Railway multi-source equipment adaptive association implementation method

Abstract

The invention provides a railway multi-source equipment self-adaptive association implementation method, which is characterized in that a water level gauge is arranged on a railway, when water overflows a rail surface, the water level gauge triggers an alarm, a self-adaptive association mechanism is automatically triggered, a nearby camera laser radar is linked according to the networking radius of the water level gauge to carry out auxiliary nuclear, and meanwhile, a nearby iron tower gunlock is triggered to carry out linkage auxiliary nuclear. According to the invention, automatic networking can be realized, self-adaptive association of multi-source equipment is supported, the influence of climate, environment and temperature is reduced, and the cost, the complexity and the maintenance cost are low.

Description

Technical Field

[0001] The invention relates to the field of train running environment safety monitoring, and in particular to a method for realizing adaptive association of railway multi-source equipment. Background Art

[0002] In the current device networking scenario, many devices are networked in a strongly associated way, which means that devices need to be bound or strongly associated through specific software. This networking method may be more applicable in certain application scenarios, such as specific industrial automation systems or specific monitoring networks. However, with the increase in the types and number of devices, and the diversification of device communication protocols, this strongly associated networking method has gradually exposed many problems and challenges.

[0003] Structural analysis 1) Device binding: In a strongly associated device networking mode, devices need to be bound to each other. This binding may be a physical connection or a logical connection achieved through specific software or protocols. The binding process is usually more complicated and requires specific equipment and software support, which increases the difficulty and cost of networking.

[0004] 2) Communication protocol: Different devices may use different communication protocols, which makes communication between devices difficult. In order to achieve interoperability between devices, it may be necessary to use devices or software such as gateways and converters to convert protocols, which undoubtedly increases the complexity and cost of networking.

[0005] 3) Network topology: Strongly associated device networking usually has a fixed network topology. When the number of devices increases or the network structure changes, it may be necessary to re-bind and reconfigure the devices, which increases the complexity and difficulty of network maintenance.

[0006] Principle Analysis The strongly associated device networking method mainly relies on direct communication and binding relationships between devices. In this networking method, communication and data transmission between devices need to go through multiple intermediate links, such as gateways, converters, etc., which increases the delay and error rate of data transmission. At the same time, since the binding relationship between devices is relatively fixed, when a device fails or needs to be replaced, the entire network may need to be reconfigured and debugged, which increases the complexity and cost of network maintenance.

[0007] Problems and shortcomings of the existing technology: 1. Difficulties in networking: Due to the increase in the types and number of devices, as well as the diversification of device communication protocols, the strongly associated device networking method faces many difficulties in the networking process. It takes a lot of time and manpower to bind and configure devices, which increases the complexity and cost of networking.

[0008] 2. High cost: In order to achieve interoperability between devices, it may be necessary to use devices or software such as gateways and converters to convert protocols. The prices of these devices and software are usually high, which increases the overall cost of networking.

[0009] 3. Complex maintenance: The strongly associated device networking mode has a fixed network topology. When the network structure changes or the device fails, complex network maintenance and debugging work is required, which increases the complexity and difficulty of network maintenance. Summary of the invention

[0010] In order to solve the problems of the prior art, the present invention provides a method for realizing adaptive association of multi-source equipment in railways, which can realize automated networking, support adaptive association of multi-source equipment, reduce the influence of climate, environment and temperature, and has low cost, low complexity and low maintenance cost.

[0011] The present invention sets a water level meter on the railway. When water overflows the rail surface, the water level meter triggers an alarm, automatically triggers an adaptive association mechanism, and links nearby camera laser radars to perform auxiliary core according to the network radius of the water level meter. At the same time, it triggers the linkage auxiliary core of the nearby tower gun machine. The adaptive association mechanism is specifically as follows: 1) Determine the association relationship: Based on the pictures or videos provided by the two nearest cameras, analyze the common area of ​​the cameras and automatically associate them; 2) Camera control parameter association calculation: Based on the common area and target calculation, the corresponding relationship between the two camera rotations is provided; 3) Single camera specific target area association calibration: Place a specific visual target in a specific area, control the camera to scan, identify the area range when the visual target is at the center of the area, and output the rotation parameters and shooting area parameters of the corresponding camera; 4) Camera and point cloud scene fusion calculation: When the camera area intersects with the lidar point cloud area, the laser integrated adaptive association is automatically triggered, and the camera scene analysis is integrated with the lidar point cloud analysis for fusion judgment analysis; 5) Based on the radius of the device that triggers the alarm, the devices within the range are automatically circled and networked.

[0012] As a further improvement, the automatic association in step 1) requires that one camera can provide another camera with a video or picture in a manually-calibrated manner or provide visual recognition correction from another perspective.

[0013] For further improvement, the rotation correspondence in step 2) specifically requires that the rotation parameters of camera B when shooting the same area can be quickly calculated based on the rotation parameters of camera A. The parameters include the camera starting angle, rotation direction, and rotation degree. The rotation correspondence mainly includes the orientation of camera A and camera B, the degree of rotation, and whether the two can capture the same scene area.

[0014] As a further improvement, in the process of the association calibration of the specific target area of ​​the single camera in step 3), a water level meter is used for anchoring to complete the adaptive association fusion analysis of the associated tower gun camera and the ball camera plus the water level meter.

[0015] The beneficial effects of the present invention are: 1. Visualized network configuration, easy to operate: Through visual interfaces and tools, users can intuitively configure and manage the network structure of railway equipment without having to deeply understand complex network protocols and configuration commands. Visual configuration can greatly reduce operational errors, improve configuration efficiency, and reduce maintenance costs.

[0016] 2. Automated networking can be achieved based on policies and algorithm models: Using preset strategies and algorithm models, the system can automatically optimize and adjust the network based on factors such as device status and network load. Automated networking can ensure the stability and reliability of the network, reduce the need for human intervention, and reduce the failure rate.

[0017] 3. Multi-source networking: Supports adaptive association of multi-source devices: The invention supports adaptive association of multiple brands and models of equipment, improving the compatibility and flexibility of the equipment. Multi-source networking can meet the networking requirements in different scenarios, making the railway equipment network more flexible and scalable.

[0018] 4. Reduce the impact of climate, environment and temperature, and provide high-quality networking in all climates: By optimizing the network protocol and hardware design, the invention creates a network performance that can maintain stability under various climate and environmental conditions. All-weather high-quality networking can reduce network failures and missed and false alarms caused by climate and environmental factors, and improve network reliability and availability.

[0019] 5. Low cost: Compared with the traditional railway equipment networking monitoring solution, the invention adopts more efficient and economical networking technology and equipment, reducing the overall cost. The cost reduction can enable more railway projects to adopt advanced networking monitoring technology and improve the safety and efficiency of railway operations.

[0020] 6. Reduced complexity: The invention reduces the complexity of the system by simplifying the network structure and configuration process. The reduction of complexity can reduce the failure rate and maintenance cost of the system and improve the stability and reliability of the system.

[0021] 7. Unified products, reduced maintenance costs: The invention adopts unified product standards and interface specifications, so that different devices and systems can be seamlessly connected and integrated. Product unification can reduce maintenance costs, improve maintenance efficiency, and reduce problems caused by incompatible devices or mismatched interfaces. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0023] Figure 1 A schematic diagram of the system architecture.

[0024] Figure 2 A schematic diagram of the application scenario.

[0025] Figure 3 Schematic diagram of a local scene for adaptive association application.

[0026] Figure 4 Schematic diagram of lidar point cloud coverage. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0028] The system used in the present invention is as follows Figure 1 As shown, the details are as follows: Cloud platform, edge computing (adaptive association program is deployed by default), front-end perception devices.

[0029] 1. The cloud platform is responsible for all northbound data access, southbound data collection, comprehensive analysis, and feedback to edge computing.

[0030] 2. Edge computing is responsible for the access of all southbound device data, including the analysis of various protocol data, such as MQTT direct access, HTTP access, MQTT BROKER access, TCP access, UDP access, WEBSOCKET access, gateway device access, plug-in device access, channel accumulation device access, etc.

[0031] Edge computing combines the characteristics of different devices, parses multi-protocol data, and performs scenario-based adaptive association.

[0032] Front-end sensing devices include lidar, cameras (tower cameras, ball cameras, gun cameras), water level meters, crack meters, Beidou, millimeter-wave radars, micro-meteorological stations, etc.

[0033] The present invention works as follows Figure 2-4 As shown, specifically: 1) Determine the association relationship: Based on the surround scene (image or video) of the two cameras, analyze the common area of ​​the cameras and automatically associate them (determine whether one camera can provide a video / image for manual verification by another camera, or provide visual recognition correction from another perspective).

[0034] 2) Camera control parameter association calculation: Based on the public area / target calculation, the rotation correspondence between the two cameras can be provided (after camera A takes a picture of the corresponding public area, the rotation parameters of camera B when taking the same area can be quickly calculated based on the rotation parameters of camera A). The parameters generally include the camera starting angle, rotation direction, rotation degree, etc. The rotation correspondence mainly includes the direction of camera A and camera B, the degree of rotation, etc., and the two can capture the same scene area.

[0035] 3) Single camera specific target area association calibration: Place a specific visual target (such as a calibration balloon, etc.) in a specific area, control the camera to scan, identify the area range when the visual target is in the center of the area, and output the corresponding camera's rotation parameters and shooting area parameters (length, width, image resolution, etc.). When used to deploy event sensors, bind the camera used to take pictures for manual verification.

[0036] 4) Camera and point cloud scene fusion calculation: When the area controlled by the camera intersects with the lidar point cloud area, the laser integrated adaptive association is automatically triggered, and fusion judgment and analysis are performed.

[0037] 5) Based on the radius of the device that triggers the alarm, the devices within the range are automatically circled and networked.

[0038] This function is used to manually check multi-angle videos or pictures of the same scene, as well as visually fuse multi-source data of the same scene (to improve visual recognition accuracy).

[0039] A specific implementation method of adaptively associating multiple source devices is as follows: Figure 2-4 As shown: Adaptive association of multiple source devices 1. Adaptive association of flooded track surface: Tower gun camera, water level meter, laser radar, ball camera 2. Network content description: The "adaptive association" method mainly uses vision + radar point cloud + pre-determined device location for association calculation and judgment. It includes: vision and vision in a specific area (discovering a common area), vision and point cloud (discovering a common area), vision and sensor devices of a specific shape (visual discovery of long sensor devices), and visual discovery of a specific area (specific sensor devices in the area).

[0040] 3. Instructions for use of networking equipment: When water overflows the track surface and the water level meter triggers an alarm, it is sent to the edge computing adaptive association program, which automatically triggers the adaptive association mechanism, and links the nearby camera lidar for auxiliary core according to the networking radius of the water level meter; at the same time, it will trigger the linkage auxiliary core of the nearby tower gun; multiple cameras will trigger algorithm association calculation for public areas and automatically network; at the same time, the lidar point cloud intersects with the camera area, triggering laser integrated fusion analysis.

[0041] (1) Water level meter triggers alarm: Enable adaptive correlation radius to trigger all devices in the nearby range for fusion analysis; (2) Pure video public area fusion analysis: ball cameras and tower gun cameras trigger pure video fusion analysis; the specific implementation method can be based on feature matching algorithms (SIFT, SURF, ORB, etc.), image registration (feature-based methods, frequency domain methods), deep learning methods (convolutional neural networks, Siamese Network), optical flow methods, and image stitching.

[0042] (3) For the single camera association area, use a water level meter for anchoring and complete the adaptive association fusion analysis of the associated tower gun camera / ball camera and water level meter.

[0043] (4) Laser integration: Camera scene analysis is integrated with LiDAR point cloud analysis to trigger laser integration adaptive correlation fusion analysis. The specific implementation method is based on feature extraction and matching (LiDAR point cloud feature extraction, image feature extraction, feature matching), external parameter calibration (calibration plate-based method, automatic calibration method), image to point cloud projection (projection matrix, registration verification), and deep learning methods (multimodal fusion network, contrastive learning).

[0044] 4. Visual configuration: Provide page drag and drop visual configuration, drag the device distribution position, set the radius; divide the area under the jurisdiction of the device; shield the device-related area; start and stop the device; adaptively associate the start and stop operation, etc.

[0045] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, the above is only a preferred embodiment of the present invention. Since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited to this. Any technician familiar with the technical field is within the technical scope disclosed by the present invention. For ordinary technicians in the technical field, changes or replacements that can be easily thought of should be covered within the protection scope of the present invention without departing from the principle of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for implementing adaptive association of railway multi-source equipment, characterized by: A water level meter is installed on the railway. When water overflows the track surface, the water level meter triggers an alarm and automatically triggers the adaptive association mechanism. The nearby camera laser radar is linked to the network radius of the water level meter for auxiliary core, and the nearby tower gun camera is triggered to link the auxiliary core. The adaptive association mechanism is as follows: 1) Determine the association relationship: Based on the pictures or videos provided by the two nearest cameras, analyze the common areas of the cameras and automatically associate them; 2) Camera control parameter association calculation: Based on the common area and target calculation, the corresponding relationship between the two camera rotations is provided; 3) Single camera specific target area association calibration: Place a specific visual target in a specific area, control the camera to scan, identify the area range when the visual target is at the center of the area, and output the rotation parameters and shooting area parameters of the corresponding camera; 4) Camera and point cloud scene fusion calculation: When the camera area intersects with the lidar point cloud area, the laser integrated adaptive association is automatically triggered, and the camera scene analysis is integrated with the lidar point cloud analysis for fusion judgment analysis; 5) Based on the radius of the device that triggers the alarm, the devices within the range are automatically circled and networked.

2. The method for realizing adaptive association of railway multi-source equipment according to claim 1, characterized in that: The automatic association in step 1) requires that a camera be able to provide another camera with a video or picture in a manually-calibrated manner or provide visual recognition correction from another perspective.

3. The method for realizing adaptive association of railway multi-source equipment according to claim 1, characterized in that: The rotation correspondence in step 2) specifically requires that the rotation parameters of camera B when shooting the same area can be quickly calculated based on the rotation parameters of camera A. The parameters include the camera starting angle, rotation direction, and rotation degree. The rotation correspondence mainly includes the orientation of camera A and camera B, the degree of rotation, and whether the two can capture the same scene area.

4. The method for realizing adaptive association of railway multi-source equipment according to claim 1, characterized in that: In step 3), during the association calibration process of the specific target area of ​​the single camera, a water level meter is used for anchoring to complete the adaptive association fusion analysis of the associated tower gun camera and ball camera plus the water level meter.

Images