Method, device, processor and medium for identifying threat event type above pipeline

Abstract

The application relates to the technical field of pipelines and discloses a pipeline overhead threat event type identification method and device, a processor and a medium. The method comprises the following steps: inputting a waterfall diagram generated according to optical fiber sensing data into a trained neural network model to obtain an identification result, wherein the identification result comprises a normal waterfall diagram, an abnormal waterfall diagram and an abnormal type; when the identification result is the abnormal waterfall diagram, determining a ground position where an anomaly occurs above the pipeline according to the source of the waterfall diagram; acquiring a video of the ground position where the anomaly occurs; detecting the video by using a video detection algorithm to obtain a detection result, wherein the detection result comprises a normal video, an abnormal video and a destructive subject; and when the identification result is the abnormal waterfall diagram, the detection result is the abnormal video, and the abnormal type corresponds to the destructive subject, determining a pipeline overhead threat event type according to the abnormal type and / or the destructive subject. The method can discover the behavior and type of pipeline damage in real time and quickly, and improves the safety of the pipeline.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of pipelines, in particular to a pipeline overhead threat event type identification method and device, a processor and a medium. BACKGROUND

[0002] Taking a gas pipeline and an oil pipeline as examples, the gas pipeline refers to a pipeline mainly used for transporting natural gas, liquefied petroleum gas and artificial gas. In long-distance transportation, the gas pipeline specifically refers to a pipeline for transporting natural gas medium; in a town, the gas pipeline refers to a pipeline for transporting natural gas, liquefied petroleum gas, artificial gas and other media. The oil pipeline (also referred to as a pipeline or a pipeline) is composed of an oil pipeline and its accessories, and is equipped with corresponding oil pump units according to the needs of the process flow, and is designed and installed into a complete pipeline system for completing the tasks of oil loading and unloading and transfer. The safe operation of the pipeline has great social and economic significance. When there is mechanical or manual construction or operation above the pipeline to threaten the safety of the pipeline, if the behavior and type of the destruction event cannot be found in time, it will bring negative effects to the society and economy. The traditional pipeline safety protection mainly installs a camera and uses manual monitoring to identify whether there is a destruction event above the pipeline and the type of the destruction event. However, manual monitoring and identification cannot find the behavior and type of the destruction pipeline in time, resulting in low safety of the pipeline. SUMMARY

[0003] In order to overcome the deficiencies of the prior art, the embodiments of the present application provide a pipeline overhead threat event type identification method, device, processor and medium.

[0004] In order to achieve the above purpose, the first aspect of the present application provides a pipeline overhead threat event type identification method, comprising:

[0005] generating a waterfall chart according to the optical fiber sensing data above the pipeline;

[0006] inputting the waterfall chart into a trained neural network model to obtain an identification result, wherein the identification result includes a normal waterfall chart and an abnormal waterfall chart, and in the case of the identification result being the abnormal waterfall chart, the identification result further includes an abnormal type corresponding to the abnormal waterfall chart;

[0007] In the case of the identification result being the abnormal waterfall chart, determining a ground position where the abnormality occurs above the pipeline according to the source of the waterfall chart;

[0008] obtaining a video of the ground position where the abnormality occurs;

[0009] detecting the video by using a video detection algorithm to obtain a detection result, wherein the detection result includes a normal video and an abnormal video, and in the case of the detection result being the abnormal video, the detection result further includes a destruction subject in the video;

[0010] In a case where the identification result is a waterfall chart anomaly, the detection result is a video anomaly, and the anomaly type corresponds to a destruction subject, the threat event type above the pipeline is determined according to the anomaly type and / or the destruction subject.

[0011] In the embodiment of the present application, the trained neural network model is obtained by the following way:

[0012] The neural network model is trained by using the unknown anomaly type waterfall chart as a sample set, and an initial neural network model is obtained.

[0013] The initial neural network model is trained by using the known anomaly type waterfall chart as a sample set, and a trained neural network model is obtained.

[0014] In the embodiment of the present application, the number of unknown anomaly type waterfall charts is greater than the number of known anomaly type waterfall charts.

[0015] In the embodiment of the present application, the anomaly type includes an abnormal waterfall chart caused by a vibration of an excavator, an abnormal waterfall chart caused by a vibration of a directional drilling machine, an abnormal waterfall chart caused by a vibration of an impact hammer, and an abnormal waterfall chart caused by a vibration of a pile driver.

[0016] In the embodiment of the present application, the video is detected by using a video detection algorithm, and the detection result includes:

[0017] The image in the video is identified frame by frame by using the video detection algorithm, so as to obtain the object in each frame of image.

[0018] The detection result is obtained according to the object.

[0019] In the embodiment of the present application, the detection result obtained according to the object includes:

[0020] In a case where the object does not include a preset destruction subject, the detection result is determined as a normal video, wherein the preset destruction subject includes an excavator, a directional drilling machine, an impact hammer, and a pile driver.

[0021] In a case where the object includes the preset destruction subject, the detection result is determined as a video anomaly.

[0022] In the embodiment of the present application, the fiber sensing data includes vibration data.

[0023] The second aspect of the present application provides a pipeline threat event type identification device, comprising:

[0024] A generation module is configured to generate a waterfall chart according to fiber sensing data above a pipeline.

[0025] The input module is configured to input the waterfall diagram into the trained neural network model to obtain an identification result, wherein the identification result includes waterfall diagram normal and waterfall diagram abnormal, and the identification result further includes an abnormal type corresponding to the waterfall diagram in the case of the identification result being waterfall diagram abnormal.

[0026] The first determination module is configured to determine a ground position where the abnormality occurs above the pipeline according to a source of the waterfall diagram in the case of the identification result being waterfall diagram abnormal.

[0027] The acquisition module is configured to acquire a video of the ground position where the abnormality occurs.

[0028] The detection module is configured to detect the video by using a video detection algorithm to obtain a detection result, wherein the detection result includes video normal and video abnormal, and the detection result further includes a destructive subject in the video in the case of the detection result being video abnormal.

[0029] The second determination module is configured to determine a threat event type above the pipeline according to the abnormal type and / or the destructive subject in the case of the identification result being waterfall diagram abnormal, the detection result being video abnormal, and the abnormal type corresponding to the destructive subject.

[0030] The third aspect of the present application provides a processor configured to execute the above-mentioned identification method of the threat event type above the pipeline.

[0031] The fourth aspect of the present application provides a machine readable storage medium, which stores instructions for causing a machine to execute the above-mentioned identification method of the threat event type above the pipeline.

[0032] In the embodiment of the present application, the identification method of the threat event type above the pipeline includes: generating a waterfall diagram according to fiber sensing data above the pipeline; inputting the waterfall diagram into a trained neural network model to obtain an identification result, wherein the identification result includes waterfall diagram normal and waterfall diagram abnormal, and the identification result further includes an abnormal type corresponding to the waterfall diagram in the case of the identification result being waterfall diagram abnormal; determining a ground position where the abnormality occurs above the pipeline according to a source of the waterfall diagram in the case of the identification result being waterfall diagram abnormal; acquiring a video of the ground position where the abnormality occurs; detecting the video by using a video detection algorithm to obtain a detection result, wherein the detection result includes video normal and video abnormal, and the detection result further includes a destructive subject in the video in the case of the detection result being video abnormal; and determining a threat event type above the pipeline according to the abnormal type and / or the destructive subject in the case of the identification result being waterfall diagram abnormal, the detection result being video abnormal, and the abnormal type corresponding to the destructive subject.

[0033] The optical fiber arranged above the pipeline is used to acquire fiber sensing data, and then a waterfall chart is generated to perceive the environmental condition above the pipeline. The trained neural network model can identify the possible construction position and type above the pipeline according to the waterfall chart. The video detection algorithm is used to detect the video of the abnormal ground position, and the detection result is used as further confirmation, so as to improve the identification accuracy of the pipeline intrusion event. Moreover, the trained neural network model and the video detection algorithm can be used to identify and detect in real time and quickly, so as to discover the behavior and type of damaging the pipeline in time, ensure the safe operation of the pipeline, improve the safety of the pipeline, and save the labor cost. BRIEF DESCRIPTION OF DRAWINGS

[0034] The accompanying drawings are included to provide a further understanding of embodiments of the application, and constitute a part of this specification that is made of the description of the application, and are used to explain the embodiments of the application together with the specific embodiments below, but do not constitute a limitation to the embodiments of the application. In the drawings:

[0035] Figure 1 A flow chart of a method for identifying the type of threat event above the pipeline according to an embodiment of the application is schematically shown;

[0036] Figure 2 An abnormal waterfall chart caused by the excavator according to an embodiment of the application is schematically shown;

[0037] Figure 3 An abnormal waterfall chart of a human damage event according to an embodiment of the application is schematically shown. DETAILED DESCRIPTION

[0038] The specific embodiments of the embodiments of the application are described in detail below in combination with the drawings. It should be understood that the specific embodiments described herein are only used to illustrate and explain the embodiments of the application, and are not used to limit the embodiments of the application.

[0039] It should be noted that if the application embodiments involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement condition, etc. between the components in a certain posture (as shown in the drawings), if the certain posture changes, the directional indications also change accordingly.

[0040] In addition, if the description of "first", "second", etc. is involved in the embodiments of the present application, the description of "first", "second", etc. is only for the purpose of description, and cannot be understood as indicating or implying the relative importance of the indicated technical features or implicitly indicating the number of the indicated technical features. Therefore, the features limited by "first", "second" can be explicitly or implicitly included at least one of the features. In addition, the technical solutions of each embodiment can be combined with each other, but it must be based on the realization of the ordinary skilled in the art, when the combination of technical solutions appears contradictory or cannot be realized, it should be considered that the combination of technical solutions does not exist, nor in the protection scope required by the present application.

[0041] Figure 1 The flow chart of the method for identifying the type of threat event above the pipeline according to the embodiment of the present application is schematically shown. As shown in the figure, Figure 1 In an embodiment of the present application, a method for identifying the type of threat event above the pipeline is provided, comprising the following steps:

[0042] Step 101, generating a waterfall chart according to the optical fiber sensing data above the pipeline;

[0043] Step 102, inputting the waterfall chart into the trained neural network model to obtain the identification result, wherein the identification result includes waterfall chart normal and waterfall chart abnormal, and in the case of waterfall chart abnormal, the identification result further includes the abnormal type corresponding to the waterfall chart;

[0044] Step 103, in the case of waterfall chart abnormal, determining the ground position where the abnormality occurs above the pipeline according to the source of the waterfall chart;

[0045] Step 104, acquiring the video of the ground position where the abnormality occurs;

[0046] Step 105, detecting the video by using the video detection algorithm to obtain the detection result, wherein the detection result includes video normal and video abnormal, and in the case of video abnormal, the detection result further includes the destructive subject in the video;

[0047] Step 106, in the case of waterfall chart abnormal, detection result abnormal, and the abnormal type corresponding to the destructive subject, determining the type of threat event above the pipeline according to the abnormal type and / or the destructive subject.

[0048] In the embodiments of the present application, the neural network model can be a convolutional neural network model, a pre-trained model, etc. The optical fiber sensing data includes vibration data. The construction environment and the quiet environment above the pipeline have differences in optical fiber disturbance, which will cause differences in the corresponding waterfall diagrams generated by the optical fiber sensing data. The trained neural network model can accurately identify such differences. In the process of sending the real-time obtained waterfall diagram into the trained neural network model, each waterfall diagram carries its own information, such as the source of the waterfall diagram, and the source of the waterfall diagram contains the pipeline position information corresponding to the waterfall diagram. Therefore, in the case of identifying the result as an abnormal waterfall diagram, the ground position above the pipeline where the anomaly occurs can be determined according to the source of the waterfall diagram. When confirming the time when the abnormal waterfall diagram is found, the time period is determined according to the time, and then the video of the ground position where the anomaly occurs in the time period is obtained.

[0049] The optical fiber arranged above the pipeline is used to obtain optical fiber sensing data, and then a waterfall diagram is generated to perceive the environmental conditions above the pipeline. The trained neural network model can identify the possible construction position and type above the pipeline according to the waterfall diagram. The video detection algorithm is used to detect the video of the ground position where the anomaly occurs. The detection result is used as further confirmation, which improves the identification accuracy of the pipeline intrusion event. Moreover, the trained neural network model and the video detection algorithm can be identified and detected in real time and quickly, the behavior and type of damaging the pipeline are found in time, the safe operation of the pipeline is ensured, the safety of the pipeline is improved, and the labor cost is saved.

[0050] In an embodiment, the trained neural network model is obtained by the following method: an abnormal waterfall diagram with an unknown abnormal type is used as a sample set to train a neural network model to obtain an initial neural network model; an abnormal waterfall diagram with a known abnormal type is used as a sample set to train the initial neural network model to obtain the trained neural network model.

[0051] In an embodiment, the neural network model can be a pre-trained model, and the pre-trained model is used to realize image recognition of the waterfall chart. A large number of abnormal fiber signals of past unknown threat events can be used to obtain a large number of abnormal fiber waterfall charts (i.e., abnormal waterfall charts of unknown abnormal types), and the excellent feature extraction capability of the pre-trained model can fully utilize the information of the part of the unlabeled data. The unlabeled data is understood as the abnormal waterfall chart of the unknown abnormal type, or is understood as the abnormal waterfall chart of the unknown threat event type above the pipeline. Then, only a relatively small number of abnormal fiber signal waterfall charts of known threat event types (i.e., abnormal waterfall charts of known abnormal types) are needed, so that the model can quickly acquire the recognition capability of the abnormal type. The number of the abnormal waterfall charts of the unknown abnormal type is greater than the number of the abnormal waterfall charts of the known abnormal type. This training method can reduce sample data, i.e., reduce training data, and reduce the training difficulty, and can make accurate prediction for the rare threat event above the pipeline, and the precision can be as high as 95% or more.

[0052] In an embodiment, a convolutional neural network (CNN) model can be used to identify the threat event category (i.e., the abnormal type) of the waterfall chart. The convolutional neural network model can be trained directly through a large number of samples of known threat event categories (abnormal waterfall charts of known abnormal types) to form a feature library. For the identification of the relatively common threat event, the precision can be as high as 95% or more.

[0053] In the embodiment of the present application, after the trained neural network model discovers the abnormality of the waterfall chart, a video detection algorithm can be used to automatically perform video recognition on the point video where the abnormality occurs, to determine the damage subject such as the excavator and the pile driver that may cause the abnormality of the fiber signal, and to further reduce the dependence on manual work. In an embodiment, the abnormal types include the abnormal waterfall chart caused by the vibration of the excavator, the abnormal waterfall chart caused by the vibration of the directional drilling machine, the abnormal waterfall chart caused by the vibration of the impact hammer, and the abnormal waterfall chart caused by the vibration of the pile driver. The damage subjects include the excavator, the directional drilling machine, the impact hammer, and the pile driver. The damage event above the pipeline can be referred to as a pipeline intrusion event, a threat event above the pipeline, a non-normal event on the ground above the pipeline, etc. When the excavator is digging and the directional drilling machine is drilling above the pipeline, the safety of the pipeline can be threatened. The identification method provided by the embodiment of the present application can be used to detect the behavior and type that can damage the pipeline in real time.

[0054] The main advantages of optical fiber are low loss, long relay distance, anti-electromagnetic interference, no crosstalk interference, good security, and optical fiber is sensitive to vibration. The optical fiber will have different waterfall graph performances under different environmental vibrations. The construction environment above the pipeline and the quiet environment have differences in disturbing the optical fiber. Generally, the ground above the pipeline is in the construction state, and a vertical line appears on the waterfall graph for a period of time. Similarly, the construction environment above the pipeline, the vibration caused by the excavator, the vibration caused by the directional drill, the vibration caused by the impact hammer, and the vibration caused by the pile driver are also different. The characteristics of the waterfall graphs of these different abnormal types are also different. The excavator, directional drill, impact hammer, and pile driver have different effects on the optical fiber when they are used for construction on the ground above the pipeline. Therefore, the differences in the waterfall graphs can be used to determine the type of non-normal event on the ground. That is, the trained neural network model can not only obtain the recognition result of the waterfall graph anomaly, but also obtain the abnormal type corresponding to the waterfall graph. Figure 2 An abnormal waterfall graph caused by vibration of an excavator according to an embodiment of the present application is schematically shown, which can be seen from Figure 2 In the waterfall graph, the image signal is strong and the duration is long.

[0055] It should be noted that there are differences between human excavation (i.e., human damage) above the pipeline and mechanical damage (excavator, directional drill, impact hammer, pile driver, etc.). Correspondingly, the waterfall graphs generated according to the optical fiber sensing data above the pipeline are also different. The trained neural network model can also identify human damage events according to the waterfall graph. Figure 3 An abnormal waterfall graph of a human damage event according to an embodiment of the present application is schematically shown.

[0056] In an embodiment, a video is detected by using a video detection algorithm to obtain a detection result, including: identifying images in the video frame by frame by using the video detection algorithm to obtain objects in each frame of image; and obtaining the detection result according to the objects. In an embodiment, obtaining the detection result according to the objects includes: in a case where the objects do not include a preset damage subject, determining the detection result as normal video, wherein the preset damage subject includes an excavator, a directional drill, an impact hammer, and a pile driver; and in a case where the objects include the preset damage subject, determining the detection result as abnormal video.

[0057] Exemplarily, if the recognition result of the trained neural network model is a waterfall chart anomaly, and the anomaly type is an abnormal waterfall chart caused by the vibration of the excavator, then, the detection result of the video detection algorithm is a video anomaly, and the object in the video includes an excavator, it is determined that the threat event type above the pipeline is that the excavator is working above the pipeline and threatens the safety of the pipeline. The waterfall chart and the video monitoring information are combined to identify the threat event (or referred to as the destruction event) above the pipeline. Based on the waterfall chart, the video detection algorithm is used to make a joint decision based on the video monitoring information of the corresponding position to locate the alarm cause.

[0058] Exemplarily, if the recognition result of the trained neural network model is a waterfall chart anomaly, and the anomaly type is an abnormal waterfall chart caused by the vibration of the pile driver, but in the detection of the video detection algorithm, the object of the image in the video is only a motorcycle (generally, a motorcycle does not threaten the safety of the pipeline, so the motorcycle is not a preset destruction subject), and no preset destruction subject is identified, at this time, it cannot be determined that the threat event type above the pipeline is that the pile driver is working above the pipeline, but a warning can be given.

[0059] Exemplarily, if the recognition result of the trained neural network model is a waterfall chart anomaly, and the anomaly type is an abnormal waterfall chart caused by the vibration of the directional drilling machine, but the video detection algorithm detects that the object in the video includes an excavator but does not detect the directional drilling machine, at this time, the threat event type above the pipeline cannot be determined, but a warning can be given, and then other determination methods are combined for confirmation.

[0060] In the embodiment of the present application, the optical fiber sensing technology is applied, the continuous sensing intensity data is converted into a waterfall chart image by processing the optical fiber sensing data, and then the threat event identification is further carried out by using image recognition. The existing optical cable in the pipeline is used as a sensor to collect the data of the optical fiber in real time, generate a waterfall chart to perceive the environmental status above the pipeline, identify the position and type of the possible construction by using the waterfall chart algorithm (neural network model), and based on the video detection algorithm, the video monitoring information of the corresponding position is used as confirmation to quickly and accurately identify the pipeline intrusion event. It can be widely applied to vibration detection of pipelines, cables and other pipelines containing underground optical fibers, or the optical fiber can be directly used as a vibration detection sensor and applied to any early warning of abnormal conditions such as vibration and destruction.

[0061] The embodiment of the present application provides a pipeline threat event type identification device, which comprises:

[0062] A generation module is configured to generate a waterfall chart according to the optical fiber sensing data above the pipeline.

[0063] The input module is configured to input the waterfall diagram into the trained neural network model to obtain an identification result, wherein the identification result includes waterfall diagram normal and waterfall diagram abnormal, and the identification result further includes an abnormal type corresponding to the waterfall diagram in the case of the identification result being waterfall diagram abnormal.

[0064] The first determination module is configured to determine a ground position where the abnormality occurs above the pipeline according to a source of the waterfall diagram in the case of the identification result being waterfall diagram abnormal.

[0065] The acquisition module is configured to acquire a video of the ground position where the abnormality occurs.

[0066] The detection module is configured to detect the video by using a video detection algorithm to obtain a detection result, wherein the detection result includes video normal and video abnormal, and the detection result further includes a destructive subject in the video in the case of the detection result being video abnormal.

[0067] The second determination module is configured to determine a threat event type above the pipeline according to the abnormal type and / or the destructive subject in the case of the identification result being waterfall diagram abnormal, the detection result being video abnormal, and the abnormal type corresponding to the destructive subject.

[0068] The identification device for the threat event type above the pipeline provided by the embodiment of the application can realize Figure 1 The method for identifying the threat event type above the pipeline in the method embodiment is not repeated here to avoid repetition.

[0069] The embodiment of the application provides a processor configured to execute the above-mentioned method for identifying the threat event type above the pipeline.

[0070] Specifically, the processor can be configured to:

[0071] generate a waterfall diagram according to optical fiber sensing data above the pipeline;

[0072] input the waterfall diagram into the trained neural network model to obtain an identification result, wherein the identification result includes waterfall diagram normal and waterfall diagram abnormal, and the identification result further includes an abnormal type corresponding to the waterfall diagram in the case of the identification result being waterfall diagram abnormal;

[0073] determine a ground position where the abnormality occurs above the pipeline according to a source of the waterfall diagram in the case of the identification result being waterfall diagram abnormal;

[0074] acquire a video of the ground position where the abnormality occurs;

[0075] detect the video by using a video detection algorithm to obtain a detection result, wherein the detection result includes video normal and video abnormal, and the detection result further includes a destructive subject in the video in the case of the detection result being video abnormal.

[0076] In a case where the identification result is a waterfall chart anomaly, the detection result is a video anomaly, and the anomaly type corresponds to a destruction subject, a threat event type above the pipeline is determined according to the anomaly type and / or the destruction subject.

[0077] In the embodiment of the present application, the processor is configured to:

[0078] The trained neural network model is obtained by the following way:

[0079] An abnormal waterfall chart of an unknown anomaly type is taken as a sample set to train the neural network model, and an initial neural network model is obtained.

[0080] An abnormal waterfall chart of a known anomaly type is taken as a sample set to train the initial neural network model, and the trained neural network model is obtained.

[0081] In the embodiment of the present application, the processor is configured to:

[0082] The number of abnormal waterfall charts of unknown anomaly types is greater than the number of abnormal waterfall charts of known anomaly types.

[0083] In the embodiment of the present application, the processor is configured to:

[0084] The anomaly types include an abnormal waterfall chart caused by vibration of an excavator, an abnormal waterfall chart caused by vibration of a directional drill, an abnormal waterfall chart caused by vibration of a percussion hammer, and an abnormal waterfall chart caused by vibration of a pile driver.

[0085] In the embodiment of the present application, the processor is configured to:

[0086] The video is detected by using a video detection algorithm, and the detection result includes:

[0087] The images in the video are identified frame by frame by using the video detection algorithm, so as to obtain objects in each frame of image.

[0088] The detection result is obtained according to the objects.

[0089] In the embodiment of the present application, the processor is configured to:

[0090] The detection result is obtained according to the objects, including:

[0091] In a case where the objects do not include a preset destruction subject, the detection result is determined as a normal video, wherein the preset destruction subject includes an excavator, a directional drill, a percussion hammer, and a pile driver.

[0092] In a case where the objects include the preset destruction subject, the detection result is determined as a video anomaly.

[0093] In the embodiment of the present application, the processor is configured to:

[0094] The fiber optic sensing data includes vibration data.

[0095] The embodiment of the present application provides a machine readable storage medium, which stores instructions for causing a machine to execute the above-mentioned method for identifying a threat event type above a pipeline.

[0096] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) containing computer usable program code.

[0097] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus produce a device implemented in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 means for performing the function specified by one or more blocks.

[0098] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer readable memory produce a manufactured product including instruction means, which implement the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 one or more blocks.

[0099] These computer program instructions can also be loaded into a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable data processing apparatus to produce a computer implemented process, so that the instructions executed on the computer or other programmable data processing apparatus provide a process for implementing the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 one or more blocks.

[0100] In one typical arrangement, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0101] Memory can include non-persistent memory and / or volatile memory, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM), electrically programmable read only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), flash memory, or other memory technologies, including, but not limited to, those that are called static RAM (SRAM), dynamic RAM (DRAM), or variants of the foregoing. Memory is an example of computer readable media.

[0102] Computer readable media includes permanent and non-permanent, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD), or other optical storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible to a computing device. According to the definition herein, computer readable media does not include transitory media, such as modulated data signals and carrier waves.

[0103] It should also be noted that the terms "comprising," "including," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements recited, but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without limitation, an element preceded by "comprises a" does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0104] The above merely provides an example of the present application and is not intended to limit the present application. The present application can have various modifications and changes for those skilled in the art. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the scope of claims of the present application.

Claims

1. A method for identifying the type of threat event above a pipeline, characterized in that: include: Generate waterfall charts based on fiber optic sensing data above the pipeline; Inputting the waterfall chart into a trained neural network model to obtain a recognition result, wherein the recognition result includes whether the waterfall chart is normal or abnormal, and if the recognition result is that the waterfall chart is abnormal, the recognition result also includes the abnormality type corresponding to the waterfall chart; In the case where the identification result is that the waterfall diagram is abnormal, determining the ground position where the abnormality occurs above the pipeline based on the source of the waterfall diagram; Acquire a video of the ground location where the anomaly occurs; Detecting the video using a video detection algorithm to obtain a detection result, wherein the detection result includes whether the video is normal or abnormal, and if the detection result is abnormal, the detection result also includes the destructive subject in the video; When the recognition result is a waterfall chart anomaly, the detection result is a video anomaly, and the anomaly type corresponds to the destructive subject, determining the type of threat event above the pipeline according to the anomaly type and / or the destructive subject; The neural network model is a pre-trained model, and the trained neural network model is obtained in the following way: The abnormal waterfall chart of unknown abnormal types is used as a sample set to train the neural network model to obtain an initial neural network model; The initial neural network model is trained using the abnormal waterfall chart of known abnormal types as a sample set to obtain a trained neural network model.

2. The method according to claim 1, characterized in that The number of the anomaly waterfall charts of the unknown anomaly type is greater than the number of the anomaly waterfall charts of the known anomaly type.

3. The method according to claim 1, characterized in that The abnormal types include abnormal waterfall diagrams of vibration caused by an excavator, abnormal waterfall diagrams of vibration caused by a directional drill, abnormal waterfall diagrams of vibration caused by an impact hammer, and abnormal waterfall diagrams of vibration caused by a pile driver.

4. The method according to claim 1, wherein The video is detected by using a video detection algorithm to obtain a detection result including: Using a video detection algorithm to identify images in the video frame by frame to obtain objects in the image of each frame; A detection result is obtained according to the object.

5. The method according to claim 4, characterized in that Obtaining a detection result according to the object includes: In a case where the object does not include a preset destructive subject, determining the detection result as a normal video, wherein the preset destructive subject includes an excavator, a directional drill, an impact hammer, and a pile driver; In a case where the object includes the preset destructive subject, the detection result is determined as a video abnormality.

6. The method according to claim 1, characterized in that The optical fiber sensing data includes vibration data.

7. A device for identifying the type of threat event above a pipeline, characterized in that: include: A generation module, for generating a waterfall chart based on optical fiber sensing data above the pipeline; An input module, configured to input the waterfall chart into a trained neural network model to obtain a recognition result, wherein the recognition result includes whether the waterfall chart is normal or abnormal, and if the recognition result is abnormal, the recognition result also includes the abnormality type corresponding to the waterfall chart; A first determining module is configured to determine, when the identification result is that the waterfall diagram is abnormal, a ground position above the pipeline where the abnormality occurs based on the source of the waterfall diagram; An acquisition module, configured to acquire a video of the ground location where the anomaly occurs; a detection module, configured to detect the video using a video detection algorithm to obtain a detection result, wherein the detection result includes whether the video is normal or abnormal, and if the detection result is abnormal, the detection result also includes a destructive subject in the video; a second determining module, for determining a type of threat event above the pipeline according to the abnormality type and / or the damaging subject, when the recognition result is a waterfall chart abnormality, the detection result is a video abnormality, and the abnormality type corresponds to the damaging subject; The neural network model is a pre-trained model, and the trained neural network model is obtained in the following way: The abnormal waterfall chart of unknown abnormal types is used as a sample set to train the neural network model to obtain an initial neural network model; The initial neural network model is trained using the abnormal waterfall chart of known abnormal types as a sample set to obtain a trained neural network model.

8. A processor, characterized in that: The device is configured to execute the method for identifying the type of threat event above a pipeline according to any one of claims 1 to 6.

9. A machine-readable storage medium having instructions stored thereon, characterized in that: The instruction is used to enable the machine to execute the method for identifying the type of threat event above the pipeline according to any one of claims 1 to 6.

Images