Railway intelligent inspection robot

Abstract

The invention provides an intelligent railway inspection robot, and relates to the technical field of railway inspection. The robot comprises a walking assembly composed of a walking unit and a first driving assembly which are matched with each other, the walking unit is matched with a rail, and the first driving assembly drives the walking assembly to move on the rail; the processing assembly is provided with a processor and a data storage unit, the processor receives and processes data, and the data storage unit stores the data; the first data acquisition assembly is arranged on the walking unit and comprises a binocular camera, a three-color warning lamp and a display screen, the binocular camera records an operation behavior video of an operator during railway temporary skylight maintenance operation and transmits the operation behavior video to the processor for processing, the three-color warning lamp comprises a red part, a green part and a yellow part, and the display screen displays an operation site monitoring picture; and the second data acquisition assembly comprises a light sensor, acquires image information of the rail and the fastener and transmits the image information to the processor for processing. According to the invention, automation and intelligentization of rail inspection are realized, and the working efficiency, safety and accuracy are greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of railway inspection, and in particular to a railway intelligent inspection robot. Background Art

[0002] Railway inspection plays a vital role in railway operation and maintenance. It means that the railway-related departments arrange professional staff to strictly follow specific standards, processes and fixed time intervals, and use a series of cutting-edge technical means and scientific methods to carry out comprehensive, systematic, detailed and in-depth inspection, monitoring and evaluation of all components of the railway system.

[0003] Railway inspection covers a wide range of areas. As far as railway line infrastructure is concerned, strict inspections must be carried out on the smoothness of the track, the wear and crack conditions of the rails, the stability of the roadbed and its drainage performance, the integrity and load-bearing capacity of the sleepers, and the switching flexibility and accuracy of the turnouts. At the same time, for the railway communication system, it is necessary to ensure smooth and unimpeded signal transmission, and to promptly detect and deal with possible communication interference or interruption problems. In terms of the signal system, it is necessary to verify the accuracy and timeliness of the signal display to ensure that the train can accurately receive operating instructions. For the inspection of the power supply system, close attention should be paid to the tension and suspension status of the contact network, the operating parameters of the substation equipment, etc., in order to ensure the continuous and stable power supply.

[0004] For example, during railway inspections in high-cold areas, workers need to pay special attention to the shrinkage and deformation of the tracks caused by low temperatures, as well as the possible icing of the contact network that affects power supply. During the inspection of mountain railways, it is important to check whether there are cracks and water leakage in the tunnel lining, and whether the bridge structure is stable, in order to prevent safety hazards caused by geological changes. In short, railway inspections are a key measure to maintain safe, stable, and efficient railway operations, and provide a solid guarantee for the smooth operation of railway transportation and the safety of passengers' lives and property.

[0005] In the prior art, when inspectors conduct foot inspections, they mainly have the following problems:

[0006] First, foot inspections are slow and have limited coverage, making it difficult to complete inspections of longer railway lines in a short period of time.

[0007] Second, for large-scale railway networks, foot inspections may not be able to detect problems in a timely manner, leading to the accumulation and aggravation of problems.

[0008] Third, workers need to walk for a long time, carrying inspection tools and equipment, which can easily cause physical fatigue. Especially in bad weather or complex terrain conditions, such as high temperature, severe cold, mountainous areas, etc., the labor intensity will be further increased, which may affect the quality of inspection.

[0009] Fourth, manual data recording is prone to errors and omissions, which affect the accuracy and completeness of the data. Moreover, it is difficult to analyze and process the collected data in real time on site, which is not conducive to quickly discovering problems and formulating solutions.

[0010] Fifth, during the maintenance work on temporary railway skylights, not only is it impossible to supervise whether the operation is compliant, but it is also difficult to ensure the safety of maintenance personnel. Summary of the invention

[0011] The purpose of the present invention is to provide a railway intelligent inspection robot, aiming to solve the technical problems in the above-mentioned background technology.

[0012] The technical solution of the present invention is achieved in this way:

[0013] The technical solution of this application provides a railway intelligent inspection robot, comprising:

[0014] A travel assembly, comprising a travel unit and a first drive assembly that cooperate with each other, wherein the travel unit is used to adapt to the rails, and the first drive assembly is used to drive the travel unit to move on the rails;

[0015] A processing component, including a processor and a data storage unit, wherein the processor is used for receiving and processing data, and the data storage unit is used for storing data;

[0016] A first data acquisition component includes a binocular camera, a three-color warning light and a display screen which are simultaneously arranged on the walking unit, the binocular camera is used to record the operation of the operator in the form of a video during the maintenance operation of the temporary railway skylight, and transmit the operation in the form of a video to the processor for data processing, the three-color warning light has a red warning part, a green warning part and a yellow warning part, and the display screen is used to display the monitoring screen of the operation site; and

[0017] The second data acquisition component includes a light sensor for acquiring image information of the rails and fasteners, and transmitting the image information to the processor for data processing.

[0018] A further technical solution is that the travel unit comprises a box body and a travel wheel set arranged on the box body, the travel wheel set is used to support the box body and can drive the box body to move on the rails;

[0019] Wherein, the above-mentioned walking wheel set is transmission connected with the above-mentioned first driving assembly.

[0020] A further technical solution is that the binocular camera and the three-color warning light can be rotatably arranged on the box;

[0021] Among them, the above-mentioned box body is provided with a first storage space and a second storage space, the above-mentioned first storage space is used to store the above-mentioned binocular camera, and the above-mentioned three-color warning light is used to store the above-mentioned three-color warning light.

[0022] A further technical solution is that the binocular camera and the three-color warning light are commonly connected to a rotating shaft, and the rotating shaft is transmission-connected to a second drive assembly. When the second drive assembly is running, the binocular camera and the three-color warning light are driven by the rotating shaft to be stored in the first storage space and the second storage space respectively.

[0023] A further technical solution is that both the first drive assembly and the second drive assembly use drive motors.

[0024] A further technical solution is that the inspection robot also includes a power supply battery, and the power supply battery is used to supply power to the first drive assembly, the second drive assembly, the binocular camera, the three-color warning light, the display screen and the optical sensor.

[0025] A further technical solution is that the bottom side of the box body is provided with guide wheels for rolling cooperation with the rails.

[0026] A further technical solution is that the box is also provided with a tool box.

[0027] Compared with the prior art, the technical solution of the present invention has at least the following advantages or beneficial effects:

[0028] (1) When the present application is applied to temporary skylight maintenance operations, the binocular camera can record the operator's operating behavior video and send the operating behavior video to the processor. The processor compares the acquired video data with its own pre-stored basic data to determine whether the operator's operating behavior is standardized. If the operator's operating behavior is normal, the processor only turns on the green light of the three-color warning light; if the operator's operating behavior is slightly irregular, the processor only turns on the yellow light of the three-color warning light; if the operator's operating behavior is seriously irregular, the processor only turns on the red light of the three-color warning light. In this way, the operator can be supervised and warned, which facilitates timely correction of erroneous operations, prevents potential accidents, and ensures that the entire operation process is safe and efficient.

[0029] (2) When the present application is conducting rail inspection, the first drive assembly is operated to cause the inspection robot to automatically walk on the rail. At the same time, the optical sensor is used to obtain image information of the rail track and fasteners, and transmit the image information to the processor, which generates high-precision three-dimensional point cloud data after processing, and analyzes and determines in real time whether there is damage, thereby ensuring the safety and reliability of railway infrastructure.

[0030] (3) The design of this application realizes the automation and intelligence of rail inspection, replacing the traditional manual inspection. Compared with the traditional manual inspection, the work efficiency, safety and accuracy are greatly improved. At the same time, this application also has the function of construction monitoring of operators, which can ensure the safety and efficiency of the entire operation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0032] Figure 1 The structure of a railway intelligent inspection robot according to an embodiment of the present invention is shown in FIG. Figure 1 ;

[0033] Figure 2 for Figure 1 A partial enlarged view of middle A;

[0034] Figure 3 for Figure 1 A partial enlarged view of B in the middle;

[0035] Figure 4 The structure of a railway intelligent inspection robot according to an embodiment of the present invention is shown in FIG. Figure 2 ;

[0036] Figure 5 The structure of a railway intelligent inspection robot according to an embodiment of the present invention is shown in FIG. Figure 3 ;

[0037] Figure 6 This is a schematic diagram of the structure of a railway intelligent inspection robot with a box removed according to an embodiment of the present invention.

[0038] Icons: 1-first storage space, 2-box, 3-display screen, 4-toolbox, 5-binocular camera, 6-three-color warning light, 7-second storage space, 8-guide wheel, 9-traveling wheel set, 10-rotating shaft, 11-first drive assembly, 12-second drive assembly, 13-optical sensor, 14-power supply. DETAILED DESCRIPTION

[0039] Example

[0040] Please refer to Figure 1-Figure 6 , a railway intelligent inspection robot applied to double-track railways, mainly composed of a walking component, a processing component, a first data acquisition component and a second data acquisition component.

[0041] Among them, the walking component includes a walking unit and a first drive assembly 11 that cooperate with each other. The walking unit is in rolling cooperation with the two tracks of the double-track railway. The first drive assembly 11 is used to drive the walking unit to move in a directional manner on the railway track so that the first data acquisition component and the second data acquisition component can carry out the collection of corresponding data.

[0042] Further, the processing component includes a processor and a data storage unit. The processor is used to analyze and process the data collected by the first data acquisition component and the second data acquisition component. At the same time, the processor can plan the walking path for the walking unit, realize the automated inspection of the present application, and improve the detection efficiency and coverage. The data storage unit is used to store data to achieve data recording and management. The first data acquisition component includes a binocular camera 5, a three-color warning light 6 and a display screen 3 that are simultaneously arranged on the walking unit. The binocular camera 5 is used to record the operation behavior video of the operator during the maintenance operation of the temporary skylight of the railway, and transmit the operation behavior video to the processor for data processing. The three-color warning light 6 has a red warning part, a green warning part and a yellow warning part. The display screen 3 is used to display the monitoring screen of the operation site. The second data acquisition component includes a light sensor 13, which is used to obtain image information of the rails and fasteners, and transmit the image information to the processor for data processing. The above-mentioned combination of the optical sensor 13 and the processor has the advantages of high precision (the optical sensor 13 can capture high-resolution three-dimensional data to ensure the accuracy of the detection results), real-time (can acquire and analyze point cloud data in real time, promptly discover and report damage to rails and fasteners) and comprehensive coverage (provides comprehensive rail and fastener detection to ensure the overall safety and reliability of the railway line).

[0043] On the one hand, when the present application is applied to temporary skylight maintenance operations, the binocular camera 5 can record the operation behavior video of the operator and send the operation behavior video to the processor. The processor compares the acquired video data with the basic data pre-stored by itself to determine whether the operation behavior of the operator is standardized. If the operation behavior of the operator is normal, the processor only turns on the green light of the three-color warning light 6; if the operation behavior of the operator is slightly irregular, the processor only turns on the yellow light of the three-color warning light 6; if the operation behavior of the operator is seriously irregular, the processor only turns on the red light of the three-color warning light 6. In this way, the supervision and warning of the operator are achieved, which facilitates timely correction of wrong operations, prevents the occurrence of potential accidents, and ensures that the entire operation process is safe and efficient. On the second hand, when the present application is conducting rail inspection, the first drive assembly 11 is operated to prompt the inspection robot to walk automatically on the rail. At the same time, the optical sensor 13 is used to obtain image information of the rail track and fasteners, and transmits the image information to the processor. After processing, the processor generates high-precision three-dimensional point cloud data, and analyzes and determines whether there is damage in real time, thereby ensuring the safety and reliability of railway infrastructure.

[0044] The design of this application realizes the automation and intelligence of rail inspection, replacing the traditional manual inspection. Compared with the traditional manual inspection, the work efficiency, safety and accuracy are greatly improved. At the same time, this application also has the function of construction monitoring of operators, which can ensure the safety and efficiency of the entire operation process.

[0045] It is worth mentioning that the processor uses the PGS-YOLO algorithm to analyze the video data when processing the video data collected by the dual-head cameras. The above processor covers the input layer, partial convolution module (PConv), GhostNetV2 module, SeaAttention module and output layer.

[0046] The role of the partial convolution module (PConv) is to solve the problem of processing irregular and partially missing input data, and enhance the robustness of the model in the face of occlusion and missing information. The purpose of the GhostNetV2 module is to reduce redundant calculations and the number of parameters, improve computational efficiency, and achieve lightweight models. The SeaAttention module aims to introduce a new attention mechanism, enhance the model's ability to capture key features, and improve the ability to detect targets in complex scenes. The output layer combines the features extracted by the above modules to perform target detection and judgment of operation specifications.

[0047] Input layer: responsible for receiving the image data recorded in real time by the binocular camera 5, and performing preprocessing operations on the input image, including normalization and data enhancement.

[0048] Partial convolution module (PConv): Performs local convolution operation on each part of the input image, performs convolution calculation only in the valid pixel area and ignores the invalid area; uses the feature map output by the partial convolution to fill in the missing part, thereby generating a complete feature map and improving the detection accuracy of the model for incomplete data.

[0049] GhostNetV2 module: The Ghost module is used to generate sparse feature maps, and feature extraction is completed with fewer computing resources and parameters. The sparse feature maps are used for further feature extraction and fusion to generate efficient and accurate feature representations. The efficient feature extraction mechanism of GhostNetV2 is combined to reduce the number of model parameters and the amount of computation to adapt to resource-constrained environments.

[0050] SeaAttention module: weights each spatial position of the input feature map to highlight key areas, suppress irrelevant areas, and generate a weighted spatial feature map; weights each channel of the input feature map to enhance the response of important channels and generate a weighted channel feature map; combines the weighted feature maps of the spatial and channel attention mechanisms to generate the final feature representation, improving the detection accuracy of the model in complex scenes.

[0051] Output layer: Input the feature maps output by each module into the fully connected layer to generate detection results; use NMS (non-maximum suppression) technology to process the detection frame, remove overlapping frames, and retain the optimal detection frame; judge whether the operator's operating behavior complies with the operating specifications based on the detection results, and output the detection results and alarm information.

[0052] The PGS-YOLO algorithm combines partial convolution (PConv) technology, GhostNetV2 module and SeaAttention mechanism. It achieves high precision and efficiency in target detection by solving the problem of processing irregular input data, reducing computational redundancy and enhancing feature capture capabilities. It significantly improves the detection performance in complex scenarios and is particularly suitable for real-time monitoring of the operating specifications of railway maintenance personnel.

[0053] Optionally, the processor of the present application supports wireless network connection, allowing managers to access the robot's recorded data and monitoring screens through remote devices, which enhances management flexibility and enables managers to understand the operation status without on-site supervision, thereby improving management efficiency.

[0054] In some embodiments of the present invention, the travel unit includes a box body 2 and a travel wheel set 9 disposed on the box body 2, and the travel wheel set 9 is used to support the box body 2 and can drive the box body 2 to move on the rails;

[0055] The traveling wheel set 9 is transmission-connected to the first driving assembly 11 .

[0056] In the above embodiment, the box 2 is placed across the double-track rails. The running wheel set 9 has four wheels, which are distributed on two tracks in pairs, thereby achieving multi-angle support for the box 2 and greatly improving the stability of the box 2 moving on the track. The first drive assembly 11 is arranged inside the box 2, achieving a hidden design of the first drive assembly 11.

[0057] In some embodiments of the present invention, the binocular camera 5 and the three-color warning light 6 can be rotatably disposed on the box 2;

[0058] The box body 2 is provided with a first storage space 1 and a second storage space 7 . The first storage space 1 is used to store the binocular camera 5 , and the three-color warning light 6 is used to store the three-color warning light 6 .

[0059] In the above embodiment, the binocular camera 5 and the three-color warning light 6 are both arranged on the upper side of the box body 2. When the binocular camera 5 and the three-color warning light 6 are not in use, the binocular camera 5 is rotated to the first storage space 1, and the three-color warning light 6 is rotated to the second storage space 7. This not only reduces the volume of the present application, but also protects the binocular camera 5 and the three-color warning light 6.

[0060] The three-color warning light 6 is equipped with a voice alarm function. When the three-color warning light 6 is in the yellow light and red light conditions, the three-color warning light 6 can emit a prompt sound.

[0061] In some embodiments of the present invention, the binocular camera 5 and the three-color warning light 6 are commonly connected to a rotating shaft 10, and the rotating shaft 10 is transmission-connected to a second drive assembly 12. When the second drive assembly 12 is running, the binocular camera 5 and the three-color warning light 6 are driven by the rotating shaft 10 to be stored in the first storage space 1 and the second storage space 7, respectively.

[0062] In the above embodiment, when the binocular camera 5 and the three-color warning light 6 are not in use, the second driving assembly 12 drives the rotating shaft 10 to rotate, and the rotating shaft 10 drives the binocular camera 5 and the three-color warning light 6 to be stored at the same time. The binocular camera 5 and the three-color warning light 6 use the same power system, which simplifies the structure of the robot, reduces the cost and the probability of failure, and is practical.

[0063] In some embodiments of the present invention, both the first drive assembly 11 and the second drive assembly 12 are drive motors.

[0064] In the above embodiment, the first driving assembly 11 and the second driving assembly 12 use driving motors as power sources, which has the advantage of stable power output.

[0065] In some embodiments of the present invention, the inspection robot further includes a power supply battery, which is used to supply power to the first drive assembly 11, the second drive assembly 12, the binocular camera 5, the three-color warning light 6, the display screen 3 and the optical sensor 13.

[0066] In the above embodiment, the power supply battery can realize the power supply of the electrical equipment in the robot. The power supply battery is a high-capacity rechargeable battery to ensure the long-term operation of the robot in the railway line maintenance operation.

[0067] Optionally, the power supply battery is equipped with an intelligent power management system, which has the following advantages:

[0068] First, real-time monitoring of battery power, optimizing power usage through intelligent power management system, and ensuring balanced power supply for each functional module. Second, when the power supply battery is low, the robot can automatically return to the charging station for charging to ensure continuous operation. Third, overcharge protection, over-discharge protection, short circuit protection and temperature monitoring of the power supply battery ensure safe operation of the battery under various conditions.

[0069] In some embodiments of the present invention, the bottom side of the box body 2 is provided with guide wheels 8 for rolling cooperation with the rails.

[0070] In the above embodiment, the guide wheels 8 play an auxiliary positioning role in the movement of the box 2, ensuring the stable travel and accurate guidance of the robot on the railway track, and preventing the robot from falling off the rail. There are four guide wheels 8, and the four guide wheels 8 are all located between the two rails, and the four guide wheels 8 are matched with the two rails in pairs.

[0071] In some embodiments of the present invention, the box body 2 is further provided with a tool box 4.

[0072] In the above embodiment, the tool box 4 is designed to store various tools and spare parts required during the maintenance and inspection process, such as wrenches, screwdrivers, measuring instruments, etc., so as to support on-site immediate maintenance and inspection and improve work efficiency and flexibility.

[0073] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A railway intelligent inspection robot, characterized in that: include: A travel assembly, comprising a travel unit and a first drive assembly (11) that cooperate with each other, wherein the travel unit is used to adapt to a rail, and the first drive assembly (11) is used to drive the travel unit to move on the rail; A processing component, comprising a processor and a data storage unit, wherein the processor is used for receiving and processing data, and the data storage unit is used for storing data; A first data acquisition component comprises a binocular camera (5), a three-color warning light (6) and a display screen (3) which are simultaneously arranged on the walking unit, the binocular camera (5) being used to record the operation of the operator in the form of a video during the maintenance operation of the temporary railway skylight, and transmitting the operation in the form of a video to the processor for data processing, the three-color warning light (6) having a red warning part, a green warning part and a yellow warning part, and the display screen (3) being used to display a monitoring screen of the operation site; as well as The second data acquisition component includes a light sensor (13) for acquiring image information of the rails and fasteners, and transmitting the image information to the processor for data processing.

2. The railway intelligent inspection robot according to claim 1, characterized in that: The walking unit comprises a box body (2) and a walking wheel set (9) arranged on the box body (2), wherein the walking wheel set (9) is used to support the box body (2) and can drive the box body (2) to move on the rails; Wherein, the running wheel set (9) and the first driving assembly (11) are transmission connected.

3. The railway intelligent inspection robot according to claim 2, characterized in that: The binocular camera (5) and the three-color warning light (6) can both be rotatably arranged on the box (2); The box body (2) is provided with a first storage space (1) and a second storage space (7); the first storage space (1) is used to store the binocular camera (5); and the three-color warning light (6) is used to store the three-color warning light (6).

4. The railway intelligent inspection robot according to claim 3, characterized in that: The binocular camera (5) and the three-color warning light (6) are commonly connected to a rotating shaft (10), and the rotating shaft (10) is drivingly connected to a second drive assembly (12). When the second drive assembly (12) is in operation, the binocular camera (5) and the three-color warning light (6) are driven by the rotating shaft (10) to be stored in the first storage space (1) and the second storage space (7) respectively.

5. The railway intelligent inspection robot according to claim 4, characterized in that: The first drive assembly (11) and the second drive assembly (12) are both driven by drive motors.

6. The railway intelligent inspection robot according to claim 4, characterized in that: The inspection robot further comprises a power supply battery, wherein the power supply battery is used to supply power to the first drive assembly (11), the second drive assembly (12), the binocular camera (5), the three-color warning light (6), the display screen (3) and the light sensor (13).

7. The railway intelligent inspection robot according to claim 2, characterized in that: The bottom side of the box body (2) is provided with guide wheels (8) for rolling cooperation with the rails.

8. The railway intelligent inspection robot according to claim 2, characterized in that: The box body (2) is also provided with a tool box (4).

Images