A rail train anti-collision recognition device and method in an unmanned state

Abstract

The application provides a rail train anti-collision identification device and method in an unmanned state, relates to the technical field of rail transit test, and comprises a data acquisition system, an intelligent detection system and a positioning system. A driver's room structure is installed on one side of a vehicle body, and a positioning system is arranged on a chassis side beam. Three-dimensional imaging data around the train body is acquired, and complete railway network big data detection clusters are formed by storage. The railway network big data detection clusters are trained by an artificial intelligence machine learning algorithm, and a railway train intruder judgment prediction model is obtained. Real-time intruders are detected, and the results after detection are comprehensively judged. The pre-judgment and response of the railway system intruders are realized. The application effectively solves the technical problems that the existing high-speed train is difficult to detect foreign matter intrusion and autonomously avoid obstacles in real time, and provides an important technical support and implementation approach for the full-scene detection capability of the rail vehicle under high-speed operation.

Description

Technical Field

[0001] This invention relates to the field of rail transit testing technology, and more specifically, to a collision avoidance identification device and method for rail trains in an unmanned driving state. Background Technology

[0002] With the rapid development of railway transportation systems, train operation safety has become a focal point of concern in the railway industry. Current technologies primarily rely on traditional sensors and manual monitoring methods to detect intrusions during train operation. These methods suffer from limitations such as limited detection range, insufficient real-time performance, and high false alarm rates. For example, traditional track inspection systems may use simple switch sensors to detect foreign objects on the track, but these systems often only detect objects in direct contact with the track, and are ineffective against non-contact or smaller intrusions.

[0003] Furthermore, manual monitoring is limited by human resources, blind spots, and response time, failing to meet the high efficiency and reliability requirements of modern railway systems. Existing technologies also lag behind in data processing and analysis, typically lacking effective algorithms to extract useful information from large amounts of sensor data, resulting in inaccurate classification, location, and risk assessment of intruders. Summary of the Invention

[0004] The purpose of this invention is to provide a collision avoidance identification device and method for unmanned rail trains to improve the aforementioned problems. To achieve the above objective, the technical solution adopted by this invention is as follows:

[0005] In a first aspect, this application provides a collision avoidance identification device for rail trains in an unmanned driving state, comprising:

[0006] The vehicle comprises a data acquisition system, an intelligent detection system, and a positioning system. The data acquisition system includes the vehicle body, with a driver's cab structure mounted on one side. A positioning system is located on the side beams of the vehicle body's underframe, and bogies are mounted on the left and right ends of the lower side of the vehicle body. The positioning system includes GPS positioning and 3D digital map imaging technology. The intelligent detection system includes machine learning-based artificial intelligence detection algorithms. A zoom camera is installed inside the windshield of the driver's cab structure, and a thermal imager is installed outside the windshield. A pitot tube is installed below the thermal imager. The thermal imager is used to detect intruding thermally sensitive organisms, thus achieving real-time imaging. The pitot tube is used to measure the outside temperature, pressure, and speed, and serves as a needle... The system provides early warnings for extreme weather and serves as input for safe train operation. A zoom camera is used for imaging at distances less than 500 meters, enabling intrusion detection and data transmission. A telephoto camera is mounted above the thermal imager. Pressure sensors are installed on the skin plate of the coupler opening and closing mechanism, and a barrier clearer is installed below the opening and closing mechanism. LiDAR is installed on the outside of the opening and closing mechanism and the skirt, with the LiDAR used to scan the surrounding environment of the rail vehicle, construct a 3D map, and detect and warn of different objects approaching the train. The telephoto camera is used for imaging at distances greater than 500 meters, enabling intrusion detection and data transmission. The GPS positioning system can also be the BeiDou navigation and positioning system.

[0007] Preferably, the bogie includes wheelsets, a frame, axle box covers, a first tether spring, and a second tether spring. A first acceleration sensor, a second acceleration sensor, and an angular momentum encoder are arranged on the axle box covers, and a frame is provided between the two wheelsets.

[0008] The first acceleration sensor, the second acceleration sensor, and the angular momentum encoder are used to monitor and identify the condition of foreign objects crushing the track, collect the vibration of the wheelset in real time, transmit the impact signal, and enable the train to slow down or brake in an emergency.

[0009] Preferably, a front-mounted camera and a rear-mounted camera are installed on the top of the vehicle body; wherein the front-mounted camera and the rear-mounted camera are used for intrusion detection and imaging of falling objects from high altitudes.

[0010] Preferably, a storage and exchange system is provided below the driver's cab structure, wherein the storage and exchange system includes a data exchanger, a data storage device, ports, and a computer; the data exchanger is used to forward the data acquisition system to the correct port by identifying the destination address in the data acquisition system, thereby realizing rapid data exchange; the data storage device is used to store and save data collected by the rail train in the unmanned state; the ports are used to distinguish different services or processes, wherein each network connection needs to specify a source port and a destination port; the computer is used to perform computing tasks and process data.

[0011] Preferably, an opening and closing device is provided below the data exchanger and data storage device.

[0012] Secondly, this application also provides a collision avoidance identification method for rail trains in an unmanned driving state, including the following steps:

[0013] Based on the lidar, three-dimensional imaging data of the area around the train body is obtained, and other on-board sensors are used to detect and determine potential intruders, thus obtaining initial data of the train operation scene.

[0014] The initial data of train operation scenarios are accumulated as initial data and combined with the collected historical detection data to form a complete railway network big data detection cluster.

[0015] By training the railway network big data detection cluster with artificial intelligence machine learning algorithms, a railway train intrusion judgment and prediction model is obtained. This model is used to classify intrusions, determine their location, and predict the degree of danger to train operation.

[0016] Based on the railway train intrusion detection and prediction model, intrusions are detected in real time and the detection results are comprehensively judged.

[0017] Based on the comprehensive judgment results, corresponding train operation instructions are output to achieve the prediction and response to intrusions into the railway system. Specifically, the intrusion judgment and prediction model will significantly increase its prediction and judgment accuracy as train operating mileage increases (in reality, trains on all lines across the country are constantly running, and the training data of the railway network big data detection cluster is constantly increasing).

[0018] Thirdly, this application also provides a collision avoidance identification device for rail trains in an unmanned driving state, including:

[0019] Memory, used to store computer programs;

[0020] A processor is used to implement the steps of the rail train collision avoidance recognition method in the unmanned driving state when executing the computer program.

[0021] Fourthly, this application also provides a readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described method for collision avoidance identification of rail trains in an unmanned driving state.

[0022] The beneficial effects of this invention are as follows:

[0023] This invention combines a data acquisition system, an intelligent detection system, and a positioning system to achieve data exchange, pre-warning of train intrusions, and automatic train control. The described method and system for collision avoidance identification of unmanned rail trains has active and semi-active environmental perception capabilities. It is of great significance for detecting static and dynamic obstacles on and beside the track, as well as for hazard assessment, and for fully realizing the autonomous operation of rail trains. It plays an indispensable key role in ensuring the safe operation of trains.

[0024] The purpose of this invention is to provide a collision avoidance identification method and system for rail trains in an unmanned driving state. It can effectively solve the technical problem that existing high-speed trains are unable to detect foreign object intrusion in real time and autonomously avoid obstacles, and provide an important technical support and implementation method for the all-scenario detection capability of rail vehicles under high-speed operation.

[0025] The method and system used in this invention are among the key technologies for realizing full-scene perception during the operation of rail vehicles. When a foreign object intrudes during train operation, the system uses onboard sensors and various detection devices, along with machine learning artificial intelligence technology for data training, identification, and analysis, to improve the environmental perception capability of the rail train. This enables the train to issue instructions to avoid collisions, thereby protecting the safety of the train during operation. Therefore, the rail vehicle data acquisition system, intelligent detection system, and rail vehicle positioning system are integrated, and artificial intelligence and big data are introduced to realize full-scene perception capability during the operation of rail vehicles, thereby obtaining automatic driving functions for detecting foreign object intrusion on the track and automatic obstacle avoidance decision-making by the train.

[0026] The system in this invention collects real-time 3D data of train operation scenarios using onboard LiDAR and other advanced sensors, constructing a comprehensive railway network big data detection cluster. Utilizing advanced data processing technologies and machine learning algorithms, the system can accurately classify, locate, and assess the risks of detected intruders. Furthermore, this application includes an adaptive control strategy that dynamically adjusts train operation commands based on real-time analysis results, such as slowing down, emergency braking, or activating obstacle clearance devices, to ensure train operation safety. Through this method, this application not only improves detection accuracy and response speed but also enhances the system's automation and intelligence levels, meeting the demands of modern railway transportation for high safety and high efficiency.

[0027] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing embodiments of the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings. Attached Figure Description

[0028] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 This is a schematic diagram of the anti-collision identification device for rail trains in an unmanned driving state as described in an embodiment of the present invention;

[0030] Figure 2 This is a schematic diagram of the collision avoidance identification method for unmanned rail trains as described in this embodiment of the invention;

[0031] Figure 3 This is a schematic diagram of the anti-collision identification device for unmanned rail trains as described in an embodiment of the present invention.

[0032] In the diagram: 1. Track; 2. Intruder; 3. Pressure sensor; 4. Pitot tube; 5. Thermal imager; 6. Zoom camera; 7. Telephoto camera; 8. Car body; 9. Front upper camera; 10. Rear upper camera; 11. GPS positioning system; 12. LiDAR; 13. Angular momentum encoder; 14. First accelerometer; 15. Second accelerometer; 16. Opening and closing device; 17. Obstacle remover; 18. Wheelset; 19. Axle box cover; 20. First tethering spring; 21. Frame; 22. Second tethering spring; 23. Data exchanger; 24. Data storage device; 25. Port; 26. Computer; 27. Bogie; 28. Skirt; 29. ​​Driver's cab structure; 800. Collision avoidance identification device for unmanned rail trains; 801. Processor; 802. Memory; 803. Multimedia component; 804. I / O interface; 805. Communication component. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0034] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0035] Example 1:

[0036] like Figure 1 As shown, this embodiment provides a collision avoidance identification device for rail trains in an unmanned driving state. See [link / reference]. Figure 1 The device includes:

[0037] The vehicle comprises a data acquisition system, an intelligent detection system, and a positioning system. The data acquisition system includes a vehicle body 8, with a driver's cab structure 29 mounted on one side. A positioning system is located on the side beams of the vehicle body 8's underframe. Bogies 27 are mounted on the lower left and right ends of the vehicle body 8. The positioning system includes a GPS positioning system 11 and 3D digital map imaging technology. The intelligent detection system includes machine learning-based artificial intelligence detection algorithms. A zoom camera 6 is installed inside the windshield of the driver's cab structure 29, and a thermal imager 5 is installed outside the windshield. A pitot tube 4 is installed below the thermal imager 5. The thermal imager 5 is used to detect intruding thermally sensitive organisms, thus achieving real-time imaging. The pitot tube 4 is used to measure the outside temperature of the vehicle. The system uses temperature, pressure, and speed as early warning for extreme weather and as input for safe train operation; zoom camera 6 is used for imaging at distances less than 500 meters, thereby enabling the detection and data transmission of intruder 2; a telephoto camera 7 is installed above thermal imager 5; a pressure sensor 3 is installed on the skin plate of the coupler opening and closing mechanism; a barrier clearer 17 is installed on the lower side of the opening and closing mechanism; and a lidar 12 is installed on the outside of the opening and closing mechanism and the skirt 28. The lidar 12 is used to detect and scan the scene around the vehicle on track 1, construct a three-dimensional map, and enable the detection and early warning of different objects approaching the train; the telephoto camera 7 is used to achieve imaging at distances greater than 500 meters, thereby enabling the detection and data transmission of intruder 2.

[0038] Specifically, the bogie 27 includes wheelsets 18, a frame 21, an axle box cover 19, a first tether spring 20 and a second tether spring 22. A first acceleration sensor 14, a second acceleration sensor 15 and an angular momentum encoder 13 are arranged on the axle box cover 19. A frame 21 is provided between the two wheelsets 18.

[0039] Understandably, in this step, a first acceleration sensor, a second acceleration sensor, and an angular momentum encoder are arranged on the axle box cover to collect data on wheelset vibration, especially for monitoring and identifying the impact of foreign objects on the track. The pressure sensor inside the closed structure of the train driver's cab can detect the damage caused by foreign objects (stones, sticks, obstacles, animals, people, etc.) hitting the end of the train. Minor impacts do not affect the normal operation of the vehicle, while more serious impacts (which can be triggered by the impact acceleration or the loss of the pressure sensor as an alarm threshold) transmit an impact signal, causing the train to slow down or brake suddenly.

[0040] The first acceleration sensor 14, the second acceleration sensor 15, and the angular momentum encoder 13 are used to monitor and identify the crushing state of foreign objects on the track 1, and to collect the vibration of the wheelset 18 in real time, transmit the impact signal, and cause the train to slow down or brake in an emergency.

[0041] Specifically, a front-mounted camera 9 and a rear-mounted camera 10 are installed on the top of the vehicle body 8;

[0042] Among them, the front upper camera 9 and the rear upper camera 10 are used for intrusion detection and imaging of falling objects from high altitudes.

[0043] Specifically, a storage and exchange system is provided below the driver's cab structure 29, wherein the storage and exchange system includes a data exchanger 23, a data storage device 24, a port 25, and a computer 26;

[0044] Among them, the data exchanger 23 is used to forward the data acquisition system to the correct port by identifying the destination address in the data acquisition system, thereby realizing the rapid exchange of data; the data storage device 24 is used to store and save the data collected by the rail train in the unmanned state; the port 25 is used to distinguish different services or processes, where each network connection needs to specify a source port and a destination port; the computer 26 is used to perform computing tasks and process data.

[0045] Specifically, an opening and closing device 16 is provided below the data exchanger 23 and the data storage device 24.

[0046] Understandably, in this step, images and data acquired by the vehicle-mounted sensors are stored, including 3D scanning imaging data from the LiDAR, threshold judgment data from the pressure sensor, biological detection data (shape, size, and location determination) from the thermal imager, detection and scene imaging data of intruders in front of the vehicle from the telephoto and zoom cameras, extreme weather detection and warning data from the pitot tube, and intrusion detection data of falling objects from above from the front and rear cameras. Multi-sensor data fusion technology ensures the accuracy and robustness of the system.

[0047] Example 2:

[0048] This embodiment provides a collision avoidance identification method for rail trains in an unmanned driving state.

[0049] See Figure 2 The figure shows that the method includes steps S100, S200, S300, S400 and S500.

[0050] S100. Based on the lidar (12), obtain the three-dimensional imaging data of the train body surroundings, and at the same time use other on-board sensors to detect and determine potential intruders, and obtain the initial data of the train operation scene.

[0051] Understandably, in this step, the train data factory is a systematic and automated data management and processing framework used to efficiently train sensor-based data acquisition systems. In active and semi-active automatic driving, it identifies and classifies obstacles to enable the observation and analysis of railway environmental conditions. The train data factory consists of three core components: train-side data sensors, trackside data contacts, and a high-performance data center. Train-side data sensors collect real-time data, while trackside data contacts act as data transmission nodes, ensuring data is uploaded to the high-performance data center, which is responsible for storing, processing, and analyzing this data. Developers can access the data in the data center for system debugging, algorithm testing, and model training to improve the reliability of train data. The train data factory continuously collects new data through a data loop mechanism to train and optimize models, adapting to changes in the track environment.

[0052] S200: Initial data from train operation scenarios is accumulated as initial data, and combined with collected historical detection data, it is stored to form a complete railway network big data detection cluster.

[0053] Understandably, in this step, specifically in step S200, initial data on the train operation scenario is first acquired from train sensors. This data includes, but is not limited to, 3D imaging data of the vehicle's perimeter captured by onboard LiDAR, as well as real-time detection information of potential intruders from other sensors. The collection of initial data forms the basis for building a big data detection cluster, providing raw materials for subsequent data analysis and machine learning model training. To form a comprehensive and in-depth railway network big data detection cluster, this initial data needs to be combined with historical detection data. This historical data may originate from previous train operation records, maintenance logs, accident reports, and other relevant data archives. The accumulation of this data is crucial for understanding long-term trends and anomaly patterns in the railway environment.

[0054] Therefore, storing this data requires an efficient database management system to ensure data accessibility, consistency, and security. The database design needs to consider query efficiency and ease of data updates to support rapid data retrieval and real-time data analysis. This cluster not only provides a large number of training samples for machine learning models but also increases the model's generalization ability to detect intruders under different operating conditions. The construction and management of big data detection clusters may also involve data fusion technology, integrating information from different sensors and data sources to obtain a more comprehensive and accurate understanding of train operation scenarios. Data fusion algorithms can improve the robustness of the detection system and reduce the risks caused by the failure or error of a single data source.

[0055] S300: The railway network big data detection cluster is trained using artificial intelligence machine learning algorithms to obtain a railway train intrusion judgment and prediction model. This model is used to classify intrusions, determine their location, and predict the degree of danger to train operation.

[0056] Specifically, in this step, the contact point downloads a large amount of sensor data, such as LiDAR data, collected by the train during its journey via wireless connection within the train's range. The data is preprocessed and reduced so that only relevant portions are transmitted to the data center. The collected data is then labeled to provide the supervisory signals needed to train machine learning models. Features that aid model learning, such as texture, shape, and color in images, are selected, and machine learning models, such as deep neural networks, are trained using the selected features and labeled data. High-performance computers are deployed in a distributed manner within the data center, using GPUs for model training and inference, and efficient optimization algorithms, such as the Adam optimization algorithm, are applied to accelerate model training and improve recognition accuracy.

[0057] It is understood that step S300 includes S301, S302, and S303, wherein:

[0058] S301. Perform data cleaning and formatting on the data in the railway network big data detection cluster to remove noise and outliers, and obtain the final railway network big data detection cluster.

[0059] S302. The final railway network big data detection cluster is labeled to distinguish different types of intruders and the different degrees of harm each intruder poses to train operation, thus obtaining the labeled railway network big data detection cluster.

[0060] S303. The labeled railway network big data detection cluster is trained based on the random forest algorithm. During the training process, the labeled railway network big data detection cluster is divided into a test set and a validation set. The test set is used for training, and the remaining validation set is used for validation. This process is repeated multiple times to obtain a railway train intrusion judgment and prediction model. The railway train intrusion judgment and prediction model is then adjusted and optimized to obtain the final railway train intrusion judgment and prediction model. The final railway train intrusion judgment and prediction model includes adjusting model parameters and selecting different feature subsets.

[0061] It should be noted that in step S301, a thorough data cleaning process is first performed on the railway network big data detection cluster. This process involves identifying and removing noise and outliers from the dataset, which may be caused by sensor errors, data transmission errors, or environmental interference. The purpose of data cleaning is to ensure the quality and consistency of the dataset, providing an accurate data foundation for subsequent analysis and model training. Next, data formatting is performed, which means converting the data into a uniform format and structure to facilitate algorithm processing. Formatting may include standardizing data types, unifying timestamp formats, and adjusting data with inconsistent units.

[0062] Specifically, in step S302, the cleaned and formatted data is labeled. Labeling is a process of classification and classification levels that requires expertise to identify different types of intruders (such as animals, vegetation, projectiles, etc.) and assess their potential hazard to train operation (such as low, medium, high).

[0063] Specifically, in step S303, the labeled dataset is trained using the Random Forest algorithm. Random Forest is an ensemble learning method that improves the accuracy and robustness of the model by constructing multiple decision trees and combining their predictions. During training, the dataset is divided into a test set and a validation set. The test set is used for actual model training, while the validation set is used to evaluate the model's performance and ensure that the model does not overfit. Cross-validation, which involves repeating this process multiple times using a different subset of data as the validation set each time, allows for a more comprehensive evaluation of the model's performance. After training, the model is tuned and optimized. This may include adjusting model parameters (such as the number of trees, maximum depth, etc.), selecting different feature subsets, or trying different algorithm configurations. Therefore, in practical research, model training and optimization may also involve more advanced techniques, such as feature selection algorithms, model fusion techniques, and hyperparameter optimization algorithms (such as Bayesian optimization). Furthermore, with the development of deep learning technology, convolutional neural networks (CNNs) or recurrent neural networks (RNNs) can be considered for processing image data or time series data to further improve model performance.

[0064] By training and processing massive amounts of scanned data using artificial intelligence (AI) software in the field of machine learning, the system can quickly locate and identify the surrounding environment and objects, such as train tracks, utility poles, and fences, from the established massive data. At the same time, it uses LiDAR data to create accurate 3D imaging models of the tracks and the surrounding environment, and establishes a digital map of the railway environment.

[0065] S400: Based on the railway train intrusion detection and prediction model, intrusions are detected in real time and the detection results are comprehensively judged.

[0066] It is understood that step S400 includes S401 and S402, wherein:

[0067] S401. Based on the railway train intrusion detection and prediction model, real-time collected data is matched with historical training data, and the presence of intrusions is identified by a judgment formula as follows:

[0068]

[0069] In the formula, d is the standardized value of the data difference, Znew is the mean of the newly collected data, Zhistory is the mean of the historical data, and σ is the standard deviation of the historical data;

[0070] S402. If there is an intrusion, locate it and comprehensively analyze its characteristics; if not, continue to collect real-time train operation scene data and perform iterative calculations.

[0071] It is understood that step S402 includes S4021, S4022, S4023, and S4024, wherein:

[0072] S4021. Obtain intrusion features from real-time data, where intrusion features include size, shape, and velocity;

[0073] S4022. Using a positioning system, the location of the intruder is determined by calculating based on the acquired characteristics of the intruder. The calculation formula is as follows:

[0074] P intruder =f(Z) new In the formula ( , Θ, R), f is the positioning function, Θ is the sensor angle, Znew is the mean of the newly acquired data, R is the distance from the sensor to the train center, and P intruder Location information;

[0075] S4023. Calculate the final size, shape, and velocity of the intruder using image processing technology, extract key features from the calculated size, shape, and velocity of the intruder, and convert the extracted key features into numerical vectors to form feature vectors;

[0076] S4024. Based on the pre-trained machine learning model, the feature vector is input into the machine learning model for classification and property judgment, and the output results are obtained. The output results are integrated to finally obtain a comprehensive analysis and description of the intruder.

[0077] By combining artificial intelligence software with powerful computing capabilities, data collected by onboard cameras, lidar, and other sensors is analyzed and compared with a digital map of the railway environment. This allows for real-time detection of static and dynamic obstacles on and beside the track. Based on the size, speed, and location of obstacles, and considering the train's safety margin and operating strategy, collision risks are dynamically calculated. Based on the risk assessment results, decisions are made regarding whether action, such as slowing down or stopping, is necessary. The risk assessment model is continuously optimized based on actual events and feedback, improving the safety and efficiency of train operations. The fault diagnosis process begins with ensuring there are no encroachments on the track. If an encroachment is found, it's determined whether it's snow accumulation or fine sand. If so, it's considered minor or harmless. If it's gravel, its size is assessed. Gravel smaller than a certain diameter is considered minor or harmless; gravel larger than a certain size (e.g., greater than 10mm but less than 100mm) is considered hazardous and can be removed using a derailment device. Gravel larger than 100mm, which the derailment device cannot handle, is considered a serious hazard and requires emergency braking to prevent a collision. If a biological intrusion occurs, it's determined whether it's a human or animal. If a human, the distance from the track is assessed, and the train is activated with a horn and brakes to remove the intruder. If the human is already on the track, and it's suspected of suicide, emergency braking is initiated. If an animal crosses the track, the distance between the track and the train is assessed. If it's a nationally protected animal, a horn is sounded, the train is slowed, and brakes are applied to remove the animal. If it's a small, ordinary animal, a horn is sounded to remove it, the derailment device is activated, and the train slows down to minimize harm.

[0078] It should be noted that step S4023 includes S40231, S40232, and S40233, wherein:

[0079] S40231. Using image processing technology, extract the features of the intruder from the three-dimensional point cloud of the three-dimensional imaging sensor of the lidar (12), and distinguish the background and the body of the intruder according to the region growing or clustering algorithm; estimate its size by calculating the range of the entire space occupied by the body of the intruder in the point cloud, and measure the size again using edge detection and geometric analysis methods to obtain the final size information.

[0080] S40232. Based on the extracted image or point cloud data of the intruder, use Fourier descriptors to characterize its shape features and obtain the final shape information.

[0081] S40233. Obtain the time series information contained in the sensor data, calculate the velocity of the intruder by comparing the changes in the position of the intruder at different time points, and then obtain the final velocity information.

[0082] It should be noted that during the train's official operation, various data collected during the preliminary testing phase and a digital map of the railway environment are pre-loaded, recording the location information of fixed equipment. GPS (or BeiDou positioning system) and sensor data (acquired in real time) are compared and contrasted with a previously established digital map of the railway environment (a digital map of the railway environment is created periodically, such as monthly or yearly, and includes the three-dimensional geometric features and corresponding geographic coordinate system of railway facilities such as surrounding buildings and utility poles (excluding the train itself)). By identifying unchanging static objects near the train (such as power towers), the features of the identified objects are matched with the features of known objects in the digital map to determine the coordinates of the objects on the map. The calibrated location information is then fused with GPS positioning data to correct the GPS positioning data, achieving higher accuracy positioning. This method, combining landmark-based positioning and GPS, can achieve centimeter-level high-precision positioning. Each railway ancillary facility possesses its own unique three-dimensional geometric features. For example, the arrangement of sleepers and their surrounding features, as well as the geometric dimensions of power towers and transmission lines, are all 3D data that can be extracted and correspond one-to-one with their geodetic coordinates. First, a rough location range is obtained. Then, using artificial intelligence and big data, highly salient features of power towers and buildings are identified. Finally, based on the geometric features of sleepers and gravel, precise geodetic coordinates are determined. Simultaneously, the presence of foreign objects and their distance from the train can be identified. The degree of danger posed to the train is judged based on the distance, and this information is cross-validated with data detected by onboard sensors.

[0083] Specifically, region growing or clustering algorithms are used to distinguish between the background and the intruder itself. Region growing algorithms start from a seed point and gradually merge adjacent points into the region until specific conditions are met. Clustering algorithms, such as K-means or DBSCAN, help identify and distinguish different objects by grouping data points. Size estimation is achieved by calculating the extent of the space occupied by the intruder itself in the point cloud, which involves geometric analysis of the point cloud data, possibly including the calculation of geometric shapes such as minimum covering rectangles, minimum covering circles, or convex hulls. Edge detection algorithms, such as the Canny algorithm, can be used to identify the boundaries of the intruder, while geometric analysis methods are used to extract size information from these boundaries. The technical effect of this step is to provide an automated method for identifying and quantifying the physical properties of intruders from complex 3D data.

[0084] In the next step, Fourier descriptors are a method that transforms image or point cloud data into the frequency domain and extracts shape features from it. By performing a Fourier transform on the boundary or contour of a shape, its frequency domain representation can be obtained, thereby extracting key features that describe the shape. This helps machine learning models understand the unique geometric characteristics of different intruders, which is crucial for classifying and identifying intruder types.

[0085] In step S40233, time-series information is acquired from the sensor data, and this information is used to calculate the velocity of the intruder. This involves comparing the positional changes of the intruder at different time points. By measuring the positional difference between two time points and dividing it by the time difference, the average velocity of the intruder can be obtained. In practical research, these steps may be combined with more advanced image and signal processing techniques, such as using deep learning for point cloud analysis, or utilizing time-series analysis methods to process sensor data. Furthermore, combining multimodal data fusion techniques can further improve the accuracy of intruder feature extraction and analysis.

[0086] S500 outputs corresponding train operation instructions based on the comprehensive judgment results, so as to realize the prediction and response to intrusions into the railway system.

[0087] Understandably, in this step, the system connects the detection results and key information to the railway network and uploads them to the cloud. Information such as common obstacles is shared to improve identification efficiency. Collaboration between different lines allows for better monitoring of train operation status, enabling intelligent scheduling and optimized operational efficiency. By connecting to the cloud, the data factory can leverage its powerful computing resources to process large-scale datasets, supporting complex data analysis and machine learning model training. The elasticity and scalability of cloud services ensure stable system operation even when data volumes surge or computing demands increase, avoiding service interruptions due to peak loads.

[0088] Specifically, in this step, logical rules or machine learning models are used to assess the specific risk level of intrusions to train operation. For example, large animals or high-speed moving objects may require a higher level of response. Based on the risk assessment results, corresponding train operation instructions are formulated, which may include slowing down, emergency braking, changing tracks, activating obstacle removal devices, or notifying maintenance personnel. Furthermore, considering the train's current status, speed, location, and the operational status of the railway network, the output instructions are optimized to ensure safe and efficient train operation. The finalized operation instructions are then communicated to the train driver or automatic driving system through the train control system to ensure the train can respond promptly and take appropriate measures.

[0089] In practical research and applications, the output of train operation commands can be further integrated with the train's automatic driving system to achieve fully automated train operation control. Furthermore, these commands can be combined with railway traffic management systems to achieve optimized scheduling of the entire railway network. With the development of IoT and vehicle-to-everything (V2X) technologies, the output and reception of train operation commands will become more real-time and precise. Communication between trains and trackside sensors, traffic control systems, and other trains will make train operation more intelligent and collaborative.

[0090] It should be noted that the specific manner in which each module performs its operation in the apparatus described in the above embodiments has been described in detail in the embodiments of the method, and will not be elaborated here.

[0091] Example 3:

[0092] Corresponding to the above method embodiments, this embodiment also provides a rail train collision avoidance identification device in an unmanned driving state. The rail train collision avoidance identification device in an unmanned driving state described below and the rail train collision avoidance identification method in an unmanned driving state described above can be referred to in correspondence with each other.

[0093] Figure 3 This is a block diagram illustrating a collision avoidance identification device 800 for a rail train in an unmanned driving state, according to an exemplary embodiment. Figure 3 As shown, the unmanned rail train collision avoidance identification device 800 includes a processor 801 and a memory 802. The unmanned rail train collision avoidance identification device 800 also includes one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0094] The processor 801 controls the overall operation of the rail train collision avoidance and identification device 800 in the unmanned driving state to complete all or part of the steps in the aforementioned rail train collision avoidance and identification method in the unmanned driving state. The memory 802 stores various types of data to support the operation of the rail train collision avoidance and identification device 800 in the unmanned driving state. This data may include, for example, instructions for any application or method operating on the rail train collision avoidance and identification device 800 in the unmanned driving state, as well as application-related data, such as contact data, sent and received messages, images, audio, video, etc. The memory 802 can be implemented using any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touchscreen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in the memory 802 or transmitted via the communication component 805. The audio component also includes at least one speaker for outputting audio signals. I / O interface 804 provides an interface between processor 801 and other interface modules, such as keyboards, mice, or buttons. These buttons can be virtual or physical. Communication component 805 is used for wired or wireless communication between the rail train collision avoidance identification device 800 and other devices in the unmanned driving state. Wireless communication includes, for example, Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, or 4G, or one or more combinations thereof. Therefore, the corresponding communication component 805 may include a Wi-Fi module, a Bluetooth module, or an NFC module.

[0095] In an exemplary embodiment, the rail train collision avoidance identification device 800 in unmanned driving mode can be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-described rail train collision avoidance identification method in unmanned driving mode.

[0096] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided, which, when executed by a processor, implement the steps of the above-described method for preventing collisions with unmanned trains. For example, the computer-readable storage medium may be the memory 802 including the program instructions, which may be executed by the processor 801 of the unmanned train collision prevention device 800 to complete the above-described method for preventing collisions with unmanned trains.

[0097] Example 4:

[0098] Corresponding to the above method embodiments, this embodiment also provides a readable storage medium. The readable storage medium described below can be referred to in conjunction with the above-described method for collision avoidance identification of rail trains in an unmanned driving state.

[0099] A computer program is stored on a readable storage medium, and when the computer program is executed by a processor, it implements the steps of the rail train collision avoidance recognition method in the unmanned state described in the above method embodiments.

[0100] Specifically, the readable storage medium can be a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, or any other readable storage medium capable of storing program code.

[0101] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0102] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A collision avoidance identification device for unmanned rail trains, characterized in that, include: The system comprises a data acquisition system, an intelligent detection system, and a positioning system. The data acquisition system includes a vehicle body (8), a driver's cab structure (29) installed on one side of the vehicle body (8), a positioning system arranged on the side beam of the underframe of the vehicle body (8), and bogies (27) installed on the left and right sides of the lower side of the vehicle body (8). The positioning system includes a GPS positioning system (11) and three-dimensional digital map imaging technology. The intelligent detection system includes machine learning artificial intelligence detection algorithms. A zoom camera (6) is installed inside the windshield of the driver's cab structure (29), and a thermal imager (5) is installed outside the windshield. A pitot tube (4) is installed below the thermal imager (5). The thermal imager (5) is used to detect intruding organisms with thermal effects, thereby achieving real-time imaging. The pitot tube (4) is used to measure the vehicle's speed. External temperature, pressure and speed are used as early warning for extreme weather and as input for safe train operation; among them, the zoom camera (6) is used to detect imaging less than 500 meters, thereby realizing the detection and data transmission of intruders (2); a telephoto camera (7) is installed above the thermal imager (5), a pressure sensor (3) is installed on the skin plate of the coupler opening and closing mechanism, a barrier clearer (17) is installed on the lower side of the opening and closing mechanism, and a lidar (12) is installed on the outside of the opening and closing mechanism and the skirt (28). Among them, the lidar (12) is used to detect and scan the scene around the vehicle on the track (1) and construct a three-dimensional map to realize the detection and early warning of different objects approaching the train; among them, the telephoto camera (7) is used to realize imaging greater than 500 meters, thereby realizing the detection and data transmission of intruders (2); Specifically, GPS and sensor data are compared and contrasted with previously established digital maps of the railway environment; by identifying unchanging static objects near the train, the features of the identified objects are matched with the features of known objects in the digital map to determine the coordinates of the objects on the map; and the calibrated position information is fused with GPS positioning data to correct the GPS positioning data and achieve high-precision positioning. The bogie (27) includes wheelsets (18), frame (21), axle box cover (19), first tether spring (20) and second tether spring (22). A first acceleration sensor (14), a second acceleration sensor (15) and an angular momentum encoder (13) are arranged on the axle box cover (19). The frame (21) is provided between the two wheelsets (18). Among them, the first acceleration sensor (14), the second acceleration sensor (15) and the angular momentum encoder (13) are used to monitor and identify the crushing state of foreign objects on the track (1), and collect the vibration of the wheelset (18) in real time, transmit the impact signal, and make the train decelerate or brake urgently. A front-mounted camera (9) and a rear-mounted camera (10) are installed on the top of the vehicle body (8). Among them, the front upper camera (9) and the rear upper camera (10) are used for intrusion detection and imaging of falling objects from high altitudes; The unmanned rail train collision avoidance identification device is used to implement the unmanned rail train collision avoidance identification method, which includes the following steps: The initial data of the train operation scene are obtained by acquiring three-dimensional imaging data around the train body using lidar (12) and using other on-board sensors to detect and determine potential intruders. The initial data of train operation scenarios are accumulated as initial data and combined with the collected historical detection data to form a complete railway network big data detection cluster. By training the railway network big data detection cluster with artificial intelligence machine learning algorithms, a railway train intrusion judgment and prediction model is obtained. This model is used to classify intrusions, determine their location, and predict the degree of danger to train operation. Based on the railway train intrusion detection and prediction model, intrusions are detected in real time and the detection results are comprehensively judged. Based on the comprehensive judgment results, corresponding train operation instructions are output to achieve the prediction and response to intrusions into the railway system; The railway train intrusion detection and prediction model detects intrusions in real time and makes a comprehensive judgment on the detection results, including: Based on the railway train intrusion detection and prediction model, real-time collected data is matched with historical training data, and the presence of intrusions is identified by a judgment formula, which is as follows: In the formula, d is the standardized value of the data difference. The mean of the newly collected data. σ is the mean of the historical data, and σ is the standard deviation of the historical data. If an intrusion is found, its location is determined, and its characteristics are comprehensively analyzed. If no intrusion is found, real-time data collection of train operation scenarios continues, and iterative calculations are performed. The process of locating the intrusion and comprehensively analyzing its characteristics includes: Acquire intrusion features from real-time data, including size, shape, and velocity; Using a positioning system, the location of the intruder is determined by calculating its characteristics. The calculation formula is as follows: In the formula, Here is the positioning function, and Θ is the sensor angle. The mean of the newly collected data. This refers to the distance from the sensor to the center of the train. Location information; Image processing techniques are used to calculate the final size, shape, and velocity of the intruder. Key features are then extracted from the calculated size, shape, and velocity of the intruder and converted into numerical vectors to form feature vectors. Based on a pre-trained machine learning model, feature vectors are input into the machine learning model for classification and property judgment, and output results are obtained. The output results are then integrated to obtain a comprehensive analysis and description of the intruder. The image processing technology is used to calculate the final size, shape, and velocity of the intruder, including: Image processing techniques are used to extract the features of the intruder from the three-dimensional point cloud of the three-dimensional imaging sensor of the lidar (12), and the background and the intruder body are distinguished according to the region growing or clustering algorithm; the size is estimated by calculating the range of the entire space occupied by the intruder body in the point cloud, and the size is measured again by edge detection and geometric analysis methods to obtain the final size information. Based on the extracted images or point cloud data of the intruder, Fourier descriptors are used to characterize its shape features to obtain the final shape information; The time-series information contained in the sensor data is acquired, and the velocity of the intruder is calculated by comparing the changes in the position of the intruder at different time points, thus obtaining the final velocity information; The method involves training a railway network big data detection cluster using artificial intelligence machine learning algorithms to obtain a railway train intrusion detection and prediction model. This model is used to classify intrusions, determine their location, and predict the degree of hazard to train operation. This includes: Data cleaning and formatting are performed on the data in the railway network big data detection cluster to remove noise and outliers, resulting in the final railway network big data detection cluster. The final railway network big data detection cluster is labeled to distinguish different types of intruders and the different degrees of harm each intruder poses to train operation, resulting in the labeled railway network big data detection cluster. The labeled railway network big data detection cluster is trained using the random forest algorithm. During the training process, the labeled railway network big data detection cluster is divided into a test set and a validation set. The test set is used for training, and the remaining validation set is used for validation. This process is repeated multiple times to obtain a railway train intrusion detection and prediction model. The railway train intrusion detection and prediction model is then adjusted and optimized to obtain the final railway train intrusion detection and prediction model. The final railway train intrusion detection and prediction model includes adjusting model parameters and selecting different feature subsets.

2. The anti-collision identification device for unmanned rail trains according to claim 1, characterized in that, A storage and exchange system is provided below the driver's cab structure (29), wherein the storage and exchange system includes a data exchanger (23), a data storage device (24), a port (25), and a computer (26). Among them, the data exchanger (23) is used to forward the data acquisition system to the correct port by identifying the destination address in the data acquisition system, thereby realizing the rapid exchange of data; the data storage device (24) is used to store and save the data collected by the rail train in the unmanned state; the port (25) is used to distinguish different services or processes, where each network connection needs to specify a source port and a destination port; the computer (26) is used to perform computing tasks and process data.

3. The anti-collision identification device for unmanned rail trains according to claim 2, characterized in that, An opening and closing device (16) is provided below the data exchanger (23) and the data storage device (24).

Images