A railway clearance detection system based on multi-radar fusion

Abstract

This utility model discloses a railway clearance detection system based on multi-radar fusion, relating to the railway transportation field. It comprises a lidar module, a wiring device, a power supply module, a switch, a positioning module, and a processing host. The lidar module includes three 128-line high-beam lidars, which are asymmetrically mounted on the roof of the inspection vehicle using customized brackets. These lidars provide high-precision distance measurement capabilities and detailed point cloud 3D information. Data collected by the three different lidars is transmitted to the wiring device on the roof, and then to the switch. The switch transmits the data via network cable to the intelligent analysis and processing host inside the vehicle. The host, located inside the inspection vehicle, runs software and algorithms to achieve clearance detection. This product is installed on the inspection vehicle and performs online clearance detection as the vehicle moves, meeting real-time detection requirements. It boasts high detection efficiency, a wide detection range, strong anti-interference capabilities, and maintains good working condition under various climatic conditions.

Description

Technical Field

[0001] This utility model relates to the field of railway transportation, and in particular to a railway clearance detection system based on multi-radar fusion. Background Technology

[0002] As a vital pillar of economic development, railway transportation's safety and reliability are crucial for ensuring stable economic operation. To guarantee safe operation, a comprehensive inspection of the actual clearance conditions along the entire railway line must be conducted before trains are put into service. This places higher demands on the efficiency and accuracy of railway clearance inspection to ensure the normal operation of railway transportation. However, traditional clearance inspection methods, such as manual measurement relying on basic tools like track gauges and theodolites, and non-contact measurement based on video cameras and handheld laser rangefinders, suffer from numerous drawbacks, including low efficiency, high risk, and significant susceptibility to environmental influences, and are no longer adequate for the rapidly developing needs of modern railways.

[0003] Existing non-contact clearance detection solutions in the railway sector mostly rely on laser sensors for system design.

[0004] A common type of railway clearance inspection vehicle on the market mainly consists of a mechanical motion structure, a computer, and a laser scanner. The vehicle is remotely controlled to move along the railway tracks, and the laser scanner slowly scans the railway clearance along the way, generating clearance data.

[0005] Another design scheme for a clearance detection device in the prior art consists of multiple laser sensors. The device is installed at the front of the vehicle and performs a 360° cross-sectional scan of the clearance in front of the vehicle as it moves.

[0006] The main problems with existing testing methods are:

[0007] (1) Low efficiency and safety hazards. The limit measurement method based on laser scanner usually uses laser rangefinder for detection. This method has high accuracy but low efficiency, long operation cycle and relies on manual operation of equipment, which poses a great safety hazard;

[0008] (2) Difficult to install and large storage requirements. The 360° method requires a detection device to be installed at the front of the vehicle, which is complicated to install and has a certain risk of the device falling. In addition, it can only scan the clearance data at the front of the vehicle and cannot provide front clearance warning. The shooting method of 5-6 cameras + point cloud has a large data storage, is sensitive to vibration interference, and is difficult to stitch together in the later stage. Utility Model Content

[0009] The purpose of this invention is to provide a railway clearance detection system based on multi-radar fusion, which offers high detection efficiency to meet the needs of modern railways for rapid and efficient inspection; a wide detection range capable of collecting clearance information over a large area ahead of the railway; and low installation and usage difficulty, allowing it to be installed on inspection vehicles such as the JX300, integrating locomotive positioning and compensation information.

[0010] The technical solution proposed in this utility model is implemented as follows: A railway clearance detection system based on multi-radar fusion, characterized in that it consists of a lidar module, a wiring device, a power supply module, a switch, a positioning module, and a processing host. The lidar module includes three 128-line high-beam lidars, which are mounted on the roof of the inspection vehicle using customized brackets in an asymmetrical manner with 30° upward, 30° horizontal, and 30° downward views respectively. The lidar module is connected to the wiring device. The data acquisition end of the wiring device is connected to the lidar module, the data transmission end is connected to the switch, and the power supply end is connected to the 24V DC voltage output end of the power supply module. The switch includes a roof-mounted switch and an in-vehicle switch. The roof-mounted switch, used to collect the raw point cloud data from the lidar module, is located in a waterproof shell in the middle of the roof. One end is connected to the wiring device, and the other end is connected to the in-vehicle switch located inside the vehicle. The in-vehicle switch is connected to the processing host located inside the vehicle. The positioning module is installed inside the inspection vehicle and includes a wheel encoder, a gyroscope, and a Beidou satellite ground receiver. The positioning information output end of the positioning module is connected to the processing host. The processing host is a high-performance server equipped with multiple point cloud processing algorithms, capable of performing relevant calculations on the three-dimensional information of the surrounding space, detecting according to clearance standards, and calculating the clearance values ​​of surrounding targets from the center point of the railway track. The processing host is connected to the positioning module and the in-vehicle switch. The power supply module includes a rectifier and filter circuit composed of multiple transistors, resistors, and capacitors. The input terminal of the power supply module is connected to a 220V AC power supply, and the output terminal with a 24V DC voltage is connected to the wiring device.

[0011] The railway clearance detection system based on multi-radar fusion provided by this utility model has higher detection efficiency: the product is installed on the inspection vehicle and follows the vehicle to perform online clearance detection, and the radar acquisition frame rate reaches 10fps per second, which can meet the requirements of real-time detection; the detection range is wider: the multi-radar fusion detection system of this invention breaks through the limitations of traditional fixed perspective and realizes full information collection of the railway line environment; and the anti-interference is stronger: it can still maintain a good working state in rainy or sunny weather. Attached Figure Description

[0012] Figure 1 System hardware structure block diagram;

[0013] Figure 2 : Installation diagram of lidar module;

[0014] Figure 3 Schematic diagram of the detection range of the lidar module;

[0015] Figure 4 Wiring diagram of the wiring device, power supply module, and radar;

[0016] Figure 5 Switch topology diagram;

[0017] Figure 6 : Positioning module topology diagram;

[0018] Figure 7 Power supply module circuit diagram Detailed Implementation

[0019] The technical solution and advantages of this utility model will be further explained below with reference to the accompanying drawings and embodiments.

[0020] Railway clearance refers to the maximum permissible distance from the centerline of a railway line to locomotives, rolling stock, or other equipment and structures. Accurate detection of railway clearance is crucial for ensuring railway operational safety. This system, based on multi-radar fusion technology, aims to achieve high-precision, real-time railway clearance detection, promptly identify clearance anomalies, and provide strong support for safe railway operations.

[0021] The hardware components of this system mainly include: a lidar module, wiring devices, a power supply module, a switch, a positioning module, and a processing host. The software installed and configured on the processing host is responsible for data processing and algorithm analysis of the data acquired by the hardware, and provides human-computer interaction functionality.

[0022] (1) LiDAR module: LiDAR has high-precision distance measurement capability and can provide detailed point cloud three-dimensional information.

[0023] The system combines three 128-line high-beam lidars, achieving a single-frame point cloud density of ≥100,000 points, a maximum detection distance of ≤100m, a minimum detection distance of ≥0.3m, a ranging accuracy of ≤2cm at distances of 20m-50m, a ranging resolution of 0.1cm, a vertical field of view of 45°, a horizontal field of view of 360°, a laser wavelength of 865nm, and a laser safety level that meets human eye safety standards. The lidar's measurement accuracy reaches 2cm, exceeding the accuracy requirements of clearance detection equipment in railways, and maintains high ranging accuracy even at long distances, providing reliable assurance for railway operational safety.

[0024] Table 1. LiDAR Technical Specifications

[0025]

[0026] The three radars are asymmetrically mounted on the roof of the inspection vehicle using customized brackets and an exterior design, such as... Figure 2 As shown, the devices are mounted on the roof at 30° upward, eye level, and 30° downward angles, respectively. The detection range covers an area up to 10 meters from the vehicle track to the air. Figure 3 As shown.

[0027] Upward-looking radar: 30° elevation angle, covering the airspace up to 10m above the vehicle track, detecting overhead facilities such as overhead contact lines and bridges;

[0028] Normal radar: horizontally installed, covering a horizontal range of 10m on both sides of the track, detecting lateral facilities of platform and tunnel wall lights;

[0029] Downward-looking radar: 30° downward angle, covering the area below the track, detecting foreign objects on the track surface, etc.

[0030] The radar directly synchronizes time between multiple radars with the PTP clock source through the hardware system, providing configuration accuracy between radar point cloud data.

[0031] (2) Wiring device: As a key connection node, it realizes system power distribution, signal transfer and hardware protection. Waterproof aviation plugs are used to connect the radar to the switch to ensure the stability and durability of the system.

[0032] (3) Power supply module: The power supply module installed on the roof provides power to the equipment on the roof, ensuring the stable operation of the radar. The equipment inside the vehicle, such as the host, server, in-vehicle switch, positioning module, etc., is powered by the inside of the detection vehicle.

[0033] The wiring relationships between the power supply module, wiring device, and radar are as follows: Figure 4 As shown, the voltage conversion circuit schematic from the power supply module to the wiring device is as follows: Figure 7 As shown;

[0034] (4) Switch: Dual-switch topology, such as Figure 5 As shown, it includes two parts: an in-vehicle switch and a roof switch, which build a local area network between devices and connect the roof devices with the main unit and other devices in the vehicle.

[0035] The rooftop switch is housed in a waterproof enclosure in the middle of the roof, and aggregates the raw point cloud data of the radar system through a waterproof aviation connector; the in-vehicle switch is located inside the vehicle, which can effectively avoid electromagnetic interference and ensure network transmission efficiency and system real-time performance.

[0036] (5) Positioning module: installed inside the testing vehicle to provide positioning information for various systems and devices on the vehicle.

[0037] A single information source has the drawback of high dependence. When the data source is unstable, an auxiliary data source can effectively maintain the stability of the location data.

[0038] This module uses a comprehensive positioning method, and the topology is as follows: Figure 6 As shown, the system obtains information such as the route and distance during driving by using basic vehicle data and wheel encoders; it locates the current latitude and longitude, vehicle speed, acceleration, and other information using BeiDou and gyroscopes; it obtains the pole number of the current vehicle using machine vision technology; and it achieves high-precision positioning in all scenarios through data fusion and judgment algorithms.

[0039] (6) Intelligent analysis and processing host: installed inside the inspection vehicle, using a high-performance computing server as the core hardware, receiving data collected by the hardware, processing raw data, running intelligent algorithms to generate inspection results, and providing a human-machine interface that meets the needs of inspection personnel.

[0040] The server performance of this system can meet the requirements of parallel processing and deep learning inference, ensuring that point cloud-related processing algorithms can be processed in real time; it also has industrial-grade protection performance, is resistant to high temperatures and impacts, and is suitable for long-term stable operation on the inspection vehicle.

[0041] The operation between systems can be divided into four layers: data acquisition layer, data transmission layer, data processing layer, and human-computer interaction layer.

[0042] Data acquisition layer: Three radars scan simultaneously in time and space to generate raw point cloud data, providing the three-dimensional spatial coordinates and timestamps of the surrounding environment; the positioning module outputs the location information of the detection vehicle and station area information in real time.

[0043] Data transmission layer: The rooftop switch aggregates radar data and transmits it to the in-vehicle host unit.

[0044] Data processing layer: Based on a variety of point cloud processing algorithms, it performs relevant calculations on the three-dimensional information of the surrounding space. It can not only detect the boundary standards, but also calculate the boundary values ​​of surrounding objects from the center point of the railway track.

[0045] Human-computer interaction layer: Online software provides a convenient interactive interface and gives warnings when limits are exceeded.

Claims

1. A railway clearance detection system based on multi-radar fusion, characterized in that: It consists of a lidar module, wiring device, power supply module, switch, positioning module and processing host. The lidar module includes three 128-line high-beam lidars. The three lidars are mounted on the roof of the inspection vehicle in an asymmetrical manner with customized brackets, with a 30° upward angle, a 30° horizontal angle and a 30° downward angle respectively. The lidar module is connected to the wiring device.

2. The railway clearance detection system based on multi-radar fusion according to claim 1, characterized in that: The data acquisition end of the wiring device is connected to the lidar module, the data transmission end is connected to the switch, and the power supply end is connected to the 24V DC voltage output end of the power supply module.

3. The railway clearance detection system based on multi-radar fusion according to claim 1, characterized in that: The switch includes a roof-mounted switch and an in-vehicle switch. The roof-mounted switch, used to collect raw point cloud data from the lidar module, is housed in a waterproof casing in the middle of the roof. One end is connected to a wiring device, and the other end is connected to an in-vehicle switch located inside the vehicle. The in-vehicle switch is connected to a processing host located inside the vehicle.

4. The railway clearance detection system based on multi-radar fusion according to claim 1, characterized in that: The positioning module is installed inside the inspection vehicle. The positioning module includes a wheel encoder, a gyroscope, and a Beidou satellite ground receiver. The positioning information output of the positioning module is connected to the processing host.

5. The railway clearance detection system based on multi-radar fusion according to claim 1, characterized in that: The processing host is a high-performance server equipped with multiple point cloud processing algorithms, capable of performing relevant calculations on the three-dimensional information of the surrounding space, detecting according to clearance standards, and calculating the clearance value of surrounding objects from the center point of the railway track. The processing host is connected to the positioning module and the in-vehicle switch.

6. The railway clearance detection system based on multi-radar fusion according to claim 1, characterized in that: The power supply module includes a rectifier and filter circuit consisting of multiple transistors, resistors, and capacitors. The input terminal of the power supply module is connected to a 220V AC power supply, and the output terminal that outputs 24V DC voltage is connected to the wiring device.

Images