An abnormal hidden danger identification system and method according to power transmission line inspection data

Abstract

The present application relates to a kind of according to the abnormal hidden danger identification system and method of transmission line inspection data, system includes: inspection console module, for by multiple inspection channels respectively management corresponding inspection equipment is inspected and its report form;Inspection behavior collection module, for receiving each inspection equipment submitted by the inspection data of inspection console module in management;Inspection behavior analysis module, for data clustering according to historical inspection data, and the clustering result is marked as abnormal by historical abnormal data, whether the current inspection data is classified as marked as abnormal clustering dataset by using nearest neighbor algorithm calculation, and the Euclidean distance between current inspection data and various hidden danger data in corresponding classified marked as abnormal clustering dataset is calculated, to calculate the probability value of occurrence certain hidden danger.Compared with prior art, the present application automatically analyzes various inspection data, can accurately and reliably calculate the probability value of occurrence hidden danger, improves inspection efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power transmission line inspection, in particular to an abnormal hidden danger identification system and method based on power transmission line inspection data. BACKGROUND

[0002] A large amount of structured and unstructured data (such as picture and video data) is generated in power transmission line inspection, including inspection data uploaded by inspection personnel using mobile terminals, monitoring data of fixed hidden danger monitoring points, and inspection data photographed by unmanned aerial vehicles. Screening and identifying hidden dangers through manual work is time-consuming and laborious, low in efficiency, high in cost, and difficult to operate. In particular, there is a certain risk in the inspection of some areas. In general, for long-distance power transmission lines, factors such as long distance, wide range, and complex environment greatly limit the working mode of traditional manual inspection, resulting in high cost, difficult operation, and difficulty in popularization. SUMMARY

[0003] The present application aims to overcome the problems of low efficiency, reliance on manual experience, and difficulty in standardization in traditional inspection during the process of power transmission line inspection, and to provide an abnormal hidden danger identification system and method based on power transmission line inspection data.

[0004] The object of the present application can be achieved by the following technical solutions:

[0005] An abnormal hidden danger identification system based on power transmission line inspection data, comprising:

[0006] An inspection console module configured to manage corresponding inspection equipment for inspection and its report through multiple inspection channels, register new inspection equipment accessed, and send electronic keys, and only process inspection data accompanied by verified electronic keys;

[0007] An inspection behavior collection module configured to establish a unified inspection behavior collection interface, receive inspection data submitted by each inspection equipment managed in the inspection console module, and store the cleaned inspection data in the database;

[0008] The inspection behavior analysis module is configured to: perform data clustering in advance according to historical inspection data, and perform anomaly labeling on the clustering results through historical abnormal data, adopt a nearest neighbor algorithm to calculate whether current inspection data obtained through the inspection behavior collection module is classified into a labeled abnormal clustering data set, and calculate the Euclidean distance between the current inspection data and various hidden danger data in the corresponding classified labeled abnormal clustering data set, so as to calculate a probability value of occurrence of a certain hidden danger.

[0009] Further, the inspection channel accessed device includes one or more of a mobile terminal, a fixed inspection camera and a drone.

[0010] The report management and electronic key sending process performed by the inspection console module is specifically as follows:

[0011] The key management process: generating an electronic key before a new inspection device is accessed, and sending the electronic key to the new inspection device after the new inspection device is accessed and registered;

[0012] The report management process: storing the inspection information submitted by each inspection device, historical inspection data and the probability value of occurrence of a hidden danger judged by the historical data of the accessed inspection device for query.

[0013] Further, the inspection data collected by the inspection behavior collection module includes: inspection device information, a geographic position, a time of performing an inspection operation, information of a target to be inspected and unstructured data generated during the inspection.

[0014] Further, the calculation process of the probability value of occurrence of a hidden danger is specifically as follows:

[0015] If the current inspection device has no historical data of occurrence of a hidden danger, the probability value of occurrence of a hidden danger is calculated according to the distance difference between the current inspection data of the current inspection device and various hidden danger data in a labeled abnormal clustering data set of the same target, the distance difference including a Euclidean distance difference, a geographic position difference, a time difference of an inspection operation and an unstructured data difference;

[0016] At this time, the first calculation expression of the probability value of occurrence of a hidden danger of the current inspection device is as follows:

[0017] The probability value = the Euclidean distance difference * the first weight + the geographic position difference * the second weight + the time difference of the inspection operation * the third weight + the unstructured data difference * the fourth weight.

[0018] Further, if the current inspection equipment has historical data of hidden trouble, a probability value of hidden trouble is calculated according to a distance difference between the current inspection data of the current inspection equipment and various hidden trouble data in the cluster data set marked as abnormal, the distance difference including a Euclidean distance difference, a geographical position difference, a time difference of inspection operation and an unstructured data difference;

[0019] At this time, the second calculation expression of the probability value of hidden trouble of the current inspection equipment is:

[0020] Probability value = historical probability value of hidden trouble + Euclidean distance difference * fifth weight + geographical position difference * sixth weight + time difference of operation request * seventh weight + unstructured data difference * eighth weight.

[0021] Further, the inspection behavior analysis module adopts a mean shift clustering algorithm for clustering calculation.

[0022] Further, the inspection behavior analysis module respectively performs clustering calculation on historical data of the same target and clustering calculation on historical data of the same target.

[0023] Further, the inspection behavior collection module transmits data information through encryption.

[0024] Further, the inspection behavior analysis module judges whether the calculated probability value of hidden trouble exceeds a preset alarm threshold, if yes, a prompt and a notification are performed, otherwise, the probability value of hidden trouble is continuously calculated in real time.

[0025] The application also provides an abnormal hidden trouble identification method according to transmission line inspection data, including the following steps:

[0026] Historical inspection data and historical abnormal data of inspection equipment in a transmission line are acquired in advance, the historical inspection data is clustered, and the clustering result is marked as abnormal according to the historical abnormal data;

[0027] Current inspection data of each inspection equipment is acquired in real time, a nearest neighbor algorithm is adopted to calculate whether the current inspection data is classified into the cluster data set marked as abnormal, and a Euclidean distance between the current inspection data and the corresponding classified cluster data set is calculated, so as to calculate the probability value of hidden trouble.

[0028] Compared with the prior art, the application has the following advantages:

[0029] (1) The inspection information is comprehensively analyzed through multiple channels, and the key point is to analyze and predict the probability of hidden trouble of the inspection target.

[0030] (2) Traditional inspection is also used as an inspection channel, and it is analyzed in combination with new inspection methods such as the Internet of Things and drones.

[0031] (3) Based on the historical inspection data and hidden danger data of this target and the historical inspection data and hidden danger data of similar targets, a hidden danger probability analysis is conducted on this target.

[0032] (4) This system realizes the collection of online inspection behavior through multiple channels and classifies inspection information in real time. When the current inspection information matches the abnormal hidden danger classification, it provides immediate hidden danger prediction prompts.

[0033] (5) The effect is that the inspection information can be accumulated for many years, which solves the problems of high cost, difficult operation and difficult promotion encountered in the inspection of transmission lines, and is conducive to the promotion and long-term development of the inspection system.

[0034] (6) The present invention calculates the probability value of potential hazards in the inspection equipment by first considering the historical data of potential hazards in the current inspection equipment and taking into account the historical probability value of potential hazards. If there is no historical data of potential hazards in the current inspection equipment, the calculation is based on the historical hazard information of similar targets. The calculation is comprehensive and accurate. In the formula for calculating the probability value of potential hazards, the distance difference between the current data and the various hazard data in the corresponding clustered data marked as abnormal is calculated. The distance difference takes into account the Euclidean distance difference, the geographical location difference, the time difference of the inspection operation, and the unstructured data difference. Combining these four parameters, the probability value of potential hazards in the current inspection equipment can be calculated more accurately and reliably. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of an abnormal hazard identification system based on power transmission line inspection data provided in an embodiment of the present invention. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0037] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0038] It should be noted that like reference numerals and characters refer to like items throughout the drawings and like items need not be further defined and explained in subsequent drawings once defined and explained in one drawing.

[0039] Embodiment 1

[0040] The inspection system needs to solve the problems of high cost, difficult operation and difficult promotion in power transmission line inspection, and can use the inspection information and abnormal hidden dangers accumulated for many years to finally ensure the analysis and prediction of abnormal hidden dangers. The embodiment provides an abnormal hidden danger identification method based on power transmission line inspection data, which comprises the following steps:

[0041] The historical inspection data and historical abnormal data of the inspection equipment in the power transmission line are acquired in advance, the historical inspection data is clustered, and the clustering results are marked as abnormal according to the historical abnormal data;

[0042] The current inspection data of each inspection equipment is acquired in real time, whether the current inspection data is classified into the clustering data set marked as abnormal is calculated by using a nearest neighbor algorithm, and the Euclidean distance between the current inspection data and the corresponding classified clustering data set is calculated, so as to calculate the probability value of the occurrence of hidden dangers.

[0043] Specifically, in the embodiment, the inspection behavior collection interface is collected, and the feature set of the historical inspection data of the inspection object and the inspection information of the same type of object are used to cluster and calculate the historical inspection data by using a mean shift clustering algorithm. This calculation is divided into two parts, that is, clustering calculation of the historical data of the same type of object and clustering calculation of the historical data of the same object. After clustering, the classification is marked by using the historical abnormal data, and whether the features of the current inspection behavior are classified into the feature set of abnormal hidden dangers is calculated by using a nearest neighbor algorithm. According to the Euclidean distance, the probability of the occurrence of a certain hidden danger is calculated according to the algorithm, and a prompt is given.

[0044] As shown in Figure 1 The embodiment also provides an abnormal hidden danger identification system based on power transmission line inspection data, which comprises:

[0045] The inspection console module is configured to manage the inspection and its report of the corresponding inspection equipment through a plurality of inspection channels, register a new inspection equipment and send an electronic key, and only process the inspection data accompanied by the electronic key verified;

[0046] The inspection behavior collection module is configured to establish a unified inspection behavior collection interface, receive the inspection data submitted by each inspection equipment managed in the inspection console module, and store the inspection data in the warehouse after cleaning the inspection data;

[0047] The inspection behavior analysis module is configured to: perform data clustering in advance according to historical inspection data, and perform abnormal marking on the clustering results through historical abnormal data, adopt a nearest neighbor algorithm to calculate whether current inspection data obtained through the inspection behavior collection module is classified into a marked abnormal clustering data set, and calculate the Euclidean distance between the current inspection data and the corresponding classified clustering data set, so as to calculate a hidden danger occurrence probability value.

[0048] The modules are further described below respectively.

[0049] 1. The inspection console module

[0050] In the inspection console module, the devices accessed by the inspection channel include one or more of a mobile terminal, a fixed inspection camera and a drone.

[0051] The report management and electronic key sending process performed by the inspection console module is specifically as follows:

[0052] The key management process: before a new inspection device is accessed, an electronic key is generated, and after the new inspection device is accessed and registered, the electronic key is sent to the new inspection device.

[0053] The report management process: for the accessed inspection devices, the inspection information submitted by each inspection device, historical inspection data and the probability value of hidden danger occurrence judged by history are respectively stored for query.

[0054] In this embodiment, the inspection console module is used to manage the functions of multiple channels using respective inspection devices for inspection and report management, when a new terminal device is accessed, terminal registration is first needed, that is, the terminal device is guided to perform authorized registration, after successful registration, a unique electronic key for subsequent initiation of inspection upload is provided, and after the subsequent inspection information collection is completed, query, report and statistics functions are also provided.

[0055] According to the embodiment of the application, the inspection behavior collection channel can include: a mobile terminal, a fixed terminal and a drone terminal. The inspection process can include: terminal registration, guiding the terminal device to perform authorized registration, after successful registration, providing a unique electronic key for subsequent initiation of inspection upload; inspection information collection, after the terminal performs an inspection behavior, the key can be used for automatic upload without secondary confirmation; and subsequent query, report and statistics functions.

[0056] 2. The inspection behavior collection module

[0057] The inspection data collected by the inspection behavior collection module includes: inspection device information, geographical position, time of inspection operation, inspected target information and unstructured data generated during inspection when a patrol inspection operation request is made.

[0058] As a preferred embodiment, the inspection behavior collection module transmits data information through encryption.

[0059] In this embodiment, the inspection behavior collection module establishes a unified inspection behavior collection interface, receives structured and unstructured data (pictures or videos) submitted by the inspection, and stores the data after data cleaning, wherein the inspection information includes device information of the inspection target (unique identification of the locatable inspection target), geographical position (precise GPS value or administrative division address obtained according to IP), time of the inspection operation (start time and end time), information of the inspected target (category information and the like), and unstructured data (i.e., picture or video information and the like) generated during the inspection.

[0060] According to the embodiment of the present application, the inspection behavior collection module can collect inspection information, all inspection messages need to be transmitted through encryption, and secondary development is used when necessary, and multiple channel access needs to be supported; after the inspection behavior collection module collects the inspection information, the inspection information data needs to be cleaned before being stored.

[0061] 3. Inspection behavior analysis module

[0062] In this embodiment, the inspection behavior analysis module performs clustering calculation on historical inspection data by using a mean shift clustering algorithm, and the calculation is divided into two parts, i.e., clustering calculation on historical data of the same target and clustering calculation on historical data of the same target. After clustering is completed, historical abnormal data is used to label the classification, and a nearest neighbor algorithm is used to calculate whether the features of the current inspection behavior belong to the feature set of abnormal hidden dangers, and according to the Euclidean distance, the algorithm is used to calculate the probability of occurrence of a certain hidden danger, and a prompt is given.

[0063] The Euclidean distance used by the inspection behavior analysis module can calculate the abnormal hidden danger probability of each inspection channel. If there is no hidden danger data of the current target in other channels, the Euclidean distance difference value, geographical position difference value, time difference value of the inspection operation, and unstructured data difference value of each abnormal hidden danger cluster of the same target are calculated according to the inspection information;

[0064] At this time, the first calculation expression of the probability value of the hidden danger of the current inspection device is:

[0065] Probability value = Euclidean distance difference value * first weight + geographical position difference value * second weight + time difference value of the inspection operation * third weight + unstructured data difference value * fourth weight.

[0066] If there is hidden danger data of the current target in other channels or the current channel, the data of the current target is preferentially used for calculation, and the calculation formula needs to increase the historical hidden danger value of the current target:

[0067] The second calculation expression of the probability value of the current inspection equipment having hidden trouble at this time is:

[0068] The probability value = the historical probability value of having hidden trouble + the Euclidean distance difference value * the fifth weight + the geographical position difference value * the sixth weight + the operation request time difference value * the seventh weight + the unstructured data difference value * the eighth weight.

[0069] In the embodiment, the inspection behavior analysis module judges whether the calculated probability value of having hidden trouble exceeds a preset alarm threshold, and if yes, a prompt and a notification are performed, otherwise, the probability value of having hidden trouble is continuously calculated in real time.

[0070] The preferred embodiments of the application are described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and changes without creative labor based on the concept of the application. Therefore, any technical solutions obtained by logical analysis, reasoning or limited experiments based on the prior art within the concept of the application should be within the protection scope determined by the claims.

Claims

1. An abnormality hazard identification system according to power transmission line inspection data, characterized by, The application comprises: An inspection console module configured to manage the inspection of corresponding inspection equipment and its report through multiple inspection channels, register and send electronic keys to new access inspection equipment, and only process inspection data with verified electronic keys; An inspection behavior collection module configured to establish a unified inspection behavior collection interface, receive inspection data submitted by each inspection equipment managed by the inspection console module, and store the cleaned inspection data in the database; An inspection behavior analysis module configured to pre-cluster data according to historical inspection data, mark the clustering results as abnormal through historical abnormal data, calculate whether the current inspection data obtained through the inspection behavior collection module is classified as abnormal clustering data set using the nearest neighbor algorithm, and calculate the Euclidean distance between the current inspection data and various hidden danger data in the corresponding classified abnormal clustering data set, thereby calculating the probability value of the occurrence of a certain hidden danger; The calculation process of the probability value of the occurrence of a hidden danger is as follows: If the current inspection equipment has no historical data of the occurrence of a hidden danger, the probability value of the occurrence of a hidden danger is calculated according to the distance difference between the current inspection data of the current inspection equipment and various hidden danger data in the abnormal clustering data set of the same target, including the Euclidean distance difference, the geographical position difference, the time difference of the inspection operation, and the unstructured data difference; At this time, the first calculation expression of the probability value of the occurrence of a hidden danger of the current inspection equipment is as follows: Probability value = Euclidean distance difference * first weight + geographical position difference * second weight + time difference of inspection operation * third weight + unstructured data difference * fourth weight.

2. The system for identifying abnormal hidden dangers according to the inspection data of a power transmission line according to claim 1, characterized in that, The equipment accessed by the inspection channel includes one or more of a mobile terminal, a fixed inspection camera, and a drone; The report management and electronic key sending process of the inspection console module is as follows: Key management process: generate an electronic key before a new inspection equipment is accessed, and send the electronic key to the new inspection equipment after the new inspection equipment is accessed and registered; Report management process: store the inspection information, historical inspection data, and the probability value of the occurrence of a hidden danger judged by the historical data submitted by each inspection equipment for query.

3. The system for identifying abnormal hidden dangers according to the inspection data of a power transmission line according to claim 1, characterized in that, The inspection data collected by the inspection behavior collection module includes the inspection equipment information, geographical position, time of inspection operation, inspected target information, and unstructured data generated during inspection.

4. The system for identifying abnormal hidden dangers according to the inspection data of a power transmission line according to claim 1, characterized in that, If the current inspection equipment has historical data of the occurrence of a hidden danger, the probability value of the occurrence of a hidden danger is calculated according to the distance difference between the current inspection data of the current inspection equipment and various hidden danger data in the abnormal clustering data set of the current inspection equipment, including the Euclidean distance difference, the geographical position difference, the time difference of the inspection operation, and the unstructured data difference; At this time, the second calculation expression of the probability value of the occurrence of a hidden danger of the current inspection equipment is as follows: The probability value = the historical probability value of hidden trouble occurrence + the Euclidean distance difference value * the fifth weight + the geographical position difference value * the sixth weight + the time difference value of operation request * the seventh weight + the unstructured data difference value * the eighth weight.

5. The system for identifying abnormal hidden dangers according to the data of power transmission line inspection according to claim 1, characterized in that, The inspection behavior analysis module adopts a mean shift clustering algorithm for clustering calculation.

6. The system for identifying abnormal hidden dangers according to the data of power transmission line inspection according to claim 1, characterized in that, The inspection behavior analysis module respectively performs clustering calculation on historical data of the same target and historical data of the same target.

7. The system for identifying abnormal hazards from inspection data of a power transmission line according to claim 1, wherein The inspection behavior collection module transmits data information through encryption.

8. The system for identifying abnormal hazards from inspection data of a power transmission line according to claim 1, wherein The inspection behavior analysis module judges whether the calculated probability value of hidden trouble occurrence exceeds a preset alarm threshold, and if yes, performs prompting and notification, and if not, continues to calculate the probability value of hidden trouble occurrence in real time.

9. A method for identifying abnormal hidden dangers from inspection data of a power transmission line, characterized by, The method comprises the following steps: Pre-acquire historical inspection data and historical abnormal data of the inspection equipment in the power transmission line, perform data clustering on the historical inspection data, and perform abnormal marking on the clustering results according to the historical abnormal data; Real-time acquire current inspection data of each inspection equipment, adopt a nearest neighbor algorithm to calculate whether the current inspection data is classified into the clustering data set marked as abnormal, and calculate the Euclidean distance between the current inspection data and the corresponding classified clustering data set, so as to calculate the probability value of hidden trouble occurrence; The calculation process of the probability value of hidden trouble occurrence is specifically as follows: If the current inspection equipment has no historical data of hidden trouble occurrence, the probability value of hidden trouble occurrence is calculated according to the distance difference value between the current inspection data of the current inspection equipment and various hidden trouble data in the clustering data set marked as abnormal in the same target, and the distance difference value includes the Euclidean distance difference value, the geographical position difference value, the time difference value of inspection operation and the unstructured data difference value; At this time, the first calculation expression of the probability value of hidden trouble occurrence of the current inspection equipment is as follows: The probability value = the Euclidean distance difference value * the first weight + the geographical position difference value * the second weight + the time difference value of inspection operation * the third weight + the unstructured data difference value * the fourth weight.

Images