calculated; the diameter of the telegraph pole is a fixed value, the distance value between the telegraph poles is also a fixed value, interference items are eliminated, the remaining point cloud is obtained, namely the telegraph pole is obtained, the obtained telegraph pole is marked as D, and the coordinates of the center of the telegraph pole are x and y; and (3) starting from the initial point of the track of the positioning and attitude determination system, obtaining a cross section every set distance, merging the obtained telegraph poles into the D, and circularly calculating to obtain all the telegraph poles D.

9. The integrated railroad inspection system of claim 7 or 8, wherein: the overhead line pole verticality inspection module (411) takes out one telegraph pole Di (xi, yi) in the D, extracts point clouds C-Di taking Di (xi, yi) as a circle center in the point clouds C, sets the point clouds C-Di in a radius range, and calculates the maximum value zmax and the minimum value zmin of the elevation coordinates zi of all points in the point clouds C-Di; setting a horizontal cross section between the maximum value zmax and the minimum value zmin at intervals of a set distance, solving an intersecting surface of the cross section and the point cloud C-Di, and solving a center coordinate oi of the point cloud of the intersecting surface; after the centers of the intersecting surfaces of all cross sections and the point clouds C-Di are obtained, all the center points are fitted into a straight line, the slope of the straight line is obtained, and the position and the slope of the telegraph pole are recorded.