Method and system for generating intelligent inspection point of substation equipment quickly

Abstract

The application provides a substation equipment intelligent inspection point position rapid generation method and system, relates to the substation inspection technical field, determines inspection demand information, extracts equipment account data set, carries out cluster analysis on multiple equipment information, generates multi-level point position class, maps multiple equipment information according to the inspection demand information according to the multi-level point position class, determines multiple initial inspection point positions, traverses multiple initial inspection point positions and the inspection point position database for matching, generates a point position image matching result, determines multiple inspection point positions, carries out detection feedback, generates feedback response data, and intelligently corrects and optimizes multiple inspection point positions. The technical problems that the existing technology has low substation equipment inspection point position generation efficiency, poor accuracy, long inspection work time and omission are solved. The technical effects of improving the substation equipment inspection point position generation efficiency and accuracy and improving the efficiency and accuracy of the inspection work are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a substation equipment intelligent patrol point rapid generation method and system. BACKGROUND

[0002] In the modern power system, as an important node of power transmission and distribution, the substation undertakes the key task of ensuring the stable operation of the power grid. In order to ensure the safety and reliability of the equipment in the substation, it is necessary to conduct regular patrol and inspection. However, in the prior art, the generation and management of patrol point of substation equipment rely on manual operation or static planning, which lacks dynamic analysis of the running state of the equipment, and it is difficult to reasonably plan the patrol according to the importance and actual state of the equipment. This way leads to low efficiency of patrol work, and the inspection range and frequency cannot be flexibly adjusted according to the risk level of the equipment, affecting the comprehensiveness and accuracy of the inspection.

[0003] In the prior art, there is a technical problem of low efficiency and poor accuracy of substation equipment patrol point generation, which leads to long time-consuming and omission of patrol work. SUMMARY

[0004] The present application provides a substation equipment intelligent patrol point rapid generation method and system, which solves the technical problem of low efficiency and poor accuracy of substation equipment patrol point generation in the prior art, which leads to long time-consuming and omission of patrol work.

[0005] In view of the above problems, the present application provides a substation equipment intelligent patrol point rapid generation method and system.

[0006] In a first aspect of the present application, a substation equipment intelligent patrol point rapid generation method is provided, which comprises: determining patrol demand information by analyzing a plurality of devices of a substation, data importing a plurality of device information according to the patrol demand information, and extracting a device account data set; clustering analysis of the plurality of device information based on the device account data set to generate a multi-level point class; mapping the plurality of device information according to the multi-level point class according to the patrol demand information to determine a plurality of initial patrol points; traversing the plurality of initial patrol points and the patrol point database to generate a point image matching result, and performing point batch processing according to the point image matching result to determine a plurality of patrol points; detecting feedback based on the plurality of patrol points combined with the plurality of device information to generate feedback response data, and intelligently correcting and optimizing the plurality of patrol points through the feedback response data.

[0007] In a second aspect of the present application, a substation equipment intelligent patrol point rapid generation system is provided, and the system comprises: a data extraction module, configured to determine patrol requirement information by analyzing a plurality of devices of a substation, perform data import on a plurality of device information according to the patrol requirement information, and extract device account data sets; an information analysis module, configured to perform cluster analysis on the plurality of device information based on the device account data sets, and generate a plurality of levels of point classes; an initial patrol point obtaining module, configured to map the plurality of device information according to the plurality of levels of point classes according to the patrol requirement information, and determine a plurality of initial patrol points; a patrol point determining module, configured to match the plurality of initial patrol points with a patrol point database, generate a point image matching result, perform point batch processing according to the point image matching result, and determine a plurality of patrol points; and a patrol point correction module, configured to perform detection feedback based on the plurality of patrol points in combination with the plurality of device information, generate feedback response data, and intelligently correct and optimize the plurality of patrol points through the feedback response data.

[0008] The one or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0009] The method provided by the embodiments of the present application performs analysis on a plurality of devices of a substation, determines patrol requirement information, performs data import on a plurality of device information according to the patrol requirement information, and extracts device account data sets; performs cluster analysis on the plurality of device information based on the device account data sets, and generates a plurality of levels of point classes; maps the plurality of device information according to the plurality of levels of point classes according to the patrol requirement information, and determines a plurality of initial patrol points; matches the plurality of initial patrol points with a patrol point database, generates a point image matching result, performs point batch processing according to the point image matching result, and determines a plurality of patrol points; performs detection feedback based on the plurality of patrol points in combination with the plurality of device information, generates feedback response data, and intelligently corrects and optimizes the plurality of patrol points through the feedback response data. The technical effect of improving the generation efficiency and accuracy of substation equipment patrol point and improving the efficiency and accuracy of patrol work is achieved. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0011] Figure 1 A flowchart of a substation equipment intelligent inspection point rapid generation method is provided for the present application.

[0012] Figure 2 A structural diagram of a substation equipment intelligent inspection point rapid generation system is provided for the present application.

[0013] Legend: data extraction module 11, information analysis module 12, initial inspection point obtaining module 13, inspection point determining module 14, and inspection point correcting module 15. DETAILED DESCRIPTION

[0014] The present application provides a substation equipment intelligent inspection point rapid generation method and system, which is used to solve the technical problem of low generation efficiency and poor accuracy of substation equipment inspection point in the prior art, resulting in long time-consuming and missing of inspection work. The technical effects of improving the generation efficiency and accuracy of substation equipment inspection point and improving the efficiency and accuracy of inspection work are achieved.

[0015] Hereinafter, the technical solutions in the present application will be described clearly and completely with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application. In addition, it should be noted that, for convenience of description, only the parts related to the present application are shown in the drawings, not all.

[0016] In one embodiment, as shown in the accompanying drawings, the present application provides a substation equipment intelligent inspection point rapid generation method, which comprises: Figure 1

[0017] By analyzing a plurality of devices of a substation, the inspection requirement information is determined, and the device information is imported according to the inspection requirement information, and the device account data set is extracted.

[0018] ​Specifically, by identifying and collecting information of all devices in the substation one by one, a comprehensive analysis of multiple devices in the substation is performed to obtain device type information and device attribute information of each device. According to the device information obtained by analysis, in combination with the actual operation condition and maintenance strategy of the substation, the inspection requirement information is determined. The inspection requirement information is obtained based on comprehensive analysis of multiple factors such as importance of the device, operating state and environmental factors, and is used to determine the priority of device inspection and the frequency of inspection. After the inspection requirement information is determined, multiple device information is imported into the substation device management system according to the inspection requirement, so as to perform subsequent data analysis and processing. During the data import process, the accuracy and integrity of the data need to be ensured to avoid data omission or error. Finally, the device account data set is extracted from the substation device management system, which is a data record set containing detailed information of multiple devices extracted from the inspection management system, including device type, attribute, priority and other key information. By analyzing and processing these data, the operating state of the device can be further understood, which provides reliable data support for subsequent inspection point generation.

[0019] Further, by analyzing multiple devices of the substation, determining the inspection requirement information, importing multiple device information according to the inspection requirement information, and extracting the device account data set, the method comprises: traversing and identifying multiple devices of the substation to determine the multiple device information, wherein the multiple device information comprises multiple device type information and multiple device attribute information; performing importance evaluation on multiple devices according to the multiple device type information to generate multiple device importance scores; performing requirement analysis based on the multiple device importance scores in combination with operating states of the multiple devices to determine the inspection requirement information; evaluating multiple devices according to the inspection requirement information to set inspection frequency information; integrating multiple sources of data according to the inspection frequency in combination with the multiple device attribute information to classify and import the integrated data set according to the inspection requirement information to obtain the device account data set.

[0020] Specifically, first, traversal recognition is performed, and a plurality of devices in the substation are analyzed and recognized one by one to determine device information of each device. The device information includes device type information such as transformers, circuit breakers, monitoring devices, etc., and device attribute information such as installation location, model, manufacturer, operating life, etc. Then, according to the plurality of device type information obtained by recognition, the importance of each device is evaluated based on the criticality and possible risks in the operation of the power grid, and a plurality of device importance scores are generated. For example, different weights are assigned to each device by the criticality and possible risks of the device, and the importance score of each device is calculated. The criticality of the device is determined based on its function and role in the substation system, and the risk assessment is obtained by comprehensive analysis of factors such as device historical failure data, operating environment, maintenance condition, etc. The weight is a value between 0 and 1, indicating the importance and risk level of the device in the overall system, and the weight allocation can be based on expert evaluation, data analysis or industry standards, etc. After obtaining the plurality of device importance scores, further demand analysis is performed in combination with the operating state of the plurality of devices to determine the inspection demand information of each device. The operating state refers to the current operating performance of the device, such as whether there is abnormal vibration, temperature deviation, etc. By comprehensively analyzing the importance of the device and its operating state, it is identified which devices need to be prioritized for inspection or more frequent monitoring, and the inspection demand information is obtained. Then, according to the inspection demand information, the inspection frequency information of each device is set, i.e., the frequency of inspection is arranged. Finally, according to the inspection frequency, the device attribute information is integrated to integrate the device information from different sources. The integrated data set is classified according to the inspection demand and imported into the substation device management system to form a device account data set, which contains all the device information related to inspection and can provide reliable data information to improve the accuracy of the inspection point.

[0021] Based on the device account data set, a plurality of device information is clustered and analyzed to generate a plurality of levels of point classes.

[0022] Specifically, using the equipment account data set containing detailed information of all equipment in the substation, multiple equipment information is analyzed by clustering through K-Means clustering algorithm, hierarchical clustering or DBSCAN clustering algorithm and other technologies. Clustering analysis is a data mining technology that can classify similar objects into a class, so that objects in the same class are as similar as possible, while objects in different classes are as different as possible. In this process, according to the physical location, functional attributes, operating characteristics and other factors of the equipment, the equipment is divided into different levels and categories, forming a multi-level point class, so as to realize more refined equipment management. For example, key equipment may be classified as a high-priority category, while auxiliary equipment is classified as a secondary category, ensuring that different types and attributes of equipment can be reasonably allocated in subsequent inspections.

[0023] According to the inspection requirement information, the multiple equipment information is mapped according to the multi-level point class, and multiple initial inspection point positions are determined.

[0024] Further, according to the inspection requirement information, the multiple equipment information is mapped according to the multi-level point class, and multiple initial inspection point positions are determined, the method comprising: classifying the inspection requirement information according to the multiple equipment importance scores to obtain multi-level inspection requirement information; performing feature analysis on the multi-level inspection requirement information in combination with the multiple equipment category information to determine multiple equipment inspection requirement features, and establishing a multi-level node structure based on the multiple equipment inspection requirement features; constructing an inspection decision tree according to the multi-level node structure according to a decision rule; traversing the inspection decision tree layer by layer according to the multi-level point class, and matching the multiple equipment information according to the traversal result to determine the multiple initial inspection point positions.

[0025] Specifically, the inspection demand information is classified according to the device importance scores, and the inspection strategies for different devices are refined to obtain multi-level inspection demand information, including high-frequency inspection demand, medium-frequency inspection demand, and low-frequency inspection demand. By classifying according to the scores, key devices such as transformers or switch devices that play a core role in the substation can be prioritized for inspection. Then, the multi-level inspection demand information is associated with the device category information, and the characteristics of each device under different inspection demands are analyzed and identified to obtain multiple device inspection demand characteristics, including but not limited to: device type, inspection frequency, priority, environmental requirements, etc., to clarify the inspection content and focus required for each device during the inspection process. Further, based on the multiple device inspection demand characteristics, the device type, importance level, and inspection frequency are comprehensively analyzed, and nodes of different levels are sequentially constructed to obtain a multi-level node structure. The multi-level node structure established can be: the first level node is used to distinguish different types of devices, such as transformers, circuit breakers, busbars, etc.; the second level node is divided according to the importance level of the device, such as high, medium, and low levels; and the third level node is further refined to specific inspection frequencies or environmental conditions, such as daily inspection, weekly inspection, and inspection under specific weather conditions, etc. Then, according to the multi-level node structure and the pre-defined decision rules, a decision tree for inspection is constructed, wherein the decision rules are defined according to the device type and inspection demand, for example, for all high-importance transformers, they are mapped to the first level inspection point, i.e. these devices need the highest level of attention and inspection. For secondary devices, they are mapped to the second or third level point according to their specific inspection demand. By associating the inspection frequency of the device with the level of the point, the decision tree model can generate reasonable point categories according to different inspection demands, thereby ensuring the pertinence and effectiveness of the inspection work. Finally, the decision tree is traversed layer by layer according to the multi-level point categories. During the traversal process, according to the matching results of the device information and the decision tree nodes, the device is mapped to the multi-level point categories to determine multiple initial inspection points, which can efficiently identify and arrange inspection points with reasonable priorities, meeting the actual needs of intelligent inspection of substation devices.

[0026] Further, the multiple initial inspection points are determined by traversing the patrol decision tree according to the multiple levels of point types, and matching the multiple device information according to the traversal result, and the method comprises: traversing the patrol decision tree to determine a root node of the patrol decision tree, and setting an initial traversal condition according to the multiple levels of point types; taking the root node as a starting point, analyzing a first layer node of the patrol decision tree according to the initial traversal condition, screening the multiple device information according to an analysis result, and generating a first screening result; analyzing a second layer node of the patrol decision tree according to the initial traversal condition according to the first screening result, generating a second screening result, and thus iteratively traversing the patrol decision tree layer by layer until the traversal stops at a leaf node of the patrol decision tree, and generating a second screening result... a Pth screening result, wherein P is an integer greater than or equal to 2, the second layer node is directly connected to the first layer node, and the second layer node is a next layer node of the first layer node; and the first screening result, the second screening result... and the Pth screening result are associated and integrated to obtain the traversal result.

[0027] Specifically, traversing the patrol decision tree first determines the root node of the patrol decision tree, which represents the starting point of the entire patrol decision tree structure. When setting the initial traversal condition, the specific initial traversal condition is defined according to the multi-level point type (such as device type, importance, etc.), to ensure the pertinence of the traversal process. Then, taking the root node as the starting point, the first layer of nodes of the decision tree is parsed according to the set initial traversal condition. At this time, the multiple device information in the layer of nodes will be screened to generate the first screening result. The first screening result contains the devices and their related information that meet the conditions, such as type, importance score and running state, etc. Then, based on the first screening result, the second layer of nodes of the patrol decision tree is parsed according to the same initial traversal condition, thereby generating the second screening result. That is, through the parsing and matching of each layer of nodes, the device information that meets the conditions is gradually screened out, and the second layer of nodes is directly connected to the first layer of nodes, and the second layer of nodes is the next layer of nodes of the first layer of nodes. This process is iterated in the patrol decision tree until the leaf nodes of the patrol decision tree are reached. The leaf node is the end of the tree structure, representing the final patrol point, which is determined based on the multi-level point type and the layer-by-layer screening condition. In the iteration process, the third screening result, the fourth screening result, …, and the Pth screening result will be generated in turn. Since the tree structure has at least two layers of structure, including the root node and the leaf node, P is an integer greater than or equal to 2, representing the depth of the patrol decision tree. Finally, based on the first screening result, the second screening result, …, and the Pth screening result, the different levels of screening results are combined to form a complete traversal result. The traversal result not only reflects the patrol priority and importance level of the device, but also ensures the comprehensiveness and pertinence of the patrol activity. According to the traversal result, a plurality of initial patrol point positions are obtained, which ensures that these point positions cover the multi-level point types parsed according to the decision tree, achieving efficient and accurate patrol planning.

[0028] The plurality of initial patrol point positions are matched with the patrol point position database to generate a point image matching result, and the point image matching result is used for batch processing of the point positions to determine a plurality of patrol point positions.

[0029] Specifically, the plurality of initial patrol point positions are compared one by one with the historical patrol point positions stored in the patrol point position database. The patrol point position database contains image data and related information of each device in the substation, which has been labeled and can be efficiently retrieved and matched. During traversal, the initial point positions and image records in the database are compared one by one based on the geographical location, device type, and patrol requirements of the point positions. By matching the real-time image of the initial patrol point position captured by the camera with the image in the patrol point position database, a point image matching result is generated. According to the point image matching result, point batch processing is performed, which refers to classifying and updating the status of the matched or unmatched points, so as to determine which points need to be prioritized, which points can be combined for processing, or which points need further analysis. According to the classification processing result of the matching result, a plurality of patrol point positions are determined, which refers to the final determined patrol point positions. Through matching processing, not only the distribution of point positions is optimized, but also it is ensured that the patrol task can cover all key devices and high-risk areas, improving the efficiency and accuracy of the patrol work.

[0030] Based on the plurality of patrol point positions and the plurality of device information, feedback detection is performed to generate feedback response data, and the plurality of patrol point positions are intelligently corrected and optimized based on the feedback response data.

[0031] Specifically, the predetermined patrol point positions are combined with the related device information for real-time detection and data collection. Through sensors, cameras, and other monitoring devices deployed at the patrol point positions, information such as device operating status and environmental parameters is obtained to ensure that the inspection results of each patrol point position are accurate and comprehensive. Device information includes device operating status, current working parameters, and fault records, etc. During the detection process, feedback response data is generated based on the current state of the device and the inspection result, which reflects the detection result of each patrol point position, such as whether the device is operating normally, whether there is a fault or potential risk. Finally, based on the actually detected feedback response data, the existing plurality of patrol point positions are intelligently corrected and optimized. Intelligent correction refers to dynamic evaluation and adjustment of the rationality of patrol point position layout, the suitability of monitoring frequency, and the accuracy of monitoring indicators, etc. For example, if the feedback data shows that the device failure rate of a certain patrol point position is abnormally high, the patrol point density in that area is increased or the patrol time is adjusted to strengthen monitoring, and if the feedback data of a certain patrol point position shows that the device is running stably, the patrol frequency can be reduced. Through the feedback and optimization mechanism, not only can device faults be discovered and solved in a timely manner, but also patrol strategies can be continuously optimized, ensuring the flexibility and adaptability of the patrol strategy, improving the comprehensiveness, accuracy, and efficiency of the patrol, optimizing resource allocation, and ultimately achieving intelligent management of substation devices.

[0032] Further, the plurality of initial patrol point positions are matched with the patrol point position database to generate a point position image matching result, the method comprising: obtaining device image data sets by image acquisition on a plurality of devices of a substation, performing image labeling processing on the device image data sets, and constructing the patrol point position database; based on the plurality of initial patrol point positions, image acquisition is performed to obtain point position image data sets; based on the patrol point position database, the plurality of initial patrol point positions are searched to obtain real-time image data sets; the point position image data sets and the real-time image data sets are subjected to vector calculation to determine point position image feature vectors and real-time image feature vectors; similarity calculation is performed on the point position image feature vectors and the real-time image feature vectors, and the point position image data sets and the real-time image data sets are matched according to the calculation result to generate the point position image matching result.

[0033] Specifically, first, image acquisition is performed on multiple devices in the substation to obtain a device image dataset, wherein the device image data is collected by monitoring devices such as cameras installed in the substation, covering different angles and details of the devices. Next, the collected device image dataset is subjected to image labeling processing, that is, each image is labeled with device type, location and other information to ensure that each image can be accurately identified and located, and a patrol point database is constructed according to the labeled images, which includes image information of each key device in the substation to support image retrieval and matching. According to the initial patrol point data, image acquisition is performed by the camera to obtain a corresponding point image dataset, which represents the state information of each patrol point in the substation. Then, based on the patrol point database, a plurality of initial patrol points are retrieved to obtain a real-time image dataset corresponding to the initial patrol points, which is obtained by real-time image acquisition by the camera. The point image dataset and the real-time image dataset are input into a feature extraction algorithm such as SIFT algorithm, HOG algorithm, etc. to determine their respective point image feature vectors and real-time image feature vectors through the feature extraction algorithm. The point image feature vector represents the image visual feature of each patrol point, while the real-time image feature vector reflects the image visual feature of the current substation device, such as the shape, color and key points of the device. Finally, the method includes algorithms based on feature point matching such as Euclidean distance, cosine similarity, Manhattan distance, etc. to calculate the similarity between the point image feature vector and the real-time image feature vector. By comparing the similarity of the point image feature vector and the real-time image feature vector, the point image data and the real-time image data are matched according to the similarity calculation result, and a point image matching result is generated, which is used to verify the correctness of the initial patrol point and the consistency of the actual device position. If the matching result shows high, it means that the patrol point matches well with the actual position of the device. Through image acquisition, labeling, vector calculation and matching analysis, the patrol point can be accurately matched with the actual device in the substation, improving the reliability and effectiveness of the patrol.

[0034] Further, the point image matching result is used for batch processing of the points to determine a plurality of patrol points. The method comprises: classifying the plurality of initial patrol points based on the point image matching result to obtain a point classification result, wherein the point classification result comprises a first point class, a second point class, and a third point class; selecting the plurality of initial patrol points according to the first point class to obtain N points, updating the states of the N points, integrating the N points according to the state information of the N points to generate a standard patrol list, wherein N is an integer greater than 0 and less than M, M is the total number of the plurality of initial patrol points; reviewing W points according to the second point class and the third point class, batch marking the W points according to the review result to generate an abnormal pending list, wherein N+W=M, W is an integer greater than or equal to 0; adjusting the states of the plurality of initial patrol points according to the abnormal pending list, batch screening the plurality of initial patrol points according to the adjustment result and the standard patrol list to determine the plurality of patrol points.

[0035] Specifically, by analyzing the point image results, the multiple initial patrol points are classified into three categories. Specifically, by setting a matching judgment standard, when the similarity is greater than a first preset threshold, such as 90%, it can be considered that the matching is successful, and the initial patrol points greater than the preset threshold are classified into a first point category, i.e., the first point category is the matching successful point. When the similarity is less than a second preset threshold, such as less than 50% or a lower value, it is considered that the points are matching recognition points and are classified into a second point category. Finally, the initial patrol points in the intermediate threshold range, for example, between 50% and 90%, are classified into a third point category, and the third point category is a low similarity point that needs to be reviewed. According to the above rules, the multiple initial patrol points are classified and processed to obtain a point classification result, which can quickly identify which points can be directly used for subsequent patrol and which points need to be further checked or adjusted. The point classification result includes the first point category, the second point category, and the third point category. Then, the points in the first point category, i.e., the matching successful points, are selected from the initial patrol points, denoted as N points, and the states of the N points are updated to matching successful. According to the state information of the N points, the points are integrated to generate a standard patrol list. The standard patrol list contains all confirmed correct patrol points, which facilitates subsequent patrol task planning and arrangement. It should be noted that N represents the number of matching successes, which is an integer greater than 0 and less than M, and M represents the total number of multiple initial patrol points. For the second point category and the third point category, i.e., the matching failed and the low similarity points that need to be reviewed, the review process includes re-collecting images or manual confirmation to determine whether there is an environmental change, image collection error, or device state anomaly. According to the review result, the W points are batch marked to generate an abnormal pending list, where W is the number of points that need to be reviewed, which is an integer greater than or equal to 0, and N+W is equal to M. The abnormal pending list is used to record points that need further attention and processing. Finally, according to the abnormal pending list, the state of the corresponding point in the multiple initial patrol points is adjusted, for example, the state is updated to pending or abnormal. The purpose of state adjustment is to modify the patrol strategy of these points according to the review result. The adjusted initial patrol points are compared with the standard patrol list to ensure that all confirmed points are included in the standard patrol list. According to the adjustment result and the standard patrol list, the initial patrol points are batch screened to finally determine multiple patrol points, which include confirmed standard patrol points and abnormal pending points after review and processing. Through this process, the states of all points are accurately reflected, achieving efficient and accurate screening and adjustment of points, and thus ensuring the accuracy and reliability of the patrol points.

[0036] Further, according to the abnormal to-be-processed list, the state of the plurality of initial inspection points is adjusted, and batch screening is performed according to the adjustment result combined with the standard inspection list to determine the plurality of inspection points. The method comprises: performing abnormal influence analysis on the abnormal to-be-processed list to obtain a plurality of influence levels, performing descending sequence processing on the plurality of influence levels to obtain an abnormal influence sequence; based on the abnormal to-be-processed list, the state of the plurality of initial inspection points is adjusted according to the abnormal influence sequence to generate an adjustment result; a screening condition is set according to the abnormal influence sequence, the adjustment result is compared with the screening condition for judgment, and the judgment result is checked and confirmed, the point that passes the check is integrated with the standard inspection list to determine the plurality of inspection points.

[0037] Specifically, first, abnormal influence analysis is performed on the abnormal to-be-processed list to identify the potential risks and importance of different points. According to the risk size and importance of the point, an influence level is assigned to each point in the abnormal to-be-processed list, and a plurality of influence levels are obtained. The plurality of influence levels are used to reflect the urgency of point processing, and the higher the level, the stronger the urgency. Then, the influence levels are processed in descending sequence to sort from high to low to generate an abnormal influence sequence for subsequent state adjustment and screening. Next, based on the information in the abnormal to-be-processed list, the state of the plurality of initial inspection points is adjusted according to the abnormal influence sequence to generate an adjustment result. State adjustment is to reconfigure the inspection strategy of the point according to its importance and current risk level, for example, to increase the inspection frequency of the emergency processing point or mark it as a high-priority processing object. After state adjustment, a screening condition is set according to the abnormal influence sequence, and the adjustment result is iterated and compared for judgment. For example, according to the state priority of the point, the point marked as "emergency processing" is arranged in the final inspection list as a high-priority processing object; the "to-be-reviewed" point is arranged as a medium-priority to ensure that these points are properly focused in subsequent inspections; and the low-priority point in the regular inspection list is retained but not the focus of this inspection, ensuring that resources are concentrated on more important points. Further, the points that meet the screening condition are checked to confirm whether their state meets the actual demand. The points that pass the check are integrated into the standard inspection list and combined with the adjusted result to form the final plurality of inspection points. This process ensures dynamic adjustment and optimization of the inspection plan, effectively responds to different states of the substation equipment, improves the accuracy and emergency response capability of the inspection points, and ensures the safe operation of the substation equipment.

[0038] Embodiment two, based on the same inventive concept as the substation equipment intelligent inspection point rapid generation method in the foregoing embodiments, such as Figure 2As shown, the present application provides a substation equipment intelligent patrol point position rapid generation system, wherein the system comprises:

[0039] A data extraction module 11 is configured to determine patrol requirement information by parsing a plurality of devices of a substation, perform data import on a plurality of device information according to the patrol requirement information, and extract device account data sets; an information analysis module 12 is configured to perform cluster analysis on the plurality of device information based on the device account data sets, and generate a plurality of point position classes; an initial patrol point position obtaining module 13 is configured to map the plurality of device information according to the plurality of point position classes according to the patrol requirement information, and determine a plurality of initial patrol point positions; a patrol point position determining module 14 is configured to match the plurality of initial patrol point positions with a patrol point position database, generate a point position image matching result, perform point position batch processing according to the point position image matching result, and determine a plurality of patrol point positions; and a patrol point position correction module 15 is configured to generate feedback response data based on detection feedback of the plurality of patrol point positions in combination with the plurality of device information, and perform intelligent correction and optimization on the plurality of patrol point positions through the feedback response data.

[0040] Further, the data extraction module 11 is configured to perform the following steps: iteratively identify a plurality of devices of a substation to determine the plurality of device information, wherein the plurality of device information comprises a plurality of device type information and a plurality of device attribute information; perform importance evaluation on a plurality of devices according to the plurality of device type information to generate a plurality of device importance scores; perform demand analysis based on the plurality of device importance scores in combination with the operating states of the plurality of devices to determine the patrol requirement information; perform evaluation on a plurality of devices according to the patrol requirement information to set patrol cycle information; perform multi-source data integration on a plurality of devices according to the patrol cycle in combination with the plurality of device attribute information, classify and import integrated data sets according to the patrol requirement information, and obtain the device account data sets.

[0041] Further, the initial patrol point position obtaining module 13 is configured to perform the following steps: classify the patrol requirement information according to the plurality of device importance scores to obtain a plurality of levels of patrol requirement information; perform feature analysis on a plurality of device category information according to the plurality of levels of patrol requirement information to determine a plurality of device patrol requirement features, establish a plurality of levels of node structures based on the plurality of device patrol requirement features; construct a patrol decision tree according to the plurality of levels of node structures according to a decision rule; perform layer-by-layer iteration on the plurality of levels of point position classes through the patrol decision tree, match the plurality of device information according to the iteration results, and determine the plurality of initial patrol point positions.

[0042] Further, the initial inspection point obtaining module 13 is further configured to perform the following steps: traversing the inspection decision tree to determine a root node of the inspection decision tree, and setting an initial traversal condition according to the multi-level point class; taking the root node as a starting point, and parsing a first layer node of the inspection decision tree according to the initial traversal condition, and filtering the device information according to the parsing result to generate a first filtering result; parsing a second layer node of the inspection decision tree according to the first filtering result and the initial traversal condition to generate a second filtering result, and thus iteratively traversing the inspection decision tree layer by layer until the leaf node of the inspection decision tree is reached to stop, and generating a second filtering result... a Pth filtering result, where P is an integer greater than or equal to 2, the second layer node is directly connected to the first layer node, and the second layer node is a next layer node of the first layer node; and performing association and integration based on the first filtering result, the second filtering result... and the Pth filtering result to obtain the traversal result.

[0043] Further, the inspection point determining module 14 is configured to perform the following steps: obtaining device image data sets by image acquisition of multiple devices of a substation, performing image labeling processing on the device image data sets, and constructing the inspection point database; obtaining point image data sets based on image acquisition of the multiple initial inspection points; obtaining real-time image data sets based on retrieval of the multiple initial inspection points according to the inspection point database; performing vector calculation on the point image data sets and the real-time image data sets to determine point image feature vectors and real-time image feature vectors; performing similarity calculation on the point image feature vectors and the real-time image feature vectors, and matching the point image data sets and the real-time image data sets according to the calculation result to generate a point image matching result.

[0044] Further, the patrol point determination module 14 is further configured to perform the following steps: classifying the plurality of initial patrol points based on the point image matching result to obtain a point classification result, the point classification result including a first point class, a second point class, and a third point class; selecting the plurality of initial patrol points according to the first point class to obtain N points, updating the states of the N points, integrating the N points according to the state information of the N points to generate a standard patrol list, wherein N is an integer greater than 0 and less than M, M is the total number of the plurality of initial patrol points; reviewing W points according to the second point class and the third point class, batch marking the W points according to the review result to generate an abnormality to-be-handled list, wherein N+W=M, W is an integer greater than or equal to 0; adjusting the states of the plurality of initial patrol points according to the abnormality to-be-handled list, batch screening according to the adjustment result and the standard patrol list to determine the plurality of patrol points.

[0045] Further, the patrol point determination module 14 is further configured to perform the following steps: performing abnormality impact analysis on the abnormality to-be-handled list to obtain a plurality of impact levels, serializing the plurality of impact levels in descending order to obtain an abnormality impact sequence; adjusting the states of the plurality of initial patrol points according to the abnormality impact sequence based on the abnormality to-be-handled list to generate an adjustment result; setting a screening condition according to the abnormality impact sequence, comparing the adjustment result with the screening condition to determine, according to the determination result, whether to pass the verification, integrating the points that pass the verification with the standard patrol list to determine the plurality of patrol points.

[0046] The above merely describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0047] The specification and drawings merely illustrate the exemplary embodiments of the present application and are considered to cover any and all modifications, variations, combinations or equivalents that fall within the scope of the present application. Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and equivalent technology, the present application is intended to include these modifications and variations.

Claims

1. A method for generating intelligent patrol point positions of substation equipment quickly, characterized in that, The method comprises: By analyzing a plurality of devices of a substation, determining inspection requirement information, importing a plurality of device information according to the inspection requirement information, and extracting a device account data set; Based on the device account data set, the plurality of device information is clustered and analyzed to generate a multi-level point site class; According to the inspection requirement information, the plurality of device information is mapped according to the multi-level point site class to determine a plurality of initial inspection point sites; Iterate the plurality of initial inspection point sites and the inspection point site database to match and generate a point site image matching result, perform point site batch processing according to the point site image matching result, and determine a plurality of inspection point sites; Based on the plurality of inspection point sites, the plurality of device information is combined for detection feedback to generate feedback response data, and the plurality of inspection point sites are intelligently corrected and optimized through the feedback response data; By analyzing a plurality of devices of a substation, determining inspection requirement information, importing a plurality of device information according to the inspection requirement information, and extracting a device account data set, the method comprises: Iterate and identify a plurality of devices of a substation to determine the plurality of device information, which includes a plurality of device type information and a plurality of device attribute information; According to the plurality of device type information, the importance of a plurality of devices is evaluated to generate a plurality of device importance scores; Based on the plurality of device importance scores, the running state of a plurality of devices is combined for demand analysis to determine the inspection requirement information; According to the inspection requirement information, a plurality of devices are evaluated, and inspection frequency information is set; According to the inspection frequency and the plurality of device attribute information, a plurality of devices are integrated with multi-source data, the integrated data set is classified and imported according to the inspection requirement information, and the device account data set is obtained; According to the inspection requirement information, the plurality of device information is mapped according to the multi-level point site class to determine a plurality of initial inspection point sites, which comprises: According to the plurality of device importance scores, the inspection requirement information is classified to obtain multi-level inspection requirement information; According to the multi-level inspection requirement information and the plurality of device category information, feature analysis is performed to determine a plurality of device inspection requirement features, and a multi-level node structure is established based on the plurality of device inspection requirement features; According to the multi-level node structure, an inspection decision tree is constructed according to the decision rule; According to the multi-level point site class, the inspection decision tree is iterated layer by layer, the plurality of device information is matched according to the iteration result, and the plurality of initial inspection point sites are determined; According to the multi-level point site class, the inspection decision tree is iterated layer by layer, the plurality of device information is matched according to the iteration result, and the plurality of initial inspection point sites are determined, which comprises: Iterate the inspection decision tree to determine the root node of the inspection decision tree, and set the initial iteration condition according to the multi-level point site class; Taking the root node as the starting point, the first layer node of the inspection decision tree is analyzed according to the initial iteration condition, the plurality of device information is filtered according to the analysis result, and a first filtering result is generated; According to the first screening result, the second layer node of the patrol decision tree is parsed according to the initial traversal condition, and the second screening result is generated, thereby the patrol decision tree is iterated layer by layer until the leaf node of the patrol decision tree is traversed to stop, and the second screening result...The P screening result is generated, wherein P is an integer greater than or equal to 2, the second layer node is directly connected with the first layer node, and the second layer node is the next layer node of the first layer node; Based on the first screening result, the second screening result...The P screening result is associated and integrated to obtain the traversal result.

2. The method of claim 1, wherein the method further comprises: Matching the plurality of initial patrol point positions with the patrol point position database generates a point position image matching result, the method comprising: Obtaining device image data set by image acquisition of a plurality of devices of a substation, image labeling processing of the device image data set, and constructing the patrol point position database; Based on the plurality of initial patrol point positions, image acquisition is performed to obtain a point position image data set; Based on the patrol point position database, the plurality of initial patrol point positions are retrieved to obtain a real-time image data set; Vector calculation is performed on the point position image data set and the real-time image data set to determine point position image feature vectors and real-time image feature vectors; According to the similarity calculation result of the point position image feature vectors and the real-time image feature vectors, the point position image data set and the real-time image data set are matched to generate the point position image matching result.

3. The method of Claim 1, wherein, According to the point position image matching result, point batch processing is performed to determine a plurality of patrol point positions, the method comprising: Based on the point position image matching result, the plurality of initial patrol point positions are classified to obtain a point classification result, and the point classification result includes a first point class, a second point class and a third point class; According to the first point class, the plurality of initial patrol point positions are selected to obtain N point positions, the states of the N point positions are updated, the N point positions are integrated according to the N point state information to generate a standard patrol list, wherein N is an integer greater than 0 and less than M, and M is the total number of the plurality of initial patrol point positions; According to the second point class and the third point class, W point positions are reviewed, and the W point positions are batch marked according to the review result to generate an abnormal pending list, wherein N+W=M, and W is an integer greater than or equal to 0; According to the abnormal pending list, the state of the plurality of initial patrol point positions is adjusted, and the plurality of patrol point positions are determined by batch screening combined with the standard patrol list according to the adjustment result.

4. The substation equipment intelligent patrol point rapid generation method of claim 3, wherein, According to the abnormal pending list, the state of the plurality of initial patrol point positions is adjusted, and the plurality of patrol point positions are determined by batch screening combined with the standard patrol list according to the adjustment result, the method comprising: Abnormal influence analysis is performed on the abnormal pending list to obtain a plurality of influence levels, the plurality of influence levels are sequentially processed in descending order to obtain an abnormal influence sequence; Adjusting states of the plurality of initial inspection points according to the abnormal influence sequence based on the abnormal to-be-handled list, to generate an adjustment result; Setting a screening condition according to the abnormal influence sequence, comparing the adjustment result with the screening condition to make a judgment, checking and confirming according to a judgment result, integrating a point position that passes the checking with the standard inspection list, and determining the plurality of inspection point positions.

5. A substation equipment intelligent patrol point position rapid generation system, characterized in that, Steps for implementing the method in any one of claims 1 to 4, comprising: A data extraction module configured to determine inspection requirement information by parsing a plurality of devices of a substation, perform data import on a plurality of device information according to the inspection requirement information, and extract device account data sets; An information analysis module configured to perform cluster analysis on the plurality of device information based on the device account data sets, to generate a plurality of point position classes; An initial inspection point position obtaining module configured to map the plurality of device information according to the plurality of point position classes according to the inspection requirement information, to determine a plurality of initial inspection point positions; An inspection point position determining module configured to match the plurality of initial inspection point positions with an inspection point position database, to generate a point position image matching result, to perform point position batch processing according to the point position image matching result, and to determine a plurality of inspection point positions; An inspection point position correction module configured to perform detection feedback on the plurality of inspection point positions based on the plurality of device information, to generate feedback response data, and to intelligently correct and optimize the plurality of inspection point positions through the feedback response data.

Images