Intelligent detection method and system for abnormal behavior in power scenarios

Abstract

The present disclosure provides an intelligent detection method and system for abnormal behavior in power scenarios, which relates to the field of abnormal detection technology. The method includes: constructing a target digital twin model; analyzing a first image set of a first power device among M power devices collected by a CCD image sensor to obtain a first 3D model; rendering the first 3D model to a target digital twin model; sensing the distance from the object to be inspected to the first power device in the target digital twin model to obtain a first spatial distance; and activating an acoustic and visual alarm device to issue an abnormal alarm when the first spatial distance does not reach a first predetermined distance threshold. The present disclosure solves the problem that the existing detection accuracy of abnormal intrusion behaviors of equipment or personnel is low, resulting in intrusion dangers in power scenarios and even threats to life and property safety, and achieves the goal of improving the detection accuracy of abnormal intrusion behaviors in power scenarios, and achieves the technical effect of reducing the incidence of abnormal intrusion danger events in power scenarios.

Description

Technical Field

[0001] The present disclosure relates to the field of abnormality detection technology, and in particular to an intelligent detection method and system for abnormal behavior in power scenarios. Background Art

[0002] With the rapid development of my country's economy and the large-scale construction of power grids, the scale of equipment in my country's power scenarios has continued to increase, and the power system has been greatly improved. However, there are always potential safety hazards in power scenarios, and safety control tasks are the basic issues in power scenarios.

[0003] In summary, the existing technology has the technical problem that the detection accuracy of abnormal intrusion behavior of equipment or personnel is low, which leads to intrusion hazards in power scenarios and even threatens the safety of life and property. Summary of the invention

[0004] The present disclosure provides an intelligent detection method and system for abnormal behavior in power scenarios, which are used to solve the technical problem in the prior art that the detection accuracy of abnormal intrusion behavior of equipment or personnel is low, resulting in intrusion hazards in power scenarios and even threats to life and property safety.

[0005] According to a first aspect of the present disclosure, there is provided an intelligent detection method for abnormal behavior in an electric power scenario, comprising: constructing a target digital twin model based on the multi-dimensional characteristics of the target electric power scenario, the target electric power scenario comprising M electric power equipment, where M is an integer greater than or equal to 1; analyzing a first image set of a first electric power equipment among the M electric power equipment acquired by the CCD image sensor through a device registration fusion model to obtain a first 3D model; rendering the first 3D model to the target digital twin model, wherein the first 3D model has an identification of a first restricted area determined based on the multi-dimensional characteristic analysis of the first electric power equipment; sensing the distance from the object to be inspected in the target digital twin model to the first electric power equipment based on the inscribed sphere principle to obtain a first spatial distance; and starting the sound and light alarm device to issue an abnormal alarm when the first spatial distance does not reach a first predetermined distance threshold, wherein the predetermined distance threshold is a distance range determined based on the analysis of the first restricted area.

[0006] According to a second aspect of the present disclosure, an intelligent detection system for abnormal behavior in an electric power scene is provided, including: a target digital twin model acquisition module, the target digital twin model acquisition module is used to construct a target digital twin model based on the multi-dimensional characteristics of the target electric power scene, the target electric power scene includes M electric power devices, M is an integer greater than or equal to 1; a first 3D model acquisition module, the first 3D model acquisition module is used to analyze a first image set of a first electric power device among the M electric power devices collected by the CCD image sensor through a device registration fusion model to obtain a first 3D model; a first restricted area identification acquisition module, the first restricted area identification acquisition module is used to The first 3D model is rendered to the target digital twin model, wherein the first 3D model has an identification of a first restricted area determined based on a multi-dimensional feature analysis of the first power equipment; a first spatial distance acquisition module, the first spatial distance acquisition module is used to sense the distance from the object to be inspected in the target digital twin model to the first power equipment based on the inscribed sphere principle, and obtain a first spatial distance; a predetermined distance threshold acquisition module, the predetermined distance threshold acquisition module is used to activate the sound and light alarm device to issue an abnormal alarm when the first spatial distance does not reach a first predetermined distance threshold, and the predetermined distance threshold is a distance range determined based on the analysis of the first restricted area.

[0007] According to a third aspect of the present disclosure, a computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor is capable of executing any one of the methods according to the first aspect.

[0008] According to a fourth aspect of the present disclosure, a computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in any one of the first aspects can be implemented.

[0009] One or more technical solutions provided in the present disclosure have at least the following technical effects or advantages: according to the method adopted in the present disclosure, a target digital twin model is constructed based on the multi-dimensional features of a target power scene, wherein the target power scene includes M power devices, where M is an integer greater than or equal to 1; a first image set of a first power device among the M power devices collected by the CCD image sensor is analyzed through a device registration fusion model to obtain a first 3D model; the first 3D model is rendered to the target digital twin model, wherein the first 3D model has an identification of a first restricted area determined based on the multi-dimensional feature analysis of the first power device; based on The inscribed sphere principle is used to sense the distance from the object to be inspected in the target digital twin model to the first power equipment to obtain a first spatial distance; when the first spatial distance does not reach a first predetermined distance threshold, the sound and light alarm device is activated to issue an abnormal alarm, and the predetermined distance threshold is a distance range determined based on the analysis of the first restricted area. This solves the technical problem in the prior art that the accuracy of detecting abnormal intrusions of equipment or personnel is low, resulting in intrusion hazards in power scenarios and even threats to life and property safety, thereby achieving the goal of improving the accuracy of detecting abnormal intrusions in power scenarios and achieving the technical effect of reducing the incidence of abnormal intrusion hazards in power scenarios.

[0010] It should be understood that the contents described in this section are not intended to indicate the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the present disclosure or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0012] Figure 1 A flow chart of an intelligent detection method for abnormal behavior in power scenarios provided by an embodiment of the present disclosure;

[0013] Figure 2 A logical schematic diagram of system equipment in an intelligent detection method for abnormal behavior in an electric power scenario according to an embodiment of the present disclosure;

[0014] Figure 3 A schematic diagram of the structure of an intelligent detection system for abnormal behavior in power scenarios provided by an embodiment of the present disclosure;

[0015] Figure 4 A schematic diagram of the structure of a computer device provided in an embodiment of the present disclosure.

[0016] Explanation of the accompanying drawings: target digital twin model acquisition module 11, first 3D model acquisition module 12, first restricted area identification acquisition module 13, first spatial distance acquisition module 14, predetermined distance threshold acquisition module 15, electronic device 600, processor 601, memory 602, bus 603. DETAILED DESCRIPTION

[0017] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0018] Embodiment 1

[0019] The present disclosure provides an intelligent detection method for abnormal behavior in power scenarios, which is referred to Figure 1 and Figure 2 Illustratively, the method comprises:

[0020] The method provided in the embodiment of the present disclosure includes:

[0021] Building a target digital twin model based on multi-dimensional features of a target power scenario, wherein the target power scenario includes M power devices, where M is an integer greater than or equal to 1;

[0022] The target power scene is a scene to be detected for abnormal behavior, such as a substation. The target power scene includes M power equipment. Since the target power scene includes at least one power equipment, M is an integer greater than or equal to 1. Further, the multi-dimensional features of the target power scene are obtained, wherein the multi-dimensional features include the scene area structure size and the equipment in the scene.

[0023] Analyzing a first image set of a first power device among the M power devices collected by the CCD image sensor through a device registration fusion model to obtain a first 3D model;

[0024] A CCD image sensor is a semiconductor device used to convert light into electric charge, which is converted into a digital signal through an analog-to-digital converter chip. Further, a first image set is obtained by performing image acquisition on a first power device randomly extracted from M power devices through a CCD image sensor. Further, point cloud data of the first image set is obtained. The first image set contains each image, and the point cloud data of each image is combined into a point cloud data set of each image in the first image set. The point cloud data set of each image in the first image set is extracted as input information and input into a device registration fusion model, and the model output result is obtained by processing the device registration fusion model. Among them, the intelligent model is constructed according to the principle of random sampling consistency algorithm to obtain a device registration fusion model. Further, a point cloud model is constructed for the model parameter estimation result in the model output result to obtain a first point cloud model. Texture mapping is performed on the first point cloud model through a CCD image sensor to obtain a texture mapping result. Among them, the texture mapping is a local feature of the point cloud model. The texture mapping result is used as the first 3D model of the first power device.

[0025] Rendering the first 3D model to the target digital twin model, wherein the first 3D model has an identification of a first restricted area determined based on a multi-dimensional feature analysis of the first power equipment;

[0026] The first 3D model is input into the target digital twin model, and the target digital twin model is used for rendering to obtain a rendering result. Further, the first 3D model is analyzed according to the multi-dimensional characteristics of the first power equipment to determine the identification of the first restricted area. The multi-dimensional characteristics represent the structural dimensions of the scene area and the equipment in the scene. The identification of the first restricted area is a location attribute identification.

[0027] Based on the inscribed sphere principle, the distance from the object to be inspected in the target digital twin model to the first power equipment is sensed to obtain a first spatial distance;

[0028] When the first spatial distance does not reach a first predetermined distance threshold, the sound and light alarm device is activated to issue an abnormal alarm, and the predetermined distance threshold is a distance range determined based on the analysis of the first restricted area.

[0029] The inscribed sphere principle is to inscribe a sphere in an irregular object, and to obtain the distance by calculating the radius of the sphere. According to the inscribed sphere principle, the distance from the object to be inspected in the target digital twin model to the first power equipment is sensed to obtain the first spatial distance. The object to be inspected is the part to be inspected.

[0030] Further, when the first spatial distance does not reach a first predetermined distance threshold, the sound and light alarm device is activated to issue an abnormal alarm, and the predetermined distance threshold is a distance range determined based on the analysis of the first restricted area.

[0031] Among them, this embodiment can solve the technical problem in the prior art that the detection accuracy of abnormal intrusion behavior of equipment or personnel is low, which leads to intrusion hazards in power scenarios and even threatens the safety of life and property, and achieves the goal of improving the detection accuracy of abnormal intrusion behavior in power scenarios, and achieves the technical effect of reducing the occurrence rate of abnormal intrusion hazards in power scenarios.

[0032] The method provided in the embodiment of the present disclosure also includes:

[0033] The CCD image sensor acquires images of the first power device based on a preset frequency to obtain the first image set;

[0034] Using the point cloud data set of each image in the first image set as input information of the device registration fusion model, and obtaining a model output result, wherein the device registration fusion model is an intelligent model constructed based on the principle of random sampling consistency algorithm;

[0035] Performing texture mapping on a first point cloud model constructed according to a model parameter estimation result in the model output result based on sensor characteristic parameters of the CCD image sensor to obtain a texture mapping result;

[0036] The texture mapping result is used as the first 3D model of the first power device.

[0037] The CCD image sensor is a semiconductor device that converts light into electric charge and converts it into a digital signal through an analog-to-digital converter chip. Furthermore, the higher the driving frequency supported by the CCD image sensor, the faster the single-frame scanning speed and the higher the system requirements. Therefore, the frequency of the CCD image sensor is customized according to the actual situation. Such as 500KHz, 1MHz, 2Mhz, 5MHz, 20MHz, 40MHz, 100MHz, etc. Further, the first power device randomly extracted from the M power devices is imaged by the CCD image sensor to obtain a first image set.

[0038] Furthermore, by directly sensing the geometric structure of each image in the first image set through the laser of the laser radar, point cloud data with high precision and high density can be obtained. The first image set contains each image, and the point cloud data of each image is combined into a point cloud data set of each image in the first image set. The point cloud data set of each image in the first image set is extracted as input information and input into the device registration fusion model, and the model output result is obtained by processing through the device registration fusion model. Among them, the intelligent model is constructed according to the principle of random sampling consistency algorithm to obtain the device registration fusion model. The principle of random sampling consistency algorithm is an iterative method for estimating the parameters of a mathematical model from a set of observation data containing outliers, and the outliers have no effect on the estimated values. Therefore, the principle of random sampling consistency algorithm is also an outlier detection method. It is a non-deterministic algorithm that generates reasonable results with a certain probability, and the probability increases with the increase of the number of iterations. For example, the principle of random sampling consistency algorithm is used to solve the problem of position determination, in which the goal is to determine points in space, and the points in space are projected onto the image to form a set of landmarks with known positions.

[0039] Further, a point cloud model is constructed based on the model parameter estimation results in the model output results to obtain a first point cloud model. Texture mapping is performed on the first point cloud model through the sensor characteristic parameters of the CCD image sensor to obtain a texture mapping result. Among them, the sensor characteristic parameters are the resolution of the CCD image sensor. The texture mapping is a local feature of the point cloud model. The texture mapping result is used as the first 3D model of the first power device.

[0040] Among them, the three-dimensional model of the First Electric Power Equipment using laser point cloud full-view scanning technology can improve the authenticity and reliability of data information, provide a target close to the real environment for three-dimensional visualization management, and improve detection efficiency.

[0041] The method provided in the embodiment of the present disclosure also includes:

[0042] The device registration fusion model obtains a first parameter estimation result of a first model according to a first sample set randomly sampled from the point cloud data set;

[0043] Screening and forming a first consistent point set, the first consistent point set refers to a set of non-samples whose distances from each non-sample point cloud data in a first non-sample set to the first model meet a second preset distance threshold, and the first non-sample set refers to a sample set obtained by excluding the first sample set from the point cloud data set;

[0044] When the first data volume of the first consistency point set meets a preset quantity threshold, a second parameter estimation result of the first model is obtained based on the first consistency point set, where the second parameter estimation result is a parameter estimation result obtained by verification re-estimation;

[0045] The first parameter estimation result is replaced by the second parameter estimation result as the model output result.

[0046] A first sample set is obtained by randomly sampling from the point cloud data set of each image in the first image set, and the first sample set is input into the device registration model, and the first model and its first parameter estimation result are obtained by training the device registration model.

[0047] Further, the point cloud data set of each image in the first image set includes a first non-sample set. The first non-sample set refers to a sample set obtained after removing the first sample set from the point cloud data set of each image in the first image set. Among them, the first non-sample set contains various non-sample point cloud data. Further, according to the actual situation, a second preset distance threshold is customized and obtained to judge the distance from each non-sample point cloud data in the first non-sample set to the first model, and then judge the reliability of the first model. Further, the distance from each non-sample point cloud data in the first non-sample set to the first model is obtained, and it is judged whether the distance from each non-sample point cloud data in the first non-sample set to the first model meets the second preset distance threshold, and the distance judgment result is obtained. If the distance judgment result is that the distance from each non-sample point cloud data in the first non-sample set to the first model meets the second preset distance threshold, it means that the first model is reliable and valid, then all non-sample point cloud data in the first non-sample set that meet the second preset distance threshold are obtained to form a first consistency point set.

[0048] Furthermore, the preset quantity threshold is used to verify the first model and remove the randomness of the values ​​of each non-sample point cloud data by judging whether the distance from most of the non-sample point cloud data in each non-sample point cloud data of the first non-sample set to the first model meets the second preset distance threshold. Among them, the judgment of most of the quantity values ​​is customized according to the actual situation. Further, when the first data volume of the first consistency point set meets the preset quantity threshold, the first consistency point set is input into the first model for verification and re-estimation to obtain the second parameter estimation result. The second parameter estimation result replaces the first parameter estimation result as the model output result.

[0049] The model output result can be used to construct the first 3D model, making the power detection more refined.

[0050] The method provided in the embodiment of the present disclosure also includes:

[0051] When the first data amount of the first consistency point set does not meet the preset quantity threshold, the device registration fusion model performs secondary random sampling from the point cloud data set.

[0052] It is determined that the distance from most of the non-sample point cloud data in each non-sample point cloud data of the first non-sample set to the first model does not meet the second preset distance threshold, or when the first data volume of the first consistency point set does not meet the preset quantity threshold, it indicates that the first data volume of the first consistency point set is insufficient to represent all point cloud data, or that the model is less reliable and effective. Further, based on the above reasons, a group of samples is re-extracted, and a second sample set is obtained by randomly sampling from the point cloud data sets of each image in the first image set. The second sample set is input into the device registration model, and the second model is obtained by training the device registration model and tested.

[0053] Among them, the device registration fusion model performs secondary random sampling from the point cloud dataset to improve the comprehensiveness of data sampling.

[0054] The method provided in the embodiment of the present disclosure also includes:

[0055] A first three-dimensional fence of the first electric equipment is set based on the first restricted area, wherein the first three-dimensional fence includes N guarding posts, where N=12, and the N guarding posts form a rectangular parallelepiped;

[0056] Deploy the infrared emitter in the infrared detector to the first enclosure column among the N enclosure columns, wherein the first enclosure column is any column located on the ground;

[0057] Deploy the infrared receiver in the infrared detector to the second enclosure column among the N enclosure columns, wherein the second enclosure column is parallel to the first enclosure column, and the plane where the second enclosure column and the first enclosure column are located is perpendicular to the ground;

[0058] generating a signal deviation curve according to the infrared signal transmitted by the infrared transmitter and the infrared signal received by the infrared receiver;

[0059] Whether the sound and light alarm device is activated is judged based on the curve characteristics of the signal deviation curve.

[0060] A three-dimensional fence of the first power equipment is set according to the first restricted area of ​​the first 3D model corresponding to the first power equipment as a first three-dimensional fence. The first three-dimensional fence includes N guarding columns, N=12, and the N guarding columns form a rectangular parallelepiped.

[0061] Furthermore, the infrared detector has an infrared emitter. The infrared emitter in the infrared detector is deployed to the first enclosure column among the N enclosure columns, wherein a enclosure column is randomly selected from the N enclosure columns as the first enclosure column. The first enclosure column is any column located on the ground.

[0062] Furthermore, the infrared detector has an infrared receiver. The infrared receiver in the infrared detector is deployed to the second enclosure column among the N enclosure columns. The first enclosure column is removed from the N enclosure columns and a enclosure column is randomly selected as the second enclosure column. The second enclosure column is a column parallel to the first enclosure column, and the plane where the second enclosure column and the first enclosure column are located is perpendicular to the ground, that is, the first enclosure column, the second enclosure column and the ground are in a perpendicular relationship.

[0063] Furthermore, the infrared signal transmitted by the infrared transmitter is compared with the infrared signal received by the infrared receiver to generate signal deviation data, and the signal deviation data is curve-fitted to generate a signal deviation curve. The first restricted area is analyzed for intrusion objects based on the difference between the transmitted infrared signal and the received infrared signal. For example, when a person passes through the first restricted area, the corresponding signal of the received infrared signal is affected.

[0064] Furthermore, the sound and light alarm device emits two alarm signals, sound and light, to give an alarm. Whether the sound and light alarm device is activated is judged according to the curve characteristics of the signal deviation curve. The curve characteristics of the signal deviation curve include characteristics such as the curve slope.

[0065] The system is connected in communication with an infrared detector and can be used to generate an alarm to warn of power anomalies.

[0066] The method provided in the embodiment of the present disclosure also includes:

[0067] Retrieving the historical power equipment infringement records in the cloud computer;

[0068] Extracting a first historical limit violation record from the historical power equipment limit violation record, wherein the first historical limit violation record includes a first historical signal deviation;

[0069] Setting a predetermined signal deviation threshold according to the first historical signal deviation;

[0070] When the maximum signal deviation in the signal deviation curve exceeds the predetermined signal deviation threshold, the sound and light alarm device is activated to issue an abnormal alarm.

[0071] The cloud computer is a data storage terminal. Further, the historical power equipment limit violation records in the cloud computer are retrieved. Further, a group of historical limit violation records are randomly extracted from the historical power equipment limit violation records as the first historical limit violation records. Further, the first historical limit violation record includes a first historical signal deviation. The first historical signal deviation is signal deviation data generated by the infrared detector.

[0072] Further, the first historical signal deviation is set as a predetermined signal deviation threshold. Further, the maximum signal deviation in the signal deviation curve is obtained according to the signal deviation curve, wherein the highest peak value of the signal deviation curve is taken as the maximum signal deviation. Further, when the maximum signal deviation in the signal deviation curve exceeds the predetermined signal deviation threshold, it indicates that the first power equipment is detected to be abnormal, and the sound and light alarm device is activated to issue an abnormal alarm.

[0073] Among them, judging whether the sound and light alarm device is activated or not based on the curve characteristics of the signal deviation curve can make power detection more refined.

[0074] The method provided in the embodiment of the present disclosure also includes:

[0075] When the maximum signal deviation in the signal deviation curve does not exceed the predetermined signal deviation threshold, the first power device is marked as risk-free.

[0076] When the maximum signal deviation in the signal deviation curve does not exceed the predetermined signal deviation threshold, it indicates that there is no abnormality in the detection of the first power device, and the first power device is marked as risk-free.

[0077] Among them, when the maximum signal deviation in the signal deviation curve does not exceed the predetermined signal deviation threshold, the first power equipment is marked as risk-free, which can improve the refinement of power detection.

[0078] Embodiment 2

[0079] Based on the same inventive concept as the intelligent detection method for abnormal behavior of power scenarios in the aforementioned embodiment, reference is made to Figure 3 For illustration, the present disclosure also provides an intelligent detection system for abnormal behavior in power scenarios, the system comprising:

[0080] A target digital twin model acquisition module, wherein the target digital twin model acquisition module is used to construct a target digital twin model based on multi-dimensional characteristics of a target power scenario, wherein the target power scenario includes M power devices, where M is an integer greater than or equal to 1;

[0081] a first 3D model acquisition module, wherein the first 3D model acquisition module is used to analyze a first image set of a first power device among the M power devices collected by the CCD image sensor through a device registration fusion model to obtain a first 3D model;

[0082] A first restricted area identification obtaining module, the first restricted area identification obtaining module is used to render the first 3D model to the target digital twin model, wherein the first 3D model has an identification of the first restricted area determined based on a multi-dimensional feature analysis of the first power equipment;

[0083] A first spatial distance acquisition module, wherein the first spatial distance acquisition module is used to sense the distance from the object to be inspected in the target digital twin model to the first power equipment based on the inscribed sphere principle to obtain a first spatial distance;

[0084] A predetermined distance threshold acquisition module is used to activate the sound and light alarm device to issue an abnormal alarm when the first spatial distance does not reach a first predetermined distance threshold, and the predetermined distance threshold is a distance range determined based on the analysis of the first restricted area.

[0085] Furthermore, the system further comprises:

[0086] A first image set acquisition module, wherein the first image set acquisition module is used for the CCD image sensor to acquire images of the first power device based on a preset frequency to obtain the first image set;

[0087] A model output result acquisition module, the model output result acquisition module is used to use the point cloud data set of each image in the first image set as input information of the device registration fusion model, and obtain a model output result, the device registration fusion model is an intelligent model constructed based on the principle of random sampling consistency algorithm;

[0088] A texture mapping result obtaining module, wherein the texture mapping result obtaining module is used to perform texture mapping on a first point cloud model constructed according to a model parameter estimation result in the model output result based on a sensor characteristic parameter of the CCD image sensor to obtain a texture mapping result;

[0089] A first 3D model processing module, wherein the first 3D model processing module is used to use the texture mapping result as the first 3D model of the first power device.

[0090] Furthermore, the system further comprises:

[0091] A first parameter estimation result obtaining module, wherein the first parameter estimation result obtaining module is used for the device registration fusion model to obtain a first parameter estimation result of a first model according to a first sample set obtained by random sampling from the point cloud data set;

[0092] A first consistency point set acquisition module, the first consistency point set acquisition module is used to screen and form a first consistency point set, the first consistency point set refers to a set of non-samples whose distances from each non-sample point cloud data in a first non-sample set to the first model meet a second preset distance threshold, and the first non-sample set refers to a sample set obtained after the first sample set is removed from the point cloud data set;

[0093] A second parameter estimation result obtaining module, wherein the second parameter estimation result obtaining module is used to obtain a second parameter estimation result of the first model based on the first consistency point set when the first data amount of the first consistency point set meets a preset quantity threshold, wherein the second parameter estimation result is a parameter estimation result obtained by verification re-estimation;

[0094] A model output result processing module, wherein the model output result processing module is used to replace the first parameter estimation result with the second parameter estimation result as the model output result.

[0095] Furthermore, the system further comprises:

[0096] A secondary random sampling processing module is used for the device registration fusion model to perform secondary random sampling from the point cloud data set when the first data amount of the first consistency point set does not meet the preset quantity threshold.

[0097] Furthermore, the system further comprises:

[0098] A first three-dimensional fence obtaining module, the first three-dimensional fence obtaining module is used to set a first three-dimensional fence of the first power equipment based on the first restricted area, the first three-dimensional fence includes N enclosure columns, N=12, and the N enclosure columns form a rectangular parallelepiped;

[0099] A guard post acquisition module, wherein the guard post acquisition module is used to deploy the infrared emitter in the infrared detector to the first guard post among the N guard posts, wherein the first guard post is any one of the posts located on the ground;

[0100] A second enclosure column obtaining module, wherein the second enclosure column obtaining module is used to deploy the infrared receiver in the infrared detector to a second enclosure column among the N enclosure columns, wherein the second enclosure column is parallel to the first enclosure column, and the plane where the second enclosure column and the first enclosure column are located is perpendicular to the ground;

[0101] A signal deviation curve acquisition module, the signal deviation curve acquisition module is used to generate a signal deviation curve according to the infrared signal transmitted by the infrared transmitter and the infrared signal received by the infrared receiver;

[0102] A signal deviation curve processing module is used to determine whether the sound and light alarm device is activated based on the curve characteristics of the signal deviation curve.

[0103] Furthermore, the system further comprises:

[0104] A module for obtaining historical electric power equipment limit violation records, the module for obtaining historical electric power equipment limit violation records being used to retrieve historical electric power equipment limit violation records in the cloud computer;

[0105] A first historical signal deviation obtaining module, the first historical signal deviation obtaining module is used to extract a first historical limit violation record from the historical power equipment limit violation record, the first historical limit violation record including a first historical signal deviation;

[0106] a predetermined signal deviation threshold value obtaining module, the predetermined signal deviation threshold value obtaining module being used to set a predetermined signal deviation threshold value according to the first historical signal deviation;

[0107] A predetermined signal deviation threshold processing module is used to start the sound and light alarm device to issue an abnormal alarm when the maximum signal deviation in the signal deviation curve exceeds the predetermined signal deviation threshold.

[0108] Furthermore, the system further comprises:

[0109] A risk-free marking processing module is used to mark the first power device as risk-free when the maximum signal deviation in the signal deviation curve does not exceed the predetermined signal deviation threshold.

[0110] The specific example of the intelligent detection method for abnormal behavior of an electric power scene in the aforementioned embodiment 1 is also applicable to an intelligent detection system for abnormal behavior of an electric power scene in this embodiment. Through the aforementioned detailed description of the intelligent detection method for abnormal behavior of an electric power scene, those skilled in the art can clearly know the intelligent detection system for abnormal behavior of an electric power scene in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0111] Embodiment 3

[0112] Figure 4 is a schematic diagram according to the third embodiment of the present disclosure, Figure 4 As shown, the computer device 600 in the present disclosure may include: a processor 601 and a memory 602 .

[0113] The memory 602 is used to store programs. The memory 602 may include volatile memory (English: volatile memory), such as random-access memory (English: random-access memory, abbreviated: RAM), such as static random-access memory (English: static random-access memory, abbreviated: SRAM), double data rate synchronous dynamic random access memory (English: Double Data Rate Synchronous Dynamic Random Access Memory, abbreviated: DDRSDRAM), etc. The memory may also include non-volatile memory (English: non-volatile memory), such as flash memory (English: flash memory). The memory 602 is used to store computer programs (such as applications, functional modules, etc. that implement the above method), computer instructions, etc. The above computer programs, computer instructions, etc. can be partitioned and stored in one or more memories 602. And the above computer programs, computer instructions, data, etc. can be called by the processor 601.

[0114] The above-mentioned computer programs, computer instructions, etc. may be stored in partitions in one or more memories 602 . And the above-mentioned computer programs, computer instructions, etc. may be called by the processor 601 .

[0115] The processor 601 is used to execute the computer program stored in the memory 602 to implement each step of the method involved in the above embodiment.

[0116] For details, please refer to the relevant description in the previous method embodiment.

[0117] The processor 601 and the memory 602 may be independent structures or integrated structures. When the processor 601 and the memory 602 are independent structures, the memory 602 and the processor 601 may be coupled and connected via a bus 603 .

[0118] The computer device of this embodiment can execute the technical solution in the above method. Its specific implementation process and technical principles are the same and will not be repeated here.

[0119] According to an embodiment of the present disclosure, the present disclosure further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed, the steps provided in any of the above embodiments are implemented.

[0120] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of this disclosure can be achieved, and this document is not limited here.

[0121] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principles of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. An intelligent detection method for abnormal behavior in power scenarios, characterized in that: The method is applied to an intelligent detection system for abnormal behavior in an electric power scene, wherein the system is communicatively connected with a CCD image sensor and an acoustic and optical alarm device, and the method comprises: Building a target digital twin model based on multi-dimensional features of a target power scenario, wherein the target power scenario includes M power devices, where M is an integer greater than or equal to 1; Analyzing a first image set of a first power device among the M power devices collected by the CCD image sensor through a device registration fusion model to obtain a first 3D model; Rendering the first 3D model to the target digital twin model, wherein the first 3D model has an identification of a first restricted area determined based on a multi-dimensional feature analysis of the first power equipment; Based on the inscribed sphere principle, the distance from the object to be inspected in the target digital twin model to the first power equipment is sensed to obtain a first spatial distance; When the first spatial distance does not reach a first predetermined distance threshold, the sound and light alarm device is activated to issue an abnormal alarm, wherein the predetermined distance threshold is a distance range determined based on the analysis of the first restricted area; The system is in communication with the infrared detector, and the method further comprises: A first three-dimensional fence of the first electric equipment is set based on the first restricted area, wherein the first three-dimensional fence includes N guarding posts, where N=12, and the N guarding posts form a rectangular parallelepiped; Deploy the infrared emitter in the infrared detector to the first enclosure column among the N enclosure columns, wherein the first enclosure column is any column located on the ground; Deploy the infrared receiver in the infrared detector to the second enclosure column among the N enclosure columns, wherein the second enclosure column is parallel to the first enclosure column, and the plane where the second enclosure column and the first enclosure column are located is perpendicular to the ground; generating a signal deviation curve according to the infrared signal transmitted by the infrared transmitter and the infrared signal received by the infrared receiver; Judging whether the sound and light alarm device is activated or not based on the curve characteristics of the signal deviation curve; The system is stored in a cloud computer, and judging whether the sound and light alarm device is activated based on the curve characteristics of the signal deviation curve includes: Retrieving the historical power equipment infringement records in the cloud computer; Extracting a first historical limit violation record from the historical power equipment limit violation record, wherein the first historical limit violation record includes a first historical signal deviation; Setting a predetermined signal deviation threshold according to the first historical signal deviation; When the maximum signal deviation in the signal deviation curve exceeds the predetermined signal deviation threshold, the sound and light alarm device is activated to issue an abnormal alarm; When the maximum signal deviation in the signal deviation curve does not exceed the predetermined signal deviation threshold, the first power device is marked as risk-free.

2. The method according to claim 1, characterized in that: The step of obtaining the first 3D model comprises: The CCD image sensor acquires images of the first power device based on a preset frequency to obtain the first image set; Using the point cloud data set of each image in the first image set as input information of the device registration fusion model, and obtaining a model output result, wherein the device registration fusion model is an intelligent model constructed based on the principle of random sampling consistency algorithm; Performing texture mapping on a first point cloud model constructed according to a model parameter estimation result in the model output result based on sensor characteristic parameters of the CCD image sensor to obtain a texture mapping result; The texture mapping result is used as the first 3D model of the first power device.

3. The method according to claim 2, characterized in that: The model output result is obtained, including: The device registration fusion model obtains a first parameter estimation result of a first model according to a first sample set randomly sampled from the point cloud data set; Screening and forming a first consistent point set, the first consistent point set refers to a set of non-samples whose distances from each non-sample point cloud data in a first non-sample set to the first model meet a second preset distance threshold, and the first non-sample set refers to a sample set obtained by excluding the first sample set from the point cloud data set; When the first data volume of the first consistency point set meets a preset quantity threshold, a second parameter estimation result of the first model is obtained based on the first consistency point set, where the second parameter estimation result is a parameter estimation result obtained by verification re-estimation; The first parameter estimation result is replaced by the second parameter estimation result as the model output result.

4. The method according to claim 3, characterized in that: When the first data amount of the first consistency point set does not meet the preset quantity threshold, the device registration fusion model performs secondary random sampling from the point cloud data set.

5. An intelligent detection system for abnormal behavior in power scenarios, characterized in that: A method for intelligently detecting abnormal behavior in an electric power scenario according to any one of claims 1 to 4, the system comprising: A target digital twin model acquisition module, wherein the target digital twin model acquisition module is used to construct a target digital twin model based on multi-dimensional characteristics of a target power scenario, wherein the target power scenario includes M power devices, where M is an integer greater than or equal to 1; a first 3D model acquisition module, wherein the first 3D model acquisition module is used to analyze a first image set of a first power device among the M power devices collected by the CCD image sensor through a device registration fusion model to obtain a first 3D model; A first restricted area identification obtaining module, the first restricted area identification obtaining module is used to render the first 3D model to the target digital twin model, wherein the first 3D model has an identification of the first restricted area determined based on a multi-dimensional feature analysis of the first power equipment; A first spatial distance acquisition module, wherein the first spatial distance acquisition module is used to sense the distance from the object to be inspected in the target digital twin model to the first power equipment based on the inscribed sphere principle to obtain a first spatial distance; A predetermined distance threshold acquisition module is used to activate the sound and light alarm device to issue an abnormal alarm when the first spatial distance does not reach a first predetermined distance threshold, and the predetermined distance threshold is a distance range determined based on the analysis of the first restricted area.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 4 are implemented.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.

Images