Rail operation vehicle auxiliary obstacle avoidance early warning device based on laser radar and image fusion

Abstract

The utility model relates to the technical field of operation safety on a railway track operation vehicle line, in particular to a track operation vehicle auxiliary obstacle avoidance early warning device based on laser radar and image fusion, which comprises two sets of environment sensing systems arranged at the front end and the rear end of a track operation vehicle. Wherein each set of environment sensing system comprises two cameras, a close-range blind compensation radar and a foresight long-distance radar which are arranged at the end part of the track operation vehicle, and a camera and a close-range blind compensation radar which are respectively arranged at the left side and the right side of the track operation vehicle; the environment sensing system is connected to the domain controller, and the domain controller is connected with an alarm device. The utility model has the advantages that: when a large machine works under different working conditions, early warning information is obtained through the man-machine interaction interface, and corresponding judgment is made in time; the personnel and equipment safety in the vehicle driving process is ensured at any time; and meanwhile, detection under various illumination and weather conditions is considered, so that the safety of the engineering operation environment is comprehensively improved.

Description

Technical Field

[0001] This utility model relates to the field of railway track maintenance vehicle operation safety technology, specifically to an auxiliary obstacle avoidance and early warning device for track maintenance vehicles based on lidar and image fusion. Background Technology

[0002] Large railway track maintenance machinery (referred to as "large machines"), such as ballast shaping cars, stabilizing cars, and tamping cars, generally lack devices for detecting obstacles during operation, often relying on manual observation. The presence of numerous construction tools from various trades and personnel from multiple units around the operating vehicles creates a complex human-machine-environment system. Due to blind spots in the large machines' field of vision, drivers manually check real-time video images to confirm the surrounding environment; however, this observation is limited by environmental interference and the driver's subjective perception. This leads to collisions between various large machines and obstacles during operation. Summary of the Invention

[0003] The purpose of this invention is to address the shortcomings of the existing technology by providing an auxiliary obstacle avoidance and early warning device for railway track maintenance vehicles based on lidar and image fusion. By combining visible light and lidar sensors, the perception system can obtain basic information about the surrounding environment and detect obstacles on the railway track maintenance vehicle lines that require human-machine collaboration under different working conditions. This ensures the safety of personnel and equipment during vehicle operation, effectively improving both the safety and operational efficiency of railway track maintenance vehicles.

[0004] The objective of this utility model is achieved through the following technical solution:

[0005] An auxiliary obstacle avoidance and early warning device for a rail maintenance vehicle based on lidar and image fusion is characterized by comprising two sets of environmental perception systems installed at the front and rear ends of the rail maintenance vehicle. Each set of environmental perception systems includes two cameras, a near-range blind spot radar, and a forward-looking long-range radar installed at the end of the rail maintenance vehicle, as well as a camera and a near-range blind spot radar installed on the left and right sides of the rail maintenance vehicle, respectively. The environmental perception system is connected to a domain controller, and the domain controller is connected to an alarm device.

[0006] The alarm device is an alarm sound and image output device.

[0007] The camera is a wide-angle high-definition camera.

[0008] The near-range blind spot filling radar uses a wide-angle lidar.

[0009] The forward-looking long-range radar is a long-range lidar.

[0010] The advantages of this invention are: based on basic environmental information, the system integrates and co-computes information from various sensors, and combines artificial intelligence, big data, image processing and other technologies to achieve the ability to obtain early warning information and make timely judgments through a human-machine interface when large machinery is operating under different working conditions; to ensure the safety of personnel and equipment during vehicle operation at all times; and to take into account the detection under various lighting and weather conditions, thereby comprehensively improving the safety of the engineering operation environment. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of the overall layout of this utility model;

[0012] Figure 2 This is a schematic diagram of the mid-end environmental sensing system of this utility model;

[0013] Figure 3 This is a schematic diagram illustrating the detection effect of this utility model;

[0014] Figure 4 This is a flowchart illustrating the early warning process during the use of this utility model. Detailed Implementation

[0015] The features and other related features of this utility model will be further described in detail below with reference to the accompanying drawings and embodiments, so as to facilitate the understanding of those skilled in the art:

[0016] like Figure 1-4 As shown in the figure, numbers 1-7 represent: Domain Controller 1, Wide-angle HD Camera 2, Wide-angle LiDAR 3, Long-range HD Camera 4, Long-range LiDAR 5, Alarm Sound and Image Output Device 6, and GYK Control Device 7.

[0017] Example: Figures 1 to 4 As shown, in this embodiment, the obstacle avoidance and early warning device for the track maintenance vehicle based on lidar and image fusion is equipped with an environmental perception system at both the front and rear ends of the track maintenance vehicle. The environmental perception system includes lidar and a camera for capturing images.

[0018] Specifically, with Figure 1For example, in the illustrated direction, with domain controller 1 as the center, there are two environmental perception systems on the left and right sides respectively. Each environmental perception system mainly consists of four cameras, three near-range blind-spot radars (detection range 0-10m) (front, left, and right), and one forward-looking long-range radar (effective detection range 10m-80m). Each environmental perception system consists of three parts: left, front, and right. The left and right parts each consist of a wide-angle high-definition camera 2 and a wide-angle lidar 3 serving as a near-range blind-spot radar, used to detect the environment on both sides of the vehicle and the status of the vehicle itself. The front sensors consist of a wide-angle high-definition camera 2, a long-range high-definition camera 4, a wide-angle lidar 3 serving as a near-range blind-spot radar, and a long-range lidar 5 serving as a forward-looking long-range radar, used for real-time perception of the environment in front of the vehicle.

[0019] The environmental perception system in this embodiment represents the optimal solution. However, in practical applications, customized perception solutions can be adopted to meet the specific needs of different working conditions. The accompanying perception algorithm can integrate the characteristics of multiple sensors, avoiding the disadvantages of a single sensor and improving the ability to perceive the railway environment.

[0020] In this embodiment, to accurately perceive the complex environment surrounding the track and improve the driving safety of the track maintenance vehicle, a combination of multiple sensors is used to achieve precise and panoramic perception of the complex external environment. By using a near-range blind-spot radar (wide-angle lidar 3) and a forward-looking long-range lidar (far-seeing lidar 5), the overall perception capability of the sensing system can be effectively improved, ensuring good performance at both near and long distances. In use, the perception information obtained through the image vision sensor can be further processed by a convolutional neural network for feature extraction to obtain information such as the category, location, and confidence level of the image target. This is then supplemented by depth image conversion using sensor parameters to obtain basic environmental geometric information.

[0021] During operation, domain controller 1 collects data from various sensors for analysis. If an obstacle is detected, domain controller 1 immediately controls alarm sound and image output device 6 to issue an alarm message. Furthermore, domain controller 1 can connect to GYK control equipment 7, and in some situations, it can also control the railcar to stop automatically if an obstacle is encountered.

[0022] In this embodiment, the detection of targets within a designated space surrounding the rail-operated vehicle is carried out. Specifically, this includes establishing effective rules for determining hazardous spaces for rail workers based on the geometric information of the vehicle and the rails in the direction of travel, designing a point cloud information filtering mechanism in the relevant space, and developing a target clustering or detection algorithm model that meets both detection accuracy and operational efficiency, taking into account the large number of lidar sensors and the large amount of data in the detection system.

[0023] To facilitate the acquisition of perception results by the driver of the rail-operated vehicle, the system will visualize the perception results using RVIZ under ROS after the sensor data is collected. The human-machine interface is designed to display the data in the form of a grid map, and includes a function to switch between real-time sensor data visualizations, allowing for easy switching between different sensor contents and the overall grid map, thus facilitating the driver's access to perception results and real-time information. Sensor data will be saved for subsequent local offline analysis. During actual system operation, there may be situations where it is necessary to save sensor data, such as saving alarm information and sensor information at the time of alarm for subsequent data analysis and processing.

[0024] Although the above embodiments have described the concept and embodiments of the present invention in detail with reference to the accompanying drawings, those skilled in the art will recognize that various improvements and modifications can still be made to the present invention without departing from the scope of the claims, and therefore will not be elaborated here.

Claims

1. A track maintenance vehicle obstacle avoidance and early warning device based on lidar and image fusion, characterized in that: The system includes two sets of environmental perception systems installed at the front and rear ends of the track maintenance vehicle. Each set of environmental perception systems includes two cameras, one near-range blind spot radar, and one forward-looking long-range radar installed at the end of the track maintenance vehicle, as well as one camera and one near-range blind spot radar installed on the left and right sides of the track maintenance vehicle, respectively. The environmental perception systems are connected to a domain controller, and the domain controller is connected to an alarm device.

2. The obstacle avoidance and early warning device for a rail vehicle based on lidar and image fusion as described in claim 1, characterized in that: The alarm device is an alarm sound and image output device.

3. The obstacle avoidance and early warning device for a rail vehicle based on lidar and image fusion as described in claim 1, characterized in that: The camera is a wide-angle high-definition camera.

4. The obstacle avoidance and early warning device for a rail vehicle based on lidar and image fusion as described in claim 1, characterized in that: The near-range blind spot filling radar uses a wide-angle lidar.

5. The obstacle avoidance and early warning device for a rail vehicle based on lidar and image fusion as described in claim 1, characterized in that: The forward-looking long-range radar is a long-range lidar.

Images