Rail locomotive remote intelligent monitoring method and system and electronic equipment

Abstract

The invention relates to the technical field of rail locomotive remote intelligent monitoring scheme design, in particular to a rail locomotive remote intelligent monitoring method and system and electronic equipment. A plurality of monitoring points are arranged on a running route of a rail locomotive, and a target area image is collected in real time and transmitted to a control module on the rail locomotive through a cloud server. And the control module identifies the image by using a trained YOLO model, judges whether a person and a track exist at the same time, constructs a dangerous area if the person and the track exist at the same time, judges whether the person exists in the dangerous area, calculates the actual distance between the person and the locomotive, and intelligently controls the running state of the track locomotive according to the distance. And the possibility that equipment in the locomotive is damaged or the management is disordered due to emergency braking of the locomotive is reduced. Meanwhile, the system can monitor the communication state between a monitoring point and the locomotive in real time, and the monitoring reliability is ensured. And the intelligent degree, the safety and the operation stability of the operation of the rail locomotive are improved to a great extent.

Description

Technical Field

[0001] The present invention relates to the technical field of the design of remote intelligent monitoring solutions for rail locomotives, and particularly relates to a method and system for remote intelligent monitoring of rail locomotives, and an electronic device. Background Art

[0002] As a key component of the modern transportation system, rail locomotives undertake the task of efficiently transporting a large number of passengers and goods, and play an irreplaceable role in the economic and social development. However, the operating environment of rail locomotives is complex and diverse, facing many potential risk factors, which poses a severe challenge to the operating safety of rail locomotives.

[0003] Traditional methods for monitoring the operation of rail locomotives mainly include manual inspections and simple sensor monitoring. Manual inspection is a common and long-established monitoring method. Inspectors need to walk along the railway tracks or conduct visual inspections during the operation of the train. Although this method can directly observe the condition of the tracks and train equipment, it has many limitations. On the one hand, the labor intensity of manual inspection is extremely high. Inspectors need to work for a long time in various harsh weather conditions, which is prone to fatigue, thus leading to missed inspections or misjudgments. On the other hand, the real-time performance of manual inspection is poor, and the inspection interval is relatively long, making it impossible to detect emergencies in a timely manner. For example, during the gap between two shifts of inspectors or during the inspection interval, dangerous situations such as unauthorized entry of people on the tracks may occur and cannot be detected in time.

[0004] Simple sensor monitoring mainly monitors some specific physical quantities, such as the vibration and temperature of the railway tracks. These sensors can timely feedback changes in certain physical characteristics of the tracks and reflect the basic health status of the tracks. However, they can only provide limited information and are difficult to comprehensively and accurately reflect the comprehensive safety situation on the tracks. For example, sensors cannot real-time sense whether there are people breaking into the vicinity of the tracks, which is often one of the important potential risks leading to railway traffic accidents.

[0005] Therefore, the existing technology still needs to be further developed. Summary of the Invention

[0006] The purpose of the present invention is to overcome the above technical deficiencies and provide a method and system for remote intelligent monitoring of rail locomotives, and an electronic device to solve the problems existing in the prior art.

[0007] To achieve the above technical objectives, according to the first aspect of the present invention, the present invention provides a method for remote intelligent monitoring of rail locomotives, including: S100. Set monitoring points at preset distances on the running route of the rail locomotive, and install monitoring devices at each of the monitoring points. The monitoring devices include cameras, which are used to capture images of the target area at preset time intervals, and send the captured images and the timestamps when the images are captured to the control module set on the rail locomotive through a cloud server; S200. The control module identifies the images of the target area, and determines whether there are target images defined as people and target images defined as tracks at the same time. If so, calculate the coordinates of the four vertices of the first anchor box for indicating each person and the second anchor box for indicating the track respectively, establish a dangerous area according to the coordinates of the four vertices of the second anchor box, calculate the centroid coordinates of each person according to the coordinates of the four vertices of the first anchor box of each person, determine whether the centroid coordinates of a certain person are located within the dangerous area. If so, calculate the first position information corresponding to the centroid coordinates using a homography matrix, obtain the second position information of the rail locomotive using the GPS sensor set on the rail locomotive, and calculate the actual distance between the first position information and the second position information; S300. Control the running state of the rail locomotive according to the actual distance.

[0008] Specifically, the controlling the running state of the rail locomotive according to the actual distance includes: Judge whether the actual distance is greater than or equal to a first preset threshold and less than a second preset threshold; If the actual distance is greater than or equal to the first preset threshold and less than the second preset threshold, control the rail locomotive to perform normal braking.

[0009] Specifically, the controlling the running state of the rail locomotive according to the actual distance includes: If the actual distance is greater than or equal to the second preset threshold, control the rail locomotive to run normally or start to resume normal running, and the second preset threshold is greater than the preset normal braking distance of the rail locomotive.

[0010] Specifically, the controlling the running state of the rail locomotive according to the actual distance includes: If the actual distance is less than the first preset threshold, control the rail locomotive to perform emergency braking, and the first preset threshold is greater than the preset emergency braking distance of the rail locomotive.

[0011] Specifically, the method includes: If there are no target images defined as people and target images defined as tracks at the same time, control the rail locomotive to run normally or start to resume normal running.

[0012] Specifically, the control module identifying the images of the target area includes: A dataset is formed using the images captured by the camera, the images in the dataset are labeled, and then the YOLO model is trained. The YOLO model is used to identify the personnel and tracks in the images of the target area.

[0013] Specifically, the method further includes: If the centroid coordinates of a certain person do not exist within the dangerous area, control the rail locomotive to travel normally or start to resume normal travel.

[0014] Specifically, the method further includes: Judge whether the control module receives the next frame of image of a certain monitoring point within a preset time after receiving the current frame of image of the monitoring point. If not, output an alarm signal regarding the communication anomaly of the monitoring point; If so, output a prompt signal regarding the normal communication of the monitoring point.

[0015] According to the second aspect of the present invention, there is provided a remote intelligent monitoring system for a rail locomotive, including: An acquisition module, including a camera for capturing images of the target area at preset time intervals, and sending the captured images and the timestamps when the images are captured to a control module set on the rail locomotive through a cloud server; A control module, used to identify the images of the target area, judge whether there are target images defined as persons and target images defined as tracks at the same time. If so, calculate the coordinates of the four vertices of the first anchor box for indicating each person and the second anchor box for indicating the track respectively, establish a dangerous area according to the coordinates of the four vertices of the second anchor box, calculate the centroid coordinates of each person according to the coordinates of the four vertices of the first anchor box of each person, judge whether there is a centroid coordinate of a certain person within the dangerous area. If so, calculate the first position information corresponding to the centroid coordinate using a homography matrix, obtain the second position information of the rail locomotive using a GPS sensor set on the rail locomotive, and calculate the actual distance between the first position information and the second position information; used to control the running state of the rail locomotive according to the actual distance.

[0016] According to the third aspect of the present invention, there is provided an electronic device, including: a memory; and a processor, where computer-readable instructions are stored on the memory, and when the computer-readable instructions are executed by the processor, the above-mentioned remote intelligent monitoring method for a rail locomotive is implemented.

[0017] Beneficial effects: 1. Improve the real-time performance and accuracy of monitoring: By setting up multiple monitoring points along the running route of the rail locomotive and installing cameras at each monitoring point to collect image information of the rail area in real time, and using deep learning algorithms to quickly and accurately identify the images, potential dangerous situations near the rails can be detected in a timely manner, such as detecting in real time the situation of personnel appearing on the rails.

[0018] 2. Intelligent control of the running state of the rail locomotive: Analyze the actual distance between the personnel and the rail locomotive based on the image information collected by the monitoring points, and combine with the preset threshold to intelligently control the running state of the rail locomotive, such as normal driving or starting to resume normal driving, normal braking and emergency braking, etc., which greatly improves the running safety of the rail locomotive.

[0019] 3. Real-time communication monitoring: The system can monitor the communication status between the monitoring points and the rail locomotive in real time. If communication anomalies occur, an alarm signal is output in a timely manner to ensure the stability and reliability of the monitoring system.

[0020] 4. Improve the traffic operation efficiency and stability: Timely and accurate monitoring and control greatly reduce the interruption of the operation of the rail locomotive caused by unexpected situations. If dangerous situations occur, while ensuring safety, the possibility of damage to in-vehicle equipment or management chaos caused by the emergency braking of the locomotive is reduced, the operation cost is reduced, and the operation efficiency and stability of the entire railway system are improved. Description of the Drawings

[0021] Figure 1 is a schematic flowchart of the remote intelligent monitoring method for rail locomotives provided in the specific embodiment of the present invention; Figure 2 is a schematic diagram of the system composition of the remote intelligent monitoring system for rail locomotives provided in the specific embodiment of the present invention. Detailed Embodiments

[0022] In order to enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Based on the embodiments in this application, other similar embodiments obtained by those of ordinary skill in the art without creative efforts shall all fall within the scope of protection of this application. In addition, the directional terms mentioned in the following embodiments, such as "up", "down", "left", "right", etc., are only with reference to the directions of the drawings. Therefore, the directional terms used are for illustration rather than limiting the present invention.

[0023] The present invention will be further described below in conjunction with the drawings and preferred embodiments.

[0024] Please refer to Figure 1 , the present invention provides a remote intelligent monitoring method for rail locomotives, including: S100. On the operating route of the rail locomotive, monitoring points are set at preset distances. A monitoring device is installed at each of the monitoring points. The monitoring device includes a camera, which is used to capture images of the target area at preset time intervals, and the captured images and the timestamps when the images are captured are sent to the control module set on the rail locomotive through a cloud server.

[0025] Specifically, the method includes: Based on the operating route of the rail locomotive, using a Geographic Information System (GIS) and professional surveying tools, the information of each key position on the operating route is measured in detail. According to the rule of every 3 kilometers (the preset distance can be adjusted according to the actual track conditions), the positions where monitoring points need to be set are accurately marked. A metal bracket with a height of 5 meters (which can also be adjusted to obtain a suitable viewing angle) is installed at each monitoring point, and the monitoring device is firmly installed on the top of the bracket to ensure that the monitoring device can cover an area 50 meters in front of the track (the adjustable viewing angle range).

[0026] The camera in the monitoring device uses a high-definition, high-frame-rate professional industrial camera. Its resolution is set to 1920×1080 pixels, and the frame rate is set to 30 frames per second (which can be adjusted according to actual needs). Images of the target area are captured at an interval of every 5 seconds (i.e., the preset time interval), and the captured images and the accurate timestamps (accurate to milliseconds) when the images are captured are sent to the control module set on the rail locomotive through a high-speed and stable cloud server.

[0027] Specifically, the method further includes: From a large number of images collected by the monitoring device, 2000 representative images are selected as the data set. These images cover rail area scenes under different lighting conditions (such as daytime, night, cloudy days, etc.), different weather conditions (such as sunny, light rain, foggy, etc.), and different positions and postures of personnel.

[0028] Using a professional image annotation tool, each image in the data set is annotated. For personnel targets, a rectangular box (i.e., the first anchor box) is used for annotation to ensure that the rectangular box can completely surround the body contour of the personnel; for rail targets, a rectangular box (i.e., the second anchor box) is also used for annotation to make it closely fit the trend and scope of the rail. The annotation information is stored in association with the corresponding image file in the form of an XML file.

[0029] Further, the present invention uses the drawing function in the Labelme tool to accurately draw prediction boxes along the edges of personnel or tracks respectively. For personnel or tracks with regular shapes, the maximum distance between the sides of the prediction box and the personnel or tracks does not exceed 2 pixels; for personnel or tracks with irregular shapes, the prediction box should fit as closely as possible to their outer contours, and the similarity between the covered area and the actual area of the personnel or tracks reaches more than 90%. At the same time, it is ensured that the prediction box completely covers the area of the cut personnel or tracks, and the number of pixel points of non-personnel or non-track areas misselected by the box does not exceed 5% of the total number of pixels in the boxed area.

[0030] Further, assign corresponding category labels to each drawn prediction box, and accurately fill in the determined category names in the annotation information column of Labelme to ensure that the labels are consistent with the pre-set "personnel" and "track", and avoid label confusion or errors.

[0031] Further, conduct a 100% comprehensive inspection on the annotated images, using a two-person cross-check method to ensure that the prediction boxes are drawn accurately and the category labels are correct. For images with doubts or inaccurate annotations (such as inconsistent inspection results of the two people, the proportion of impurities covered by the prediction box is less than 90% or the misselection ratio is higher than 5%, etc.), return them for re-observation and annotation until the annotation quality of all images meets the standards.

[0032] After all the images are annotated, export the data file with prediction box annotation information and category labels in the JSON format specified by Labelme and save it to a dedicated folder for subsequent use in the training of the YOLO model.

[0033] Specifically, before step S100, it includes: Preset a preset distance, a first preset threshold, a second preset threshold, a preset time interval, a preset time, a preset emergency braking distance, a preset normal braking distance, and a preset multiple in the control module.

[0034] It can be understood that the values of the preset distance, the first preset threshold, the second preset threshold, the preset time interval, the preset time, and the preset multiple can be specifically set according to the actual needs of the users of the present invention, and the present invention does not limit this here, as long as it is applicable to the remote intelligent monitoring method of the rail locomotive proposed by the present invention.

[0035] Preferably, in the present invention, the preset distance is preferably 3000 meters, the first preset threshold is preferably 1000 meters, the second preset threshold is preferably 2000 meters, the preset time interval is set to 5 seconds, the preset time is preferably 10 seconds, and the preset multiple is preferably 1.4 times. Since the maximum emergency braking distance of the rail locomotive used in the present invention is 700 meters and the maximum normal braking distance is 1400 meters, that is, the preset emergency braking distance of the rail locomotive used in the present invention is 800 meters and the preset normal braking distance is 1500 meters. The above settings are based on the rail locomotive used in the present invention. It can well implement the rail locomotive remote intelligent monitoring method described in the present invention.

[0036] S200. The control module identifies the images in the target area, determines whether there are target images defined as humans and target images defined as tracks at the same time. If so, it calculates the coordinates of the four vertices of the first anchor box for indicating each person and the second anchor box for indicating the track respectively, establishes a dangerous area according to the coordinates of the four vertices of the second anchor box, calculates the centroid coordinates of each person according to the coordinates of the four vertices of the first anchor box of each person, determines whether there is a centroid coordinate of a certain person located in the dangerous area. If so, it calculates the first position information corresponding to the centroid coordinate by using the homography matrix, obtains the second position information of the rail locomotive by using the GPS sensor arranged on the rail locomotive, and calculates the actual distance between the first position information and the second position information.

[0037] Specifically, the establishing a dangerous area according to the coordinates of the four vertices of the second anchor box includes: Calculating the geometric center coordinates of the second anchor box according to the coordinates of the four vertices of the second anchor box. Taking the geometric center coordinates of the second anchor box as the base point, magnifying the second anchor box by a preset multiple, taking the magnified second anchor box as the safety boundary box, and obtaining the coordinates of the four vertices of the dangerous area.

[0038] Specifically, the determining whether there is a centroid coordinate of a certain person located in the dangerous area includes: Judging whether there is a centroid coordinate of a certain person located inside the dangerous area according to the relative position relationship between the coordinates of the four vertices of the dangerous area and the centroid coordinates of each person.

[0039] It should be noted here that the calculating the centroid coordinates of the person includes: Making a piecewise linear change to the gray value of the current frame infrared thermal imaging image according to the mean and standard deviation of the gray value of the current frame infrared thermal imaging image, obtaining the gray value of each point of the 8-bit single-channel image of the current frame image, and calculating the centroid coordinates of the person according to the gray value of each point of the 8-bit single-channel image of the current frame infrared thermal imaging image.

[0040] The formula for the piecewise linear transformation is as follows: where μ and σ are the mean and standard deviation of the gray values of the current frame infrared thermal imaging image respectively, x is the gray value of each point of the current frame infrared thermal imaging image, y is the gray value of each point of the 8-bit single-channel image obtained after linear transformation, and is the floor operation.

[0041] The calculation formula for centroid positioning is as follows: where M and N represent the width and height of the first anchor box respectively, (Mmin, Nmin) and (Xmax, Ymax) represent the upper left coordinate and the lower right coordinate of the first anchor box respectively, Iij is the gray value of the pixel point within the rectangular area of the first anchor box, i is the row position of the pixel point, j is the column position of the pixel point, and (Xc, Yc) represents the centroid coordinates of the person.

[0042] It should be noted here that the application of the homography matrix in the conversion between pixel coordinates and geographic coordinates is a prior art, and the present invention will not elaborate on it too much.

[0043] S300. Control the running state of the rail locomotive according to the actual distance.

[0044] Specifically, the controlling the running state of the rail locomotive according to the actual distance includes: Judge whether the actual distance is greater than or equal to a first preset threshold and less than a second preset threshold; If the actual distance is greater than or equal to the first preset threshold and less than the second preset threshold, control the rail locomotive to perform normal braking.

[0045] Specifically, the controlling the running state of the rail locomotive according to the actual distance includes: If the actual distance is greater than or equal to the second preset threshold, control the rail locomotive to run normally or start to resume normal running, and the second preset threshold is greater than the preset normal braking distance of the rail locomotive.

[0046] Specifically, the controlling the running state of the rail locomotive according to the actual distance includes: If the actual distance is less than the first preset threshold, control the rail locomotive to perform emergency braking, and the first preset threshold is greater than the preset emergency braking distance of the rail locomotive.

[0047] It should be noted here that the actual distance is preferably the Euclidean distance.

[0048] Specifically, the method includes: If there is no target image defined as a person and a target image defined as a track at the same time, control the normal driving of the track locomotive or start to resume normal driving.

[0049] It can be understood that the steps related to program operation in S100-S300 of the present invention are loop operations. Whenever the track locomotive receives a frame of image, the program of S100-S300 is executed once in a loop.

[0050] It can be understood that the remote intelligent monitoring system of the track locomotive in the present invention can be installed on multiple track locomotives on the same line respectively.

[0051] Specifically, the method further includes: If there are two or more track locomotives in a driving state on the same line, when controlling the track locomotive in front of the line to start normal braking or emergency braking, mark this track locomotive as the first track locomotive, and mark the track locomotive behind the first track locomotive as the second track locomotive. Use the GPS sensors respectively set on the first track locomotive and the second track locomotive to obtain the position data of the first track locomotive and the second track locomotive, and then calculate the second actual Euclidean distance between the first track locomotive and the second track locomotive, and judge whether the second actual Euclidean distance is greater than or equal to the first preset threshold and less than the second preset threshold; If the second actual Euclidean distance is greater than or equal to the first preset threshold and less than the second preset threshold, control the second track locomotive to perform normal braking; If the second actual Euclidean distance is greater than or equal to the second preset threshold, control the second track locomotive to drive normally or start to resume normal driving, and the second preset threshold is greater than the preset normal braking distance of the second track locomotive; If the second actual Euclidean distance is less than the first preset threshold, control the second track locomotive to perform emergency braking, and the first preset threshold is greater than the preset emergency braking distance of the second track locomotive.

[0052] It can be understood that through the above solution, it can be ensured that after the previous track locomotive on the same line starts braking due to dangerous reasons, it can be ensured that the following track locomotive will not collide with the previous track locomotive. At the same time, while ensuring safety, it reduces the possibility of equipment damage or management chaos in the following second track locomotive due to emergency braking, and greatly improves the safety, reliability and intelligence of the present invention.

[0053] Specifically, the control module recognizes the image of the target area, including: Use the images taken by the camera to form a data set, annotate the images in the data set, and then train the YOLO model, and use the YOLO model to recognize the people and tracks in the images of the target area.

[0054] Specifically, the training process of the YOLO model is as follows: YOLO model training preparation: Use the Python programming language and related deep learning frameworks (such as PyTorch) to build a training environment. Divide the labeled dataset into a training set and a validation set in a ratio of 8:2. The training set contains 1,600 images, and the validation set contains 400 images.

[0055] Training parameter settings: Set the training batch size to 16, that is, process 16 images simultaneously for model training each time to balance training efficiency and memory usage. Set the initial learning rate to 0.001, and adopt a cosine annealing learning rate scheduling strategy during training to gradually adjust the learning rate as the number of training rounds increases, so as to improve the convergence speed and accuracy of the model. Set the number of training epochs to 50, that is, the model performs 50 complete learning processes on the entire training set.

[0056] Model training process: Input the training set data into the YOLO model. The model calculates the prediction results through forward propagation, compares the prediction results with the annotation information, and calculates the loss function (such as a combined loss of mean squared error MSE and intersection over union IoU). According to the value of the loss function, use the backpropagation algorithm to update the weight parameters of the model to minimize the loss function. After each round of training, input the validation set into the model for validation, and evaluate the performance metrics of the model on the validation set, such as mean average precision (mAP). If the performance metrics of the validation set do not improve significantly (the improvement amplitude is less than the set threshold, such as 0.01) in several consecutive rounds (such as 3 rounds) of training, stop training and save the weight parameters of the model with the best performance.

[0057] Specifically, the method further includes: If the centroid coordinates of a certain person are not within the dangerous area, control the rail locomotive to drive normally or start to resume normal driving.

[0058] Specifically, the method further includes: Judge whether the control module receives the next frame image of a certain monitoring point within a preset time after receiving the current frame image of the monitoring point. If not, output an alarm signal regarding the communication anomaly of the monitoring point; If so, output a prompt signal regarding the normal communication of the monitoring point.

[0059] Specifically, after receiving the current frame image of a certain monitoring point, the control module on the rail locomotive starts an internal timer and sets the timing time to a preset communication time threshold (such as 10 seconds). If the next frame image of this monitoring point can be received within 10 seconds, by comparing with the normal communication signal (the preset normal communication signal mode), a prompt signal indicating normal communication of this monitoring point is output. If the next frame image is not received within 10 seconds, the communication exception handling mechanism is triggered, and an alarm signal indicating communication exception of this monitoring point is output through the network alarm module to notify the relevant responsible personnel for timely handling. At the same time, this information and time are recorded in the log file for subsequent analysis and fault troubleshooting, further improving the intelligence, safety, and reliability of the present invention.

[0060] It can be understood that the present invention has the following beneficial effects: 1. Improve the real-time and accuracy of monitoring: By setting multiple monitoring points on the running route of the rail locomotive and installing cameras at each monitoring point to collect image information of the rail area in real time, and using deep learning algorithms to quickly and accurately identify the images, potential dangerous situations near the rails can be detected in time, such as detecting the situation of people appearing on the rails in real time.

[0061] 2. Intelligent control of the running state of the rail locomotive: According to the image information collected by the monitoring points, analyze the actual distance between the personnel and the rail locomotive, and combine with the preset threshold to intelligently control the running state of the rail locomotive, such as normal driving or starting to resume normal driving, normal braking, and emergency braking, etc., greatly improving the safety of the rail locomotive operation.

[0062] 3. Real-time communication monitoring: The system can monitor the communication status between the monitoring points and the rail locomotive in real time. If communication anomalies occur, alarm signals are output in time to ensure the stability and reliability of the monitoring system.

[0063] 4. Improve traffic operation efficiency and stability: Timely and accurate monitoring and control greatly reduce the interruption of the rail locomotive operation caused by unexpected situations. If dangerous situations occur, while ensuring safety, it reduces the possibility of damage to in-vehicle equipment or management chaos caused by the emergency braking of the locomotive, reduces operation costs, and improves the operation efficiency and stability of the entire railway system.

[0064] Please refer to Figure 2 , the present invention provides another embodiment. This embodiment provides a remote intelligent monitoring system for rail locomotives, and the remote intelligent monitoring system for rail locomotives includes: An acquisition module 100, including a camera for taking images of the target area at preset time intervals, and sending the taken images and the timestamps when the images are taken to the control module set on the rail locomotive through the cloud server; The control module 200 is configured to identify an image of a target area, and determine whether there are both a target image defined as a person and a target image defined as a track. If so, calculate the coordinates of the four vertices of the first anchor box for indicating each person and the second anchor box for indicating the track, establish a danger area according to the coordinates of the four vertices of the second anchor box, calculate the centroid coordinates of each person according to the coordinates of the four vertices of the first anchor box of each person, determine whether the centroid coordinates of a certain person are located within the danger area. If so, calculate the first position information corresponding to the centroid coordinates by using a homography matrix, obtain the second position information of the rail locomotive by using a GPS sensor disposed on the rail locomotive, and calculate the actual distance between the first position information and the second position information; and is configured to control the running state of the rail locomotive according to the actual distance.

[0065] It should be noted here that the present invention has the following beneficial effects: 1. Improve the real-time performance and accuracy of monitoring: By setting multiple monitoring points on the running route of the rail locomotive and installing cameras at each monitoring point to collect image information of the rail area in real time, and using deep learning algorithms to quickly and accurately identify the images, potential dangerous situations near the track can be detected in time, such as detecting the situation where a person appears on the track in real time.

[0066] 2. Intelligently control the running state of the rail locomotive: Analyze the actual distance between the person and the rail locomotive according to the image information collected by the monitoring points, and intelligently control the running state of the rail locomotive, such as normal driving or starting to resume normal driving, normal braking and emergency braking, etc., greatly improving the running safety of the rail locomotive.

[0067] 3. Real-time communication monitoring: The system can monitor the communication state between the monitoring points and the rail locomotive in real time. If a communication anomaly occurs, an alarm signal is output in time to ensure the stability and reliability of the monitoring system.

[0068] 4. Improve traffic operation efficiency and stability: Timely and accurate monitoring and control greatly reduce the interruption of the running of the rail locomotive caused by unexpected situations. If a dangerous situation occurs, while ensuring safety, the possibility of damage to in-vehicle equipment or management chaos caused by the emergency braking of the locomotive is reduced, the operation cost is reduced, and the operation efficiency and stability of the entire railway system are improved.

[0069] In a preferred embodiment, the present application further provides an electronic device, and the electronic device includes: A memory; and a processor, wherein computer-readable instructions are stored on the memory, and when the computer-readable instructions are executed by the processor, the remote intelligent monitoring method for rail locomotives is implemented. This computer device can generally be a server, a terminal, or any other electronic device with the necessary computing and / or processing capabilities. In one embodiment, the computer device may include a processor, a memory, a network interface, a communication interface, etc. connected through a system bus. The processor of the computer device can be used to provide the necessary computing, processing, and / or control capabilities. The memory of the computer device may include a non-volatile storage medium and an internal memory. An operating system, computer programs, etc. may be stored in or on the non-volatile storage medium. The internal memory can provide an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface and communication interface of the computer device can be used to connect and communicate with external devices through a network. When the computer program is executed by the processor, the steps of the method of the present invention are executed.

[0070] The present invention can be implemented as a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method of the embodiments of the present invention are caused to be executed. In one embodiment, the computer program is distributed among a plurality of network-coupled computer devices or processors, so that the computer program is stored, accessed, and executed in a distributed manner by one or more computer devices or processors. A single method step / operation, or two or more method steps / operations, can be executed by a single computer device or processor or by two or more computer devices or processors. One or more method steps / operations can be executed by one or more computer devices or processors, and one or more other method steps / operations can be executed by one or more other computer devices or processors. One or more computer devices or processors can execute a single method step / operation, or execute two or more method steps / operations.

[0071] Those of ordinary skill in the art can understand that the method steps of the present invention can be completed by a computer program instructing relevant hardware such as a computer device or a processor. The computer program can be stored in a non-transitory computer-readable storage medium, and when the computer program is executed, the steps of the present invention are caused to be executed. Depending on the situation, any reference to a memory, storage, database, or other medium herein may include non-volatile and / or volatile memory. Examples of non-volatile memory include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, magnetic tape, floppy disk, magneto-optical data storage device, optical data storage device, hard disk, solid state disk, etc. Examples of volatile memory include random access memory (RAM), external cache memory, etc.

[0072] It is understandable that the present invention has the following beneficial effects: 1. Improve the real-time performance and accuracy of monitoring: By setting multiple monitoring points on the running route of the rail locomotive and installing cameras at each monitoring point to collect image information of the rail area in real time, and using deep learning algorithms to quickly and accurately identify the images, potential dangerous situations near the rails can be detected in a timely manner, such as detecting the situation where a person appears on the rails in real time.

[0073] 2. Intelligent control of the running state of the rail locomotive: According to the image information collected by the monitoring points, analyze the actual distance between the person and the rail locomotive, and combine with the preset threshold to intelligently control the running state of the rail locomotive, such as normal driving or starting to resume normal driving, normal braking and emergency braking, etc., greatly improving the running safety of the rail locomotive.

[0074] 3. Real-time communication monitoring: The system can monitor the communication status between the monitoring points and the rail locomotive in real time. If a communication anomaly occurs, an alarm signal is output in a timely manner to ensure the stability and reliability of the monitoring system.

[0075] 4. Improve traffic operation efficiency and stability: Timely and accurate monitoring and control greatly reduce the interruption of the operation of the rail locomotive caused by unexpected situations. If a dangerous situation occurs, while ensuring safety, it reduces the possibility of damage to in-vehicle equipment or management chaos caused by the emergency braking of the locomotive, reduces the operation cost, and improves the operation efficiency and stability of the entire railway system.

[0076] The above-described technical features can be combined arbitrarily. Although not all possible combinations of these technical features are described, any combination of these technical features should be considered to be covered by this specification as long as such a combination does not exist in contradiction.

[0077] The specific implementation manners of the present invention described above do not constitute a limitation to the protection scope of the present invention. Any other corresponding changes and deformations made according to the technical concept of the present invention should be included within the protection scope of the claims of the present invention.

Claims

1. A remote intelligent monitoring method for rail locomotives, characterized in that, The method includes: S100. Set monitoring points at a preset distance on the running route of the rail locomotive, and install monitoring devices at each of the monitoring points. The monitoring device includes a camera, which is used to capture images of the target area at a preset time interval, and send the captured images and the timestamps when the images are captured to a control module set on the rail locomotive through a cloud server; S200. The control module identifies the images of the target area, and determines whether there are target images defined as people and target images defined as rails at the same time. If so, calculate the coordinates of the four vertices of the first anchor box for indicating each person and the second anchor box for indicating the rail respectively, establish a danger area according to the coordinates of the four vertices of the second anchor box, calculate the centroid coordinates of each person according to the coordinates of the four vertices of the first anchor box of each person, and determine whether there is a centroid coordinate of a certain person located within the danger area. If so, calculate the first position information corresponding to the centroid coordinate by using a homography matrix, obtain the second position information of the rail locomotive by using a GPS sensor set on the rail locomotive, and calculate the actual distance between the first position information and the second position information; S300. Control the running state of the rail locomotive according to the actual distance.

2. The remote intelligent monitoring method for rail locomotives according to claim 1, wherein The controlling the running state of the rail locomotive according to the actual distance includes: Judging whether the actual distance is greater than or equal to a first preset threshold and less than a second preset threshold; If the actual distance is greater than or equal to the first preset threshold and less than the second preset threshold, control the rail locomotive to perform normal braking.

3. The remote intelligent monitoring method for a rail locomotive according to claim 2, wherein The controlling the running state of the rail locomotive according to the actual distance includes: If the actual distance is greater than or equal to the second preset threshold, control the rail locomotive to run normally or start to resume normal running, and the second preset threshold is greater than the preset normal braking distance of the rail locomotive.

4. The remote intelligent monitoring method for rail locomotives according to claim 2, characterized in that, The controlling the running state of the rail locomotive according to the actual distance includes: If the actual distance is less than the first preset threshold, control the rail locomotive to perform emergency braking, and the first preset threshold is greater than the preset emergency braking distance of the rail locomotive.

5. The remote intelligent monitoring method for rail locomotives according to claim 1, wherein, The method includes: If there are no target images defined as people and target images defined as rails at the same time, control the rail locomotive to run normally or start to resume normal running.

6. The remote intelligent monitoring method for rail locomotives according to claim 1, wherein The control module's identifying the images of the target area includes: Form a data set with the images captured by the camera, annotate the images in the data set, and then train a YOLO model, and use the YOLO model to identify people and rails in the images of the target area.

7. The remote intelligent monitoring method for rail locomotives according to claim 1, characterized in that The method further includes: If there is no centroid coordinate of a certain person located within the danger area, control the rail locomotive to run normally or start to resume normal running.

8. The remote intelligent monitoring method for rail locomotives according to claim 1, characterized in that, The method further includes: Judge whether the control module receives the next frame image of a certain monitoring point within a preset time after receiving the current frame image of the monitoring point. If not, output an alarm signal regarding communication abnormality of the monitoring point; If so, output a prompt signal regarding normal communication of the monitoring point.

9. A remote intelligent monitoring system for rail locomotives, characterized in that, including: An acquisition module, including a camera for taking images of a target area at preset time intervals, and sending the taken images and the timestamps when the images are taken to a control module set on a rail locomotive through a cloud server; A control module, configured to identify the images of the target area, determine whether there are target images defined as a person and target images defined as a rail simultaneously. If so, calculate the coordinates of the four vertices of the first anchor box for indicating each person and the second anchor box for indicating the rail respectively, establish a dangerous area according to the coordinates of the four vertices of the second anchor box, calculate the centroid coordinates of each person according to the coordinates of the four vertices of the first anchor box of each person, determine whether the centroid coordinate of a certain person is located within the dangerous area. If so, calculate the first position information corresponding to the centroid coordinate by using a homography matrix, obtain the second position information of the rail locomotive by using a GPS sensor set on the rail locomotive, and calculate the actual distance between the first position information and the second position information; and is used to control the running state of the rail locomotive according to the actual distance.

10. An electronic device, characterized in that, It includes: A memory; And a processor, wherein computer-readable instructions are stored on the memory, and when the computer-readable instructions are executed by the processor, the rail locomotive remote intelligent monitoring method according to any one of claims 1 to 8 is implemented.

Images