Analytic model training method and device for cable path diagram, analytical method and device and medium

Abstract

The invention relates to an analytic model training method, an analytic method and device for a cable path diagram and a medium, and relates to the technical field of artificial intelligence, and the method comprises the steps: obtaining at least one first railway signal cable path diagram; intercepting at least one first screenshot containing railway signal equipment from each first railway signal cable path diagram, and generating a corresponding railway signal equipment position mask diagram according to the first screenshot; determining an associated cable path diagram of the railway signal equipment in each first screenshot in the corresponding first railway signal cable path diagram; and training a preset image segmentation network based on the railway signal equipment position mask graph corresponding to each first screenshot, the associated cable path graph and the first railway signal cable path graph to obtain a cable path graph analysis model. The method has the effect of improving the troubleshooting efficiency.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a cable path diagram analysis model training method, analysis method, device and medium. Background Art

[0002] A cable routing diagram is an engineering drawing that shows in detail the physical orientation and layout path of cables or wires in a specific area. The most common cable routing diagram is the railway signal cable routing diagram, which is an engineering drawing that describes the layout, connection and transmission path of various signal cables in the railway signal system, including different railway signal equipment with connected relationships. Railway signal equipment refers to various equipment used to control and supervise the operation of trains to ensure safe and efficient train travel.

[0003] At present, railway signal equipment occasionally has abnormalities. Once the railway signal equipment has an abnormality and cannot be used normally, it will directly affect the safety of railway driving. The usual method is to quickly find the cause of the abnormality by checking the cable path. In this method, experienced personnel are required to find the cable path that needs to be checked from the engineering drawings. However, the various cable paths in the engineering drawings are complicated, and it takes a long time to determine the cable path that needs to be checked, resulting in low troubleshooting efficiency. Summary of the invention

[0004] In order to improve the efficiency of troubleshooting, the present application provides a cable path diagram analysis model training method, analysis method, device and medium.

[0005] In a first aspect of the present application, a cable path diagram analysis model training method is provided, which specifically includes: Acquire at least one first railway signal cable routing diagram, wherein the first railway signal cable routing diagram includes different railway signal devices having a connection relationship; intercepting at least one first screenshot including railway signal equipment from each of the first railway signal cable path diagrams, and generating a corresponding railway signal equipment position mask diagram according to the first screenshot; Determine an associated cable path diagram of each railway signal device in the first screenshot in the corresponding first railway signal cable path diagram, wherein the associated cable path diagram is a portion of the cable path diagram that is most associated with the corresponding railway signal device when an abnormality occurs; Based on the railway signal equipment position mask map, the associated cable path map and the first railway signal cable path map corresponding to each of the first screenshots, the preset image segmentation network is trained to obtain a cable path map parsing model. The cable path map parsing model is a model that outputs the associated cable path map of the corresponding railway signal equipment based on the input railway signal cable path map and the railway signal equipment position mask map.

[0006] By adopting the above technical solution, after obtaining the first railway signal cable path map, a first screenshot containing railway signal equipment is intercepted from the first railway signal cable path map, and then the corresponding railway signal equipment position mask map is determined, thereby characterizing the position information of the railway signal equipment in the corresponding first railway signal cable path map. Furthermore, the railway signal equipment position mask map corresponding to a single first screenshot, the first railway signal cable path map and the associated cable path map are used as training data to train the image segmentation network so that it can continuously learn and have the ability to segment the associated cable path map corresponding to the railway signal equipment from the first railway signal cable path map, and finally obtain a cable path map parsing model. Once an abnormality occurs in the railway signal equipment, only the position mask map of the abnormal railway signal equipment and the railway signal cable path map where it is located need to be input into this cable path map parsing model, so that the cable path that needs to be checked can be quickly determined, thereby improving the efficiency of troubleshooting.

[0007] Optionally, the training of a preset image segmentation network based on the railway signal equipment position mask map, the associated cable path map and the first railway signal cable path map corresponding to each of the first screenshots specifically includes: Converting the first railway signal cable path diagram corresponding to each of the first screenshots into an input cable path diagram in a picture format, and adjusting the size of the input cable path diagram to a first preset size to obtain corresponding first input data; Adjusting the size of the railway signal equipment position mask image corresponding to each of the first screenshots to the first preset size to obtain corresponding second input data; Performing a concatenation operation on the first input data and the second input data corresponding to the same first screenshot to obtain final input data; Based on the associated cable path diagrams corresponding to the final input data and the first screenshots, a preset image segmentation network is trained.

[0008] By adopting the above technical solution, the first railway signal cable path diagram is converted into a picture format and the picture size is adjusted to a first preset size. At the same time, the size of the railway signal equipment position mask map corresponding to the first screenshot is also adjusted to the same first preset size. Finally, the first input data and the second input data after the size adjustment are concatenated to obtain the final input data, so that the final input data is input into the cable path diagram parsing model for training, thereby improving the training effect of the cable path diagram parsing model.

[0009] In a second aspect of the present application, a cable path diagram analysis method is provided, which specifically includes: Obtain a second railway signal cable path diagram including abnormal railway signal equipment, a cable path diagram parsing model obtained by using the cable path diagram parsing model training method described in the first aspect, and obtain a target equipment position mask diagram corresponding to the abnormal railway signal equipment; Preprocessing the second railway signal cable path map and the target equipment location mask map to obtain target input data; Inputting the target input data into the cable path diagram analysis model to obtain a target associated cable path diagram corresponding to the abnormal railway signal equipment; The target associated cable route diagram is superimposed on the second railway signal cable route diagram and highlighted.

[0010] By adopting the above technical solution, the trained cable path diagram parsing model is called, and the second railway signal cable path diagram containing abnormal railway signal equipment and the target equipment position mask map are preprocessed and input into the cable path diagram parsing model. The cable path diagram parsing model predicts the target associated cable path diagram corresponding to the abnormal railway signal equipment, and finally the target associated cable path diagram is superimposed on the second railway signal cable path diagram. In this way, personnel can determine the cable path that needs to be troubleshooted based on the target associated cable path diagram, without having to manually search for the cable path that needs to be troubleshooted from the complex cable paths in the second railway signal cable path diagram, thereby improving the troubleshooting efficiency.

[0011] Optionally, the step of obtaining a target device location mask map corresponding to the abnormal railway signal device specifically includes: Acquire at least one third railway signal cable routing diagram, and capture at least one second screenshot not including railway signal equipment and a third screenshot including railway signal equipment from each of the third railway signal cable routing diagrams; Determine each of the second screenshots and each of the third screenshots as an image binary classification data set, and train a preset binary classification neural network through the image binary classification data set to obtain a trained final binary classification network and store it in a preset database, wherein the final binary classification network can determine whether the input screenshot contains railway signal equipment; Intercepting an initial screenshot corresponding to the abnormal railway signal device from the second railway signal cable path diagram, and inputting the initial screenshot into the final binary classification network to obtain a corresponding output result; According to the output result, determining whether the corresponding initial screenshot contains railway signal equipment, and if so, generating a target equipment position mask map corresponding to the abnormal railway signal equipment according to the initial screenshot; If not, repeat the step of capturing the initial screenshot corresponding to the abnormal railway signal equipment from the second railway signal cable route diagram.

[0012] By adopting the above technical solution, if it is determined according to the output result that the initial screenshot contains railway signal equipment, it means that the initial screenshot successfully captures the abnormal railway signal equipment, and then the corresponding target device position mask map is determined; conversely, if the initial screenshot does not contain this railway signal equipment, it means that the initial screenshot fails to successfully capture the abnormal railway signal equipment, and then the step of capturing the initial screenshot corresponding to the abnormal railway signal equipment from the second railway signal cable path map is repeated, and then the re-captured initial screenshot is input into the final binary classification network, and it is determined whether the initial screenshot contains abnormal railway signal equipment according to the output result, so that the determined target device position mask map is more accurate.

[0013] Optionally, preprocessing the second railway signal cable path map and the target device location mask map to obtain target input data specifically includes: Converting the second railway signal cable path diagram into a picture format and adjusting the size to a second preset size to obtain first target input data; The size of the target device position mask image is adjusted to the second preset size to obtain second target input data, and the first target input data and the second target input data are concatenated to obtain target input data.

[0014] By adopting the above technical solution, the second railway signal cable path diagram is converted into a picture format and the picture size is adjusted to a second preset size. At the same time, the size of the target device position mask map is also adjusted to the same second preset size. Finally, the first target input data after the size adjustment is concatenated with the second target input data to obtain the target input data, so that the target input data is input into the cable path diagram parsing model, thereby improving the performance of the cable path diagram parsing model and making the accuracy of predicting the associated cable path diagram of abnormal railway signal equipment through the cable path diagram parsing model higher.

[0015] Optionally, the method further includes: Obtaining historical anomalies of historical railway signal equipment in which anomalies have occurred, counting the first occurrence times of each of the historical anomalies, and selecting the first occurrence times of each of the historical anomalies in descending order of the first occurrence times as an anomaly that is likely to occur; Obtaining the abnormal environment factor of the historical associated cable path diagram corresponding to the occurrence of a single abnormality in the historical railway signal equipment, and counting the second occurrence number of each abnormal environment factor, wherein the abnormal environment factor is the abnormal environment in which the cable corresponding to the associated cable path diagram in which the fault is found is located; According to the descending order of the second occurrence times, the abnormal environmental factors with the second number are selected from each of the abnormal environmental factors and determined as the corresponding abnormal factors that are prone to abnormality; Calculating a first weight of each of the prone abnormalities and a second weight of each corresponding prone abnormality factor, wherein the first weight is a ratio of a first occurrence number of each prone abnormality to a sum of first occurrence numbers of all prone abnormalities, and the second weight is a ratio of a second occurrence number of a single prone abnormality factor corresponding to the prone abnormality to a sum of second occurrence numbers of all corresponding prone abnormality factors; When the target abnormality occurring in the abnormal railway signal equipment is the prone abnormality, the target associated cable path diagram is verified based on the first weight of the target abnormality and the second weight of the prone abnormality-inducing factor in the abnormal environmental factors of the target associated cable path diagram.

[0016] By adopting the above technical solution, the more the first occurrence times, the more likely the railway signal equipment is to have the corresponding historical anomaly. Then, according to the order of the first occurrence times from large to small, the historical anomalies with the first number are selected from each historical anomaly and are determined as abnormalities that are prone to occur, that is, abnormalities that are prone to occur in the railway signal equipment. Further, the more the second occurrence times, the more likely the cables in the corresponding historical association path diagram are to have corresponding abnormal environmental factors. Further, according to the order of the second occurrence times from large to small, the abnormal environmental factors with the second number are selected from each abnormal environmental factor, that is, factors that are prone to induce abnormalities. Finally, according to the first weight of the target abnormality and the second weight of the factors that are prone to induce abnormalities corresponding to the target-associated cable path diagram, the rationality of troubleshooting through the target-associated cable path diagram is analyzed, and then the verification of the target-associated cable path diagram is realized, thereby improving the efficiency of troubleshooting.

[0017] Optionally, the verifying the target-associated cable path diagram based on the first weight of the target abnormality and the second weight of each easily-inducing abnormality factor in the abnormal environment factor of the target-associated cable path diagram specifically includes: Calculate the first weight of the target abnormality and the second weight of each abnormality-prone factor in the abnormal environment factor of the target-associated cable path diagram to obtain a corresponding weight product; Summing the weighted products to obtain a weighted product sum, and comparing the weighted product sum with a preset product sum threshold; If the weight product sum is greater than the product sum threshold, it is determined that the target associated cable path diagram has been verified, and the target cable path corresponding to the target associated cable path diagram is checked.

[0018] By adopting the above technical scheme, when the target abnormality of the abnormal railway signal equipment is prone to abnormality, the weight product of the first weight of the target abnormality and the second weight of each easily induced abnormal factor in the abnormal environmental factor of the target associated cable path diagram is calculated. The larger the weight product, the greater the probability that the target abnormality is caused by the corresponding easily induced abnormal factor. Finally, the weight products are summed to obtain the weight product sum. The larger the weight product sum, the greater the possibility that the target abnormality is caused by the easily induced abnormal factor in the abnormal environmental factor of the target associated cable path diagram. The weight product is compared with the preset product and threshold value. If the weight product sum is greater than the product and threshold value, it means that the target abnormality is likely to be caused by the easily induced abnormal factor in the abnormal environmental factor of the target associated cable path diagram. It means that the target cable path in the target associated cable path diagram is indeed the cable path that needs to be checked, then it is determined that the target associated cable path diagram is verified. By checking the target cable path in this target associated cable path diagram, it is highly likely that the fault point can be found, thereby improving the efficiency of the check.

[0019] Optionally, the checking the target cable path corresponding to the target associated cable path diagram specifically includes: Determine at least one target cable segment from the target cable path corresponding to the target associated cable path diagram, wherein the target cable segment is a portion of the cables in the target cable path where there are factors that are prone to inducing abnormalities; Determine the troubleshooting priority of the corresponding target cable segment according to the weight product, the larger the weight product is, the higher the troubleshooting priority of the corresponding target cable segment is; According to the order of the checking priority from high to low, the corresponding target cable segments in the target cable path are checked in sequence.

[0020] By adopting the above technical solution, the inspection priority of the corresponding target cable segment is determined according to the weight product corresponding to the easily induced abnormal factors in the target cable path. The larger the weight product, the greater the probability that the corresponding easily induced abnormal factors will cause the target abnormality, and the higher the inspection priority of the corresponding target cable segment. Finally, according to each inspection priority, the corresponding target cable segment in the target cable path is inspected in turn, thereby further improving the inspection efficiency of the fault that causes the target abnormality.

[0021] In a third aspect of the present application, a cable path diagram analytical model training device is provided, specifically comprising: An information acquisition module, used to acquire at least one first railway signal cable path diagram, wherein the first railway signal cable path diagram includes different railway signal devices having a connection relationship; A mask determination module, configured to capture at least one first screenshot containing railway signal equipment from each of the first railway signal cable path diagrams, and generate a corresponding railway signal equipment position mask diagram according to the first screenshot; A path determination module, used to determine an associated cable path diagram of each railway signal device in the first screenshot in the corresponding first railway signal cable path diagram, wherein the associated cable path diagram is a cable path diagram of a portion of the railway signal device that is most associated with the corresponding railway signal device when an abnormality occurs; A model training module is used to train a preset image segmentation network based on the railway signal equipment position mask map, the associated cable path map and the first railway signal cable path map corresponding to each of the first screenshots to obtain a cable path map parsing model. The cable path map parsing model is a model that outputs an associated cable path map of the corresponding railway signal equipment based on the input railway signal cable path map and the railway signal equipment position mask map.

[0022] By adopting the above technical solution, the information acquisition module obtains the first railway signal cable path map, the mask determination module generates the corresponding railway signal equipment position mask map according to the first screenshot, and then the path determination module determines the associated cable path map corresponding to the railway signal equipment in the first screenshot. Finally, based on the railway signal equipment position mask map corresponding to the first screenshot, the associated cable path map and the first railway signal cable path map, the image segmentation network is trained to obtain a cable path map parsing model.

[0023] In a fourth aspect of the present application, a computer-readable storage medium is provided, in which a computer program is stored. When the computer program is loaded and executed by a processor, the method steps described in any one of the first aspects are performed.

[0024] In summary, the present application includes at least one of the following beneficial technical effects: a first screenshot containing railway signal equipment is intercepted from a first railway signal cable path diagram, and then the corresponding railway signal equipment position mask diagram is determined, thereby characterizing the position information of the railway signal equipment in the corresponding first railway signal cable path diagram. Furthermore, the railway signal equipment position mask diagram corresponding to a single first screenshot, the first railway signal cable path diagram, and the associated cable path diagram are used as training data to train the image segmentation network, so that it continuously learns and has the ability to segment the associated cable path diagram corresponding to the railway signal equipment from the first railway signal cable path diagram, and finally obtains a cable path diagram parsing model. Once an abnormality occurs in the railway signal equipment, only the position mask diagram of the abnormal railway signal equipment and the railway signal cable path diagram in which it is located need to be input into this cable path diagram parsing model, so that the cable path that needs to be checked can be quickly determined, thereby improving the efficiency of troubleshooting. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is a flow chart of a cable path diagram analysis model training method provided in an embodiment of the present application; Figure 2 It is a schematic diagram of a flow chart of a cable path diagram analysis method provided in an embodiment of the present application; Figure 3 It is a structural schematic diagram of a cable path diagram analytical model training device provided in an embodiment of the present application; Figure 4 It is a structural schematic diagram of a cable path diagram analysis device provided in an embodiment of the present application; Figure 5 It is a structural schematic diagram of another cable path diagram analysis device provided in an embodiment of the present application.

[0026] Explanation of the accompanying drawings: 11. Information acquisition module; 12. Mask determination module; 13. Path determination module; 14. Model training module; 21. Data acquisition module; 22. Image processing module; 23. Model prediction module; 24. Path superposition module; 25. Path verification module. DETAILED DESCRIPTION

[0027] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0028] In the description of the embodiments of the present application, words such as "illustrative", "for example" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "illustrative", "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "illustrative", "for example" or "for example" is intended to present related concepts in a concrete way.

[0029] In the description of the embodiments of the present application, the term "and / or" is only a kind of association relationship describing the associated objects, indicating that there may be three kinds of relationships, for example, A and / or B, which can represent: A exists alone, B exists alone, and A and B exist at the same time. In addition, unless otherwise specified, the meaning of the term "multiple" refers to two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Thus, the features defined as "first" and "second" can explicitly or implicitly include one or more of the features. The terms "include", "comprise", "have" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0030] See also Figure 1 The present application embodiment discloses a flow chart of a cable path diagram analysis model training method, which can be implemented by a computer program or run on a cable path diagram analysis model training device based on the von Neumann system. The computer program can be integrated into an application or run as an independent tool application, specifically including: S101: Obtain at least one first railway signal cable route map.

[0031] Specifically, the first railway signal cable path diagram refers to a schematic diagram used to describe the layout, connection and transmission paths of various signal cables in the railway signal system, which is very important for maintenance personnel to perform troubleshooting and line modification. In an embodiment of the present application, the first railway signal cable path diagram includes different railway signal equipment that are connected. The first railway signal cable path diagram can be an electronic diagram in Computer-Aided Design (CAD) format. In other embodiments, it can also be an electronic diagram in vector graphics format. Among them, railway signal equipment refers to various equipment used to control and supervise the operation of trains to ensure safe and efficient travel of trains. Exemplarily, railway signal equipment can be signal machines, switch machines, track circuits, and the like.

[0032] In addition, the embodiment of the present application discloses a method for training an analytical model of a cable path diagram, the execution subject of which is a server. The server may be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. The server and the terminal are directly or indirectly connected through a wired or wireless network. Furthermore, a feasible way to obtain the first railway signal cable path diagram is: personnel upload the first railway signal cable path diagram corresponding to different railway projects through the terminal, and the terminal sends each first railway signal cable path diagram to the server, and the server finally obtains at least one first railway signal cable path diagram. In other embodiments, at least one first railway signal cable path diagram corresponding to a single railway project may also be sent to the server through the terminal. Among them, the terminal may be a smart phone or a personal computer.

[0033] S102: capturing at least one first screenshot including railway signal equipment from each first railway signal cable routing diagram, and generating a corresponding railway signal equipment position mask diagram according to the first screenshot.

[0034] Specifically, after the first railway signal cable route map is obtained, random screenshots are taken from each first railway signal cable route map, and the size of the screenshot is a preset fixed size. A feasible way to capture the first screenshot is: a person sends an interception instruction to the server through a terminal, and the screenshot instruction includes an area for capturing a screenshot of the first railway signal cable route map (an area including railway signal equipment). After the server obtains the interception instruction, at least one first screenshot including railway signal equipment is captured from each first railway signal cable route map through a preset screenshot tool. The screenshot tool can be a Snipaste tool or a ShareX tool. In other embodiments, a person can also complete the interception of the first screenshot of each first railway signal cable route map through a terminal, and finally send it to the server through the terminal.

[0035] Furthermore, a corresponding railway signal equipment position mask map is generated according to each first screenshot, that is, an equipment position mask map of the railway signal equipment contained in each first screenshot, so as to characterize the position information of the railway signal equipment in the corresponding first railway signal cable path map. A feasible generation method is: input the first screenshot and the corresponding first railway signal cable path map into a preset OpenCV tool to generate a corresponding railway signal equipment position mask map.

[0036] S103: Determine the associated cable routing diagram of the railway signal equipment in each first screenshot in the corresponding first railway signal cable routing diagram, the associated cable routing diagram being the cable routing diagram of the part that is most associated with the corresponding railway signal equipment when an abnormality occurs.

[0037] Specifically, after the railway signal equipment position mask diagram corresponding to each first screenshot is determined, it is necessary to determine the associated cable path diagram of the railway signal equipment in the first screenshot in the first railway signal cable path diagram corresponding to the first screenshot. The associated cable path diagram is the part of the cable path diagram that is most associated with the corresponding railway signal equipment when an abnormality occurs, that is, when a cable in the associated cable path diagram has a fault point, it is most likely to cause the corresponding railway signal equipment to have an abnormality. A feasible determination method is: obtain the historical cable path diagram of a single railway signal equipment in the historical abnormality from the cable path inspection record, filter out the successful historical cable path diagram of the troubleshooting that successfully finds the fault point from each historical cable path diagram, and count the number of occurrences of each successful historical cable path diagram, and determine the successful historical cable path diagram with the most occurrences as the associated cable path diagram of the corresponding railway signal equipment. In other embodiments, the path diagram selection instruction sent by the receiving terminal for a single first railway signal cable path diagram can also be determined as the associated cable path diagram of the railway signal equipment. Exemplarily, the associated cable path diagram of railway signal equipment A includes railway signal equipment A itself. It should be noted that the partial cable path diagram is a portion of the cable path in the corresponding first railway signal cable path diagram, and the cable path refers to the specific route of the cable from one connection point to another, connecting at least one railway signal equipment. The first railway signal cable path diagram includes a plurality of different cable paths.

[0038] S104: training a preset image segmentation network based on the railway signal equipment position mask map, the associated cable path map, and the first railway signal cable path map corresponding to each first screenshot to obtain a cable path map parsing model.

[0039] Specifically, after the railway signal equipment position mask map and the associated cable path map corresponding to each first screenshot are determined, the first railway signal cable path map corresponding to each first screenshot is converted into an input cable path map in a picture format. In the embodiment of the present application, the picture format may be a JPG format, and in other embodiments, it may also be a PNG format. Then, the size of the input cable path map is adjusted to a first preset size to obtain the corresponding first input data. In the embodiment of the present application, the first preset size may be (1024, 1024), and in other embodiments, the first preset size may also be (512, 512), so as to meet the input size requirements of the deep learning model as much as possible and improve the performance of the deep learning model. Furthermore, the size of the railway signal equipment position mask map corresponding to the same first screenshot is also adjusted to the first preset size accordingly to obtain the corresponding second input data. Furthermore, the first input data and the second input data corresponding to the same first screenshot are concatenated by a preset concate function to obtain the final input data. Furthermore, the final input data and the associated cable path diagram corresponding to each first screenshot are summarized as a training data set, and the training data set is divided into a training set, a validation set, and a test set according to a first division ratio. The preset image segmentation network is trained by the training set, the validation set, and the test set to obtain a cable path diagram parsing model. Among them, the first division ratio can be 7:2:1, or 6:3:1. In other embodiments, the training data set can also be divided into a training set and a test set according to a second division ratio, and the second division ratio can be 8:2 or 7:3. The railway signal cable path diagram and the railway signal equipment position mask diagram are input into this cable path diagram parsing model, and the model can output the associated cable path diagram of the corresponding railway signal equipment. In addition, the image segmentation network can be a unet network. In other embodiments, the image segmentation network can be a DeepLab network. The image segmentation network mainly uses deep learning technology to achieve image segmentation tasks.

[0040] It should be noted that when an abnormality occurs in railway signal equipment, in order to quickly determine the cable path that needs to be checked and improve the efficiency of fault detection, the railway signal cable path map where the abnormal railway signal equipment is located and the location mask map of the abnormal railway signal equipment are input into this cable path map parsing model, so as to quickly determine the cable path that needs to be checked, that is, the associated cable path map.

[0041] See also Figure 2 The present application embodiment discloses a flow chart of a cable path diagram analysis method, which can be implemented by a computer program or run on a cable path diagram analysis device based on the von Neumann system. The computer program can be integrated into an application or run as an independent tool application, specifically including: S201: Obtain a second railway signal cable path map including abnormal railway signal equipment, obtain a cable path map parsing model using the cable path map parsing model training method in any of the above embodiments, and obtain a target equipment position mask map corresponding to the abnormal railway signal equipment.

[0042] Specifically, after the cable path diagram analysis model is determined, if there is currently an abnormal railway signal device, in order not to affect the railway driving safety, it is necessary to check the cable path as soon as possible and quickly determine the fault point, then obtain the second railway signal cable path diagram containing this abnormal railway signal device. A feasible way to obtain it is: the railway operation and maintenance personnel send the CAD electronic drawing of the overall cable path containing this abnormal railway signal device, that is, the second railway signal cable path diagram, to the server through the terminal, and the server finally obtains the second railway signal cable path diagram. Further, obtain the target device location mask map corresponding to this abnormal railway signal device, that is, the mask map that characterizes the location information of the abnormal railway signal device in the second railway signal cable path map. A feasible acquisition method is: intercepting the initial screenshot corresponding to the abnormal railway signal device from the second railway signal cable path map, then calling the preset final binary classification network from the database, inputting the initial screenshot into the final binary classification network, and obtaining an output result. In the embodiment of the present application, the output result is the predicted probability value of the abnormal railway signal device contained in the initial screenshot, and the value range is [0,1]. If the output result exceeds the preset critical value, the critical value is 0.5. For example, if the output result is 0.8, which exceeds the critical value of 0.5, then it is determined that the corresponding initial screenshot contains abnormal railway signal equipment, then the initial screenshot and the second railway signal cable path map are input into the preset OpenCV tool to obtain the corresponding target device position mask map; conversely, if the output result does not exceed the critical value, it means that the abnormal railway signal equipment is not included in the initial screenshot, then repeat the step of intercepting the initial screenshot corresponding to the abnormal railway signal equipment from the second railway signal cable path map, and then input the re-intercepted initial screenshot into the final binary classification network, and judge whether the initial screenshot contains abnormal railway signal equipment according to the output result. Among them, the final binary classification network can determine whether the input screenshot contains railway signal equipment. In other embodiments, the size of the initial screenshot can be adjusted to (32,32) before being input into the final binary classification network, thereby improving the performance of the model.

[0043] Further, the method for determining the final binary classification network is as follows: obtaining at least one third railway signal cable path map, and intercepting at least one second screenshot that does not include railway signal equipment and a third screenshot that includes railway signal equipment from each third railway signal cable path map. For details, please refer to steps S101 and S102, which will not be repeated here. Then, each second screenshot and each third screenshot are determined as an image binary classification data set, and this image binary classification data set is divided into a training set, a verification set and a test set according to a first division ratio. In other embodiments, this image binary classification data set can also be divided into a training set and a test set according to a second division ratio. Further, the preset binary classification neural network is trained by the divided training set, verification set and test set to obtain the final binary classification network and store it in the database. Among them, the binary classification neural network can adopt a convolutional neural network. In other embodiments, the binary classification can also adopt a recurrent neural network. It should be noted that the second railway signal cable path map includes several different cable paths.

[0044] S202: Preprocess the second railway signal cable path map and the target equipment location mask map to obtain target input data.

[0045] Specifically, after the target device position mask map corresponding to the abnormal railway signal device is determined, the second railway signal cable path map and the target device position mask map are preprocessed respectively, so that the input data is more in line with the input requirements of the model. The specific preprocessing process is: convert the second railway signal cable path map into an image format and adjust the size to a second preset size to obtain the first target input data, wherein the size adjustment can be performed through the cv2.resize function in the OpenCV library. In other embodiments, the size adjustment can also be performed through the transforms.Resize function in PyTorch. In addition, similarly, the size of the target device position mask map is adjusted to the second preset size to obtain the second target input data. The second preset size can be (1024, 1024) or (512, 512).

[0046] Finally, the first target input data and the second target input data are concatenated to obtain the target input data. For details, please refer to step S104, which will not be described again.

[0047] S203: Input the target input data into the cable path diagram analysis model to obtain the target associated cable path diagram corresponding to the abnormal railway signal equipment.

[0048] S204: superimposing the target associated cable route map onto the second railway signal cable route map and highlighting it.

[0049] Specifically, the target input data is input into the cable path diagram analysis model to obtain the cable path diagram that needs to be troubleshooted for this abnormal railway signal equipment, thereby improving the efficiency of troubleshooting and avoiding personnel spending a long time to determine the cable path diagram that needs to be troubleshooted. Furthermore, the target associated cable path diagram is finally superimposed on the second railway signal cable path diagram and highlighted, thereby providing a reference for subsequent operation and maintenance personnel to troubleshoot.

[0050] In other embodiments, historical anomalies of historical railway signal equipment where anomalies have occurred are obtained from the historical investigation records of cable paths, and the first occurrence times of each historical anomaly are counted. The more the first occurrence times, the more likely the railway signal equipment is to have the corresponding historical anomaly. Then, in order of the first occurrence times from large to small, the historical anomalies with the first number are selected from each historical anomaly and are determined as easily occurring anomalies, that is, easily occurring anomalies of the railway signal equipment. Further, when a single easily occurring anomaly occurs in the historical railway signal equipment, the abnormal environmental factors of the corresponding historical associated cable path diagram are obtained from the historical investigation records of the cable path, that is, the abnormal environment in which the cable corresponding to the associated cable path diagram where the fault is found is located, such as high temperature, high pressure, etc. The second occurrence times of each abnormal environmental factor are counted. The more the second occurrence times, the more likely the cable in the corresponding historical associated path diagram is to have the corresponding abnormal environmental factor. Further, in order of the second occurrence times from large to small, the abnormal environmental factors with the second number are selected from each abnormal environmental factor and determined as the easily induced abnormal factors corresponding to this single easily occurring anomaly, that is, the abnormal environmental factors that are easily induced in the associated cable path to cause the railway signal equipment to have the corresponding easily occurring anomaly.

[0051] Furthermore, the first weight of each prone abnormality and the corresponding second weight of each prone abnormality factor are calculated, the first weight being the ratio of the first occurrence number of each prone abnormality to the sum of the first occurrence numbers of all prone abnormalities, and the second weight being the ratio of the second occurrence number of a single prone abnormality factor corresponding to the prone abnormality to the sum of the second occurrence numbers of all prone abnormalities. Further, in the case where the target abnormality of the abnormal railway signal equipment is prone to abnormality, the weight product of the first weight of the target abnormality and the second weight of each prone abnormality factor in the abnormal environmental factors of the target-associated cable path diagram is calculated. The larger the weight product, the greater the probability that the target abnormality is caused by the corresponding prone abnormality factor. Finally, the sum of the weight products is obtained to obtain the weight product sum. The larger the weight product sum, the greater the possibility that the target abnormality is caused by the prone abnormality factor in the abnormal environmental factors of the target-associated cable path diagram. The weight product is compared with the preset product and threshold. If the weight product is greater than the product and threshold, it means that the target abnormality is likely to be caused by the easily induced abnormality factor in the abnormal environmental factors of the target-associated cable path diagram, indicating that the target cable path in the target-associated cable path diagram is indeed the cable path that needs to be checked. Then it is determined that the target-associated cable path diagram has passed the verification. By checking the target cable path in this target-associated cable path diagram, the fault point can be found with a high probability, which improves the efficiency of the investigation. Finally, the operation and maintenance personnel check the target cable path corresponding to the target-associated cable path diagram. It should be noted that the abnormal environmental factors of the target-associated cable path diagram are determined by obtaining the corresponding picture information through the camera on the cable path corresponding to the target-associated cable path diagram, and the personnel determine the abnormal environmental factors therein and send them to the server through the terminal.

[0052] In another embodiment, the target cable path corresponding to the target associated cable path diagram is checked. An achievable implementation method is: the part of the cables in the target cable path that have the factors that easily induce abnormalities is determined as the target cable segment, and then the checking priority of the corresponding target cable segment is determined according to the weight product. The larger the weight product, the greater the probability that the corresponding factors that easily induce abnormalities will cause the target abnormality, and the higher the checking priority of the corresponding target cable segment. Specifically, the checking priority corresponding to each weight product can be matched through a preset priority matching table. The priority matching table includes different weight product range intervals and corresponding checking priorities. The checking priority takes values ​​of 1-10. For example, the weight product range interval corresponding to the checking priority 10 is a1~a2. If the weight product A is in this a1~a2 interval, it means that the highest checking priority of the corresponding target cable segment is 10. Finally, according to each checking priority, the corresponding target cable segments in the target cable path are checked in turn, so as to further improve the efficiency of troubleshooting faults that cause target abnormalities.

[0053] The implementation principle of the cable path diagram analysis model training method of the embodiment of the present application is: intercepting a first screenshot containing railway signal equipment from the first railway signal cable path diagram, and then determining its corresponding railway signal equipment position mask map, thereby characterizing the position information of the railway signal equipment in the corresponding first railway signal cable path diagram. Furthermore, the railway signal equipment position mask map corresponding to the single first screenshot, the first railway signal cable path diagram, and the associated cable path diagram are used as training data to train the image segmentation network so that it can continuously learn and have the ability to segment the associated cable path diagram corresponding to the railway signal equipment from the first railway signal cable path diagram, and finally obtain the cable path diagram analysis model. Once an abnormality occurs in the railway signal equipment, it only needs to input the abnormal railway signal equipment position mask map and the railway signal cable path diagram where it is located into this cable path diagram analysis model, so that the cable path that needs to be checked can be quickly determined, thereby improving the efficiency of troubleshooting.

[0054] The following is an embodiment of the device of the present application, which can be used to execute the embodiment of the method of the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the embodiment of the method of the present application.

[0055] See also Figure 3 , is a schematic diagram of the structure of the analytical model training device for the cable path diagram provided in the embodiment of the present application. The analytical model training device applied to the cable path diagram can be implemented as all or part of the device through software, hardware or a combination of both. The device includes an information acquisition module 11, a mask determination module 12, a path determination module 13 and a model training module 14.

[0056] The information acquisition module 11 is used to acquire at least one first railway signal cable path diagram, where the first railway signal cable path diagram includes different railway signal devices that are connected to each other; The mask determination module 12 is used to intercept at least one first screenshot containing railway signal equipment from each first railway signal cable path diagram, and generate a corresponding railway signal equipment position mask diagram according to the first screenshot; A path determination module 13 is used to determine an associated cable path diagram of the railway signal equipment in each first screenshot in the corresponding first railway signal cable path diagram, wherein the associated cable path diagram is a cable path diagram of a part of the railway signal equipment that is most associated with the corresponding railway signal equipment when an abnormality occurs; The model training module 14 is used to train the preset image segmentation network based on the railway signal equipment position mask map, the associated cable path map and the first railway signal cable path map corresponding to each first screenshot to obtain a cable path map parsing model. The cable path map parsing model is a model that outputs the associated cable path map of the corresponding railway signal equipment based on the input railway signal cable path map and the railway signal equipment position mask map.

[0057] Optionally, the model training module 14 is specifically used for: Converting the first railway signal cable path diagram corresponding to each first screenshot into an input cable path diagram in a picture format, and adjusting the size of the input cable path diagram to a first preset size to obtain corresponding first input data; Adjusting the size of the railway signal equipment position mask image corresponding to each first screenshot to a first preset size to obtain corresponding second input data; Perform a concatenation operation on the first input data and the second input data corresponding to the same first screenshot to obtain final input data; Based on the final input data and the associated cable path diagrams corresponding to the first screenshots, the preset image segmentation network is trained.

[0058] It should be noted that the analytical model training device for a cable path diagram provided in the above embodiment only uses the division of the above-mentioned functional modules as an example when executing the analytical model training method for a cable path diagram. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the analytical model training device for a cable path diagram provided in the above embodiment and the analytical model training method embodiment for a cable path diagram belong to the same concept, and the implementation process thereof is detailed in the method embodiment, which will not be repeated here.

[0059] See also Figure 4 , is a schematic diagram of the structure of the cable path diagram analysis device provided in the embodiment of the present application. The cable path diagram analysis device can be implemented as all or part of the device through software, hardware or a combination of both. The device includes a data acquisition module 21, an image processing module 22, a model prediction module 23 and a path superposition module 24.

[0060] The data acquisition module 21 is used to obtain a second railway signal cable path map containing abnormal railway signal equipment, a cable path map parsing model obtained by using the cable path map parsing model training method in any of the above embodiments, and obtain a target device position mask map corresponding to the abnormal railway signal equipment.

[0061] The image processing module 22 is used to pre-process the second railway signal cable path map and the target equipment position mask map to obtain target input data.

[0062] The model prediction module 23 is used to input the target input data into the cable path diagram analysis model to obtain the target associated cable path diagram corresponding to the abnormal railway signal equipment.

[0063] The path superposition module 24 is used to superimpose the target associated cable path diagram onto the second railway signal cable path diagram and highlight it.

[0064] Optionally, the image processing module 22 is specifically used for: Converting the second railway signal cable path diagram into a picture format and adjusting the size to a second preset size to obtain first target input data; The size of the target device position mask image is adjusted to a second preset size to obtain second target input data, and the first target input data and the second target input data are concatenated to obtain target input data.

[0065] Optional, such as Figure 5 As shown, the device also includes a path verification module 25, which is specifically used to: Obtain historical anomalies of historical railway signal equipment where anomalies have occurred, count the first occurrence times of each historical anomaly, and select the first historical anomaly from each historical anomaly in descending order of the first occurrence times to determine it as an anomaly that is likely to occur; Obtain the abnormal environment factors of the historical associated cable path diagram corresponding to the occurrence of a single abnormality in the historical railway signal equipment, and count the second occurrence times of each abnormal environment factor. The abnormal environment factor is the abnormal environment in which the cable corresponding to the associated cable path diagram where the fault is found is located; According to the order of the second occurrence number from large to small, the abnormal environmental factor with the second number is selected from each abnormal environmental factor and determined as the corresponding abnormal inducing factor that is prone to abnormality; Calculate the first weight of each prone abnormality and the second weight of each corresponding prone abnormality inducing factor, the first weight being the ratio of the first occurrence number of each prone abnormality to the sum of the first occurrence numbers of all prone abnormalities, and the second weight being the ratio of the second occurrence number of a single prone abnormality inducing factor corresponding to the prone abnormality to the sum of the second occurrence numbers of all corresponding prone abnormality inducing factors; When the target abnormality of the abnormal railway signal equipment is prone to abnormality, the target associated cable path diagram is verified based on the first weight of the target abnormality and the second weight of the prone to abnormality factor in the abnormal environmental factors of the target associated cable path diagram.

[0066] Optionally, the path verification module 25 is specifically used for: Calculate the first weight of the target abnormality and the second weight of each factor that is prone to abnormality in the abnormal environment factor of the target-related cable path diagram to obtain a corresponding weight product; The weighted products are summed to obtain a weighted product sum, and the weighted product sum is compared with a preset product sum threshold; If the weight product sum is greater than the product sum threshold, it is determined that the target associated cable path map has been verified, and the target cable path corresponding to the target associated cable path map is checked.

[0067] Optionally, the path verification module 25 is specifically used for: Determine at least one target cable segment from the target cable path corresponding to the target associated cable path diagram, the target cable segment being a portion of the cables in the target cable path that have factors that are prone to inducing abnormalities; According to the weight product, the troubleshooting priority of the corresponding target cable segment is determined. The larger the weight product is, the higher the troubleshooting priority of the corresponding target cable segment is. In descending order of priority, the corresponding target cable segments in the target cable path are checked in sequence.

[0068] The embodiment of the present application further discloses a computer-readable storage medium, and the computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, a cable path diagram analysis model training method and an analysis method of the above-mentioned embodiment are adopted.

[0069] Among them, the computer program can be stored in a computer-readable medium, the computer program includes computer program code, the computer program code can be in the form of source code, object code, executable file or certain middleware, etc. The computer-readable medium includes any entity or device that can carry the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the computer-readable medium includes but is not limited to the above-mentioned components.

[0070] Among them, through this computer-readable storage medium, a cable path diagram analysis model training method and analysis method of the above embodiment are stored in a computer-readable storage medium, and are loaded and executed on a processor to facilitate the storage and application of the above method.

[0071] An embodiment of the present application also discloses an electronic device, in which a computer program is stored in a computer-readable storage medium. When the computer program is loaded and executed by a processor, the above-mentioned cable path diagram analysis model training method and analysis method are adopted.

[0072] The electronic device may be a desktop computer, a laptop computer, a cloud server or other electronic device, and the electronic device includes but is not limited to a processor and a memory. For example, the electronic device may also include input and output devices, a network access device, and a bus.

[0073] Among them, the processor can adopt a central processing unit (CPU). Of course, according to actual usage, other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. can also be adopted. The general-purpose processor can adopt a microprocessor or any conventional processor, etc., and this application does not impose any restrictions on this.

[0074] Among them, the memory can be an internal storage unit of the electronic device, such as a hard disk or memory of the electronic device, or it can be an external storage device of the electronic device, such as a plug-in hard disk, a smart memory card (SMC), a secure digital card (SD) or a flash memory card (FC) equipped on the electronic device. Moreover, the memory can also be a combination of an internal storage unit and an external storage device of the electronic device. The memory is used to store computer programs and other programs and data required by the electronic device. The memory can also be used to temporarily store data that has been output or is to be output, and this application does not impose any restrictions on this.

[0075] Among them, through this electronic device, the analytical model training method and analytical method of a cable path diagram in the above-mentioned embodiment are stored in the memory of the electronic device, and are loaded and executed on the processor of the electronic device for easy use.

[0076] The above is only an exemplary embodiment of the present disclosure, and the scope of the present disclosure cannot be limited thereto. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the specification and practicing the disclosure here, those skilled in the art will easily think of other embodiments of the present disclosure. This application is intended to cover any modification, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the technical field not recorded in the present disclosure. The description and examples are only regarded as exemplary, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A cable path diagram analytical model training method, characterized in that: The method comprises: Acquire at least one first railway signal cable routing diagram, wherein the first railway signal cable routing diagram includes different railway signal devices having a connection relationship; intercepting at least one first screenshot including railway signal equipment from each of the first railway signal cable path diagrams, and generating a corresponding railway signal equipment position mask diagram according to the first screenshot; Determine an associated cable path diagram of each railway signal device in the first screenshot in the corresponding first railway signal cable path diagram, wherein the associated cable path diagram is a portion of the cable path diagram that is most associated with the corresponding railway signal device when an abnormality occurs; Based on the railway signal equipment position mask map, the associated cable path map and the first railway signal cable path map corresponding to each of the first screenshots, the preset image segmentation network is trained to obtain a cable path map parsing model. The cable path map parsing model is a model that outputs the associated cable path map of the corresponding railway signal equipment based on the input railway signal cable path map and the railway signal equipment position mask map.

2. The cable path diagram analytical model training method according to claim 1, characterized in that: The training of a preset image segmentation network based on the railway signal equipment position mask map, the associated cable path map and the first railway signal cable path map corresponding to each of the first screenshots specifically includes: Converting the first railway signal cable path diagram corresponding to each of the first screenshots into an input cable path diagram in a picture format, and adjusting the size of the input cable path diagram to a first preset size to obtain corresponding first input data; Adjusting the size of the railway signal equipment position mask image corresponding to each of the first screenshots to the first preset size to obtain corresponding second input data; Performing a concatenation operation on the first input data and the second input data corresponding to the same first screenshot to obtain final input data; Based on the associated cable path diagrams corresponding to the final input data and the first screenshots, a preset image segmentation network is trained.

3. A cable path diagram analysis method, characterized in that: The method comprises: Obtain a second railway signal cable path diagram including abnormal railway signal equipment, a cable path diagram parsing model obtained by the cable path diagram parsing model training method according to any one of claims 1 to 2, and obtain a target equipment position mask diagram corresponding to the abnormal railway signal equipment; Preprocessing the second railway signal cable path map and the target equipment location mask map to obtain target input data; Inputting the target input data into the cable path diagram analysis model to obtain a target associated cable path diagram corresponding to the abnormal railway signal equipment; The target associated cable route diagram is superimposed on the second railway signal cable route diagram and highlighted.

4. The cable path diagram analysis method according to claim 3, characterized in that: The step of obtaining a target device location mask map corresponding to the abnormal railway signal device specifically includes: Acquire at least one third railway signal cable routing diagram, and capture at least one second screenshot not including railway signal equipment and a third screenshot including railway signal equipment from each of the third railway signal cable routing diagrams; Determine each of the second screenshots and each of the third screenshots as an image binary classification data set, and train a preset binary classification neural network through the image binary classification data set to obtain a trained final binary classification network and store it in a preset database, wherein the final binary classification network can determine whether the input screenshot contains railway signal equipment; Intercepting an initial screenshot corresponding to the abnormal railway signal device from the second railway signal cable path diagram, and inputting the initial screenshot into the final binary classification network to obtain a corresponding output result; According to the output result, determining whether the corresponding initial screenshot contains railway signal equipment, and if so, generating a target equipment position mask map corresponding to the abnormal railway signal equipment according to the initial screenshot; If not, repeat the step of capturing the initial screenshot corresponding to the abnormal railway signal equipment from the second railway signal cable route diagram.

5. The cable path diagram analysis method according to claim 3, characterized in that: The preprocessing of the second railway signal cable path map and the target equipment location mask map to obtain target input data specifically includes: Converting the second railway signal cable path diagram into a picture format and adjusting the size to a second preset size to obtain first target input data; The size of the target device position mask image is adjusted to the second preset size to obtain second target input data, and the first target input data and the second target input data are concatenated to obtain target input data.

6. The cable path diagram analysis method according to claim 3, characterized in that: The method further comprises: Obtaining historical anomalies of historical railway signal equipment in which anomalies have occurred, counting the first occurrence times of each of the historical anomalies, and selecting the first occurrence times of each of the historical anomalies in descending order of the first occurrence times as an anomaly that is likely to occur; Obtaining the abnormal environment factor of the historical associated cable path diagram corresponding to the occurrence of a single abnormality in the historical railway signal equipment, and counting the second occurrence number of each abnormal environment factor, wherein the abnormal environment factor is the abnormal environment in which the cable corresponding to the associated cable path diagram in which the fault is found is located; According to the descending order of the second occurrence times, the abnormal environmental factors with the second number are selected from each of the abnormal environmental factors and determined as the corresponding abnormal factors that are prone to abnormality; Calculating a first weight of each of the prone abnormalities and a second weight of each corresponding prone abnormality factor, wherein the first weight is a ratio of a first occurrence number of each prone abnormality to a sum of first occurrence numbers of all prone abnormalities, and the second weight is a ratio of a second occurrence number of a single prone abnormality factor corresponding to the prone abnormality to a sum of second occurrence numbers of all corresponding prone abnormality factors; When the target abnormality occurring in the abnormal railway signal equipment is the prone abnormality, the target associated cable path diagram is verified based on the first weight of the target abnormality and the second weight of the prone abnormality-inducing factor in the abnormal environmental factors of the target associated cable path diagram.

7. The cable path diagram analysis method according to claim 6, characterized in that: The verifying of the target-associated cable path diagram based on the first weight of the target abnormality and the second weight of each easily-induced abnormality factor in the abnormal environment factor of the target-associated cable path diagram specifically includes: Calculate the first weight of the target abnormality and the second weight of each abnormality-prone factor in the abnormal environment factor of the target-associated cable path diagram to obtain a corresponding weight product; Summing the weighted products to obtain a weighted product sum, and comparing the weighted product sum with a preset product sum threshold; If the weight product sum is greater than the product sum threshold, it is determined that the target associated cable path diagram has been verified, and the target cable path corresponding to the target associated cable path diagram is checked.

8. The cable path diagram analysis method according to claim 7, characterized in that: The checking of the target cable path corresponding to the target associated cable path diagram specifically includes: Determine at least one target cable segment from the target cable path corresponding to the target associated cable path diagram, wherein the target cable segment is a portion of the cables in the target cable path where there are factors that are prone to inducing abnormalities; Determine the troubleshooting priority of the corresponding target cable segment according to the weight product, the larger the weight product is, the higher the troubleshooting priority of the corresponding target cable segment is; According to the order of the checking priority from high to low, the corresponding target cable segments in the target cable path are checked in sequence.

9. A cable path diagram analytical model training device, characterized in that: include: An information acquisition module (11) is used to acquire at least one first railway signal cable path diagram, wherein the first railway signal cable path diagram includes different railway signal devices that are connected to each other; A mask determination module (12) is used to capture at least one first screenshot containing railway signal equipment from each of the first railway signal cable path diagrams, and generate a corresponding railway signal equipment position mask diagram based on the first screenshot; A path determination module (13) is used to determine an associated cable path diagram of each railway signal device in the first screenshot in the corresponding first railway signal cable path diagram, wherein the associated cable path diagram is a cable path diagram of a portion of the railway signal device that is most associated with the corresponding railway signal device when an abnormality occurs; A model training module (14) is used to train a preset image segmentation network based on the railway signal equipment position mask map, the associated cable path map and the first railway signal cable path map corresponding to each of the first screenshots to obtain a cable path map analysis model, wherein the cable path map analysis model is a model that outputs an associated cable path map of the corresponding railway signal equipment based on the input railway signal cable path map and the railway signal equipment position mask map.

10. A computer-readable storage medium having a computer program stored therein, characterized in that: When the computer program is loaded and executed by a processor, the method according to any one of claims 1 to 8 is adopted.

Images