A method and system for abnormal identification of distribution network equipment

Abstract

The present application relates to a method and system for identifying abnormalities in distribution network equipment. After obtaining the set of operation status information of distribution network equipment, it is possible to further determine the set of operation status information in the defective environment. In this way, different state description mining subnets can be used to specifically mine and identify different operation status information, so as to accurately and completely obtain different state description distributions. Further, by integrating different state description distributions and invoking the equipment defect identification subnet, it is possible to accurately locate the quantization possibility of the defect theme pointed to by the operation status information of the distribution network equipment. In this way, it is possible to accurately locate the abnormal identification situation corresponding to the set of operation status information of the distribution network equipment based on the quantization possibility of the defect theme, and thus analyze in combination with the equipment operation level and the defective environment level to improve the accuracy and credibility of the abnormal identification of distribution network equipment.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of distribution network security analysis, and in particular to a method and system for identifying abnormal distribution network equipment. Background Art

[0002] Distribution substations (distribution rooms) / substations are important facilities in the power system. The abnormal monitoring and identification of distribution substations and substations are important guarantees for the safety of the entire distribution network facilities and operation safety. Ensuring the safety of the equipment on-site environment plays an important role in the stable and reliable operation of the power system. However, in practical applications, the scale of the distribution network is relatively large, which may increase the difficulty of abnormal identification. For example, related technologies often only perform abnormal identification of the distribution network based on one level, making it difficult to ensure the accuracy and credibility of abnormal identification. Summary of the Invention

[0003] In view of this, the embodiments of the present application provide a method and system for identifying abnormal distribution network equipment.

[0004] The embodiments of the present application provide a method for identifying abnormal distribution network equipment, which is applied to an abnormal distribution network equipment identification system. The method at least includes: collecting a set of distribution network equipment operation status information; wherein, the set of distribution network equipment operation status information includes several groups of distribution network equipment operation status information with a sequential order; using the set of distribution network equipment operation status information to determine a set of defect environment operation status information; wherein, the set of defect environment operation status information includes several groups of defect environment operation status information with a sequential order; combining the set of distribution network equipment operation status information, and determining a set of distribution network equipment status description distributions through a first status description mining subnet covered by the distribution network equipment abnormal identification network; wherein, the set of distribution network equipment status description distributions includes several distribution network equipment status descriptions; combining the set of defect environment operation status information, and determining a set of defect environment status description distributions through a second status description mining subnet covered by the distribution network equipment abnormal identification network; wherein, the set of defect environment status description distributions includes several defect environment status descriptions; combining the set of distribution network equipment status description distributions and the set of defect environment status description distributions, and determining the quantization possibility of the defect theme pointed to by the set of distribution network equipment operation status information through an equipment defect identification subnet covered by the distribution network equipment abnormal identification network; using the quantization possibility of the defect theme to determine the abnormal identification situation of the set of distribution network equipment operation status information.

[0005] For some preferred embodiments, combining the distribution set of the distribution network device status descriptions and the distribution set of the defect environment status descriptions, determining the quantization possibility of the defect theme pointed to by the distribution network device operation status information set through the device defect identification subnet included in the distribution network device anomaly identification network, includes: combining the distribution set of the distribution network device status descriptions, and determining a plurality of first array-type device status contents through the first device line structure local attention subnet included in the distribution network device anomaly identification network; wherein each first array-type device status content is matched with a distribution network device status description distribution; combining the distribution set of the defect environment status descriptions, and determining a plurality of second array-type device status contents through the second device line structure local attention subnet included in the distribution network device anomaly identification network; wherein each second array-type device status content is matched with a defect environment status description distribution; performing a merging operation on the plurality of first array-type device status contents and the plurality of second array-type device status contents to obtain a plurality of target array-type device status contents; wherein each target array-type device status content covers a first array-type device status content and a second array-type device status content; combining the plurality of target array-type device status contents, and determining the quantization possibility of the defect theme pointed to by the distribution network device operation status information set through the device defect identification subnet included in the distribution network device anomaly identification network.

[0006] For some preferred embodiments, by combining the distribution set of the distribution network equipment status descriptions, a plurality of first array-type equipment status contents are determined through the first equipment line structure local attention subnet included in the distribution network equipment anomaly recognition network, including: for each distribution network equipment status description distribution in the distribution set of the distribution network equipment status descriptions, a first individual refined status description distribution is determined through the individual description refinement model included in the first equipment line structure local attention subnet; wherein, the first equipment line structure local attention subnet belongs to the distribution network equipment anomaly recognition network; for each distribution network equipment status description distribution in the distribution set of the distribution network equipment status descriptions, a first group refined status description distribution is determined through the group description refinement model included in the first equipment line structure local attention subnet; for each distribution network equipment status description distribution in the distribution set of the distribution network equipment status descriptions, by combining the first individual refined status description distribution and the first group refined status description distribution, a first globalized status description distribution is determined through the moving average processing model included in the first equipment line structure local attention subnet; for each distribution network equipment status description distribution in the distribution set of the distribution network equipment status descriptions, by combining the first globalized status description distribution and the distribution network equipment status description distribution, a first array-type equipment status content is determined through the first group description refinement model included in the first equipment line structure local attention subnet.

[0007] For some preferred embodiments, the combination of the defect environment state description distribution sets determines a number of second array-type device state contents through the second device line structure partial attention subnet included in the distribution network device anomaly recognition network, including: for each defect environment state description distribution in the defect environment state description distribution sets, determining a second individual refined state description distribution through the individual description refinement model included in the second device line structure partial attention subnet; wherein, the second device line structure partial attention subnet belongs to the distribution network device anomaly recognition network; for each defect environment state description distribution in the defect environment state description distribution sets, determining a second group refined state description distribution through the group description refinement model included in the second device line structure partial attention subnet; for each defect environment state description distribution in the defect environment state description distribution sets, combining the second individual refined state description distribution and the second group refined state description distribution, and determining a second globalized state description distribution through the moving average processing model included in the second device line structure partial attention subnet; for each defect environment state description distribution in the defect environment state description distribution sets, combining the second globalized state description distribution and the defect environment state description distribution, and determining the second array-type device state content through the second group description refinement model included in the second device line structure partial attention subnet.

[0008] For some preferred embodiments, the combination of the several target array-type device state contents determines the defect topic quantification possibility pointed to by the distribution network device operation state information set through the device defect recognition subnet included in the distribution network device anomaly recognition network, including: combining the several target array-type device state contents, and determining the globalized array-type device state content through the device real-time demand partial attention subnet included in the distribution network device anomaly recognition network; wherein, the globalized array-type device state content is determined by using the several target array-type device state contents and several time dimension importance indexes, and each target array-type device state content is matched with a time dimension importance index; combining the globalized array-type device state content, and determining the defect topic quantification possibility pointed to by the distribution network device operation state information set through the device defect recognition subnet included in the distribution network device anomaly recognition network.

[0009] For some preferred embodiments, combining the several target array-type device status contents to determine the globalized array-type device status content through the device real-time demand local attention subnet included in the distribution network device anomaly recognition network includes: combining the several target array-type device status contents to determine several first local array-type device status contents through the first local attention model included in the device real-time demand local attention subnet; wherein, the device real-time demand local attention subnet belongs to the distribution network device anomaly recognition network; combining the several first local array-type device status contents to determine several second local array-type device status contents through the second local attention model included in the device real-time demand local attention subnet; using the several second local array-type device status contents to determine several time dimension importance indexes; wherein, each time dimension importance index is matched with a target array-type device status content; using the several target array-type device status contents and the several time dimension importance indexes to determine the globalized array-type device status content.

[0010] For some preferred embodiments, combining the distribution set of the distribution network equipment status descriptions and the distribution set of the defect environment status descriptions, determining the quantization possibility of the defect theme pointed to by the distribution network equipment operation status information set through the equipment defect identification subnet included in the distribution network equipment anomaly identification network, includes: combining the distribution set of the distribution network equipment status descriptions, and determining a number of first array-type equipment status contents through the first group description reduction model included in the distribution network equipment anomaly identification network; wherein, each first array-type equipment status content is matched with a distribution network equipment status description distribution; combining the distribution set of the defect environment status descriptions, and determining a number of second array-type equipment status contents through the second group description reduction model included in the distribution network equipment anomaly identification network; wherein, each second array-type equipment status content is matched with a defect environment status description distribution; performing a merging operation on the number of first array-type equipment status contents and the number of second array-type equipment status contents to obtain a number of target array-type equipment status contents; wherein, each target array-type equipment status content covers a first array-type equipment status content and a second array-type equipment status content; combining the number of target array-type equipment status contents, and determining a globalized array-type equipment status content through the equipment real-time demand local attention subnet included in the distribution network equipment anomaly identification network; wherein, the globalized array-type equipment status content is determined by using the number of target array-type equipment status contents and a number of time dimension importance indexes, and each target array-type equipment status content is matched with a time dimension importance index; combining the globalized array-type equipment status content, and determining the quantization possibility of the defect theme pointed to by the distribution network equipment operation status information set through the equipment defect identification subnet included in the distribution network equipment anomaly identification network.

[0011] For some preferred embodiments, using the distribution network equipment operation status information set to determine the defect environment operation status information set, includes: for each distribution network equipment operation status information in the distribution network equipment operation status information set, determining a first defect environment information set, a second defect environment information set, and a third defect environment information set through the operation status information optimization network; using the first defect environment information set, the second defect environment information set, and the third defect environment information set pointed to by each distribution network equipment operation status information to determine the defect environment operation status information pointed to by each distribution network equipment operation status information.

[0012] The embodiment of the present application further provides a distribution network equipment anomaly identification system, which is characterized by including a processor, a network module, and a memory; the processor communicates with the memory through the network module, and the processor reads and runs a computer program from the memory to execute the above method.

[0013] The embodiments of the present application also provide a computer-readable storage medium, on which a computer program is stored, and the computer program implements the above method when running.

[0014] The embodiments of the present application also provide a computer storage medium, which stores a computer program, and the computer program implements the above method when running.

[0015] Applied to the embodiments of the present application, first, a set of operation state information of distribution network devices is collected, then a set of operation state information of the defect environment is determined according to the set of operation state information of distribution network devices. Based on the set of operation state information of distribution network devices, a set of distribution of state descriptions of distribution network devices is determined through a first state description mining subnet included in the distribution network device anomaly recognition network. And based on the set of operation state information of the defect environment, a set of distribution of state descriptions of the defect environment is determined through a second state description mining subnet included in the distribution network device anomaly recognition network. Then, based on the set of distribution of state descriptions of distribution network devices and the set of distribution of state descriptions of the defect environment, a quantization possibility of a defect theme pointed to by the operation state information of distribution network devices is determined through a device defect recognition subnet included in the distribution network device anomaly recognition network. Subsequently, the anomaly recognition situation of the set of operation state information of distribution network devices is determined according to the quantization possibility of the defect theme.

[0016] Through the above technical solution, after obtaining the set of operation state information of distribution network devices, the set of operation state information of the defect environment can be further determined. In this way, different operation state information can be specifically mined and recognized through different state description mining subnets to accurately and completely obtain different distributions of state descriptions. Further, by integrating different distributions of state descriptions and invoking the device defect recognition subnet, the quantization possibility of the defect theme pointed to by the operation state information of distribution network devices can be accurately located. In this way, the anomaly recognition situation corresponding to the set of operation state information of distribution network devices can be accurately located according to the quantization possibility of the defect theme, so as to analyze in combination with the device operation level and the defect environment level to improve the accuracy and credibility of the anomaly recognition of distribution network devices.

[0017] In the following description, some other features will be partly stated. When examining the following content and the drawings, those skilled in the art will partly discover these features, or these features can be learned through production or application. Through practicing or using various aspects of the methods, tools, and combinations listed in the detailed examples described later, the features in the current application can be implemented and obtained. Description of the Drawings

[0018] To more clearly illustrate the technical solutions of the embodiments of the present application, the accompanying drawings required for the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present application and should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can be obtained based on these drawings.

[0019] Figure 1 It is a block diagram of a distribution network equipment anomaly recognition system provided by an embodiment of the present application.

[0020] Figure 2 It is a flowchart of a distribution network equipment anomaly recognition method provided by an embodiment of the present application. Detailed implementation manners

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all of them. Usually, the components of the embodiments of the present application described and illustrated in the drawings here can be arranged and designed in various different configurations.

[0022] Therefore, the detailed description of the embodiments of the present application provided in the drawings below is not intended to limit the scope of the claimed present application, but merely represents the selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present application.

[0023] It should be noted that similar reference numerals and letters indicate similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0024] Figure 1 It shows a block diagram of a distribution network equipment anomaly recognition system 10 provided by an embodiment of the present application. The distribution network equipment anomaly recognition system 10 in the embodiments of the present application can be a server with data storage, transmission, and processing functions, such as Figure 1 As shown, the distribution network equipment anomaly recognition system 10 includes: a memory 11, a processor 12, a network module 13, and a distribution network equipment anomaly recognition device 20.

[0025] The memory 11, the processor 12, and the network module 13 are electrically connected directly or indirectly to achieve data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines. The memory 11 stores the distribution network equipment anomaly recognition device 20. The distribution network equipment anomaly recognition device 20 includes at least one software function module that can be stored in the memory 11 in the form of software or firmware. The processor 12 executes various functional applications and data processing by running the software programs and modules stored in the memory 11, such as the distribution network equipment anomaly recognition device 20 in the embodiments of the present application, thereby implementing the distribution network equipment anomaly recognition method in the embodiments of the present application.

[0026] Among them, the memory 11 can be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), etc. Among them, the memory 11 is used to store programs, and the processor 12 executes the programs after receiving execution instructions.

[0027] The processor 12 may be an integrated circuit chip with data processing capabilities. The above-mentioned processor 12 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0028] The network module 13 is used to establish a communication connection between the distribution network equipment anomaly recognition system 10 and other communication terminal devices through the network to achieve the transceiver operation of network signals and data. The above-mentioned network signals can include wireless signals or wired signals.

[0029] It can be understood that Figure 1 The structure shown is only schematic. The distribution network equipment anomaly recognition system 10 may also include more or fewer components than those Figure 1 shown, or have a different configuration from that Figure 1 shown.Figure 1 Each component shown can be implemented by hardware, software, or a combination thereof.

[0030] The embodiments of the present application also provide a computer storage medium storing a computer program which, when running, implements the above method.

[0031] Figure 2 The flowchart of a method for identifying anomalies in distribution network equipment provided by the embodiments of the present application is shown. The method steps defined by the relevant processes are applied to the distribution network equipment anomaly identification system 10 and can be implemented by the processor 12. The method includes the content described in the following steps.

[0032] Step 21, collect the set of operating state information of the distribution network equipment.

[0033] In the embodiments of the present application, the set of operating state information of the distribution network equipment includes several groups of operating state information of the distribution network equipment (including but not limited to voltage state information, current state information, power consumption state information, etc., meter state information, transformer state information, etc.) that have a sequential order (continuous in time).

[0034] Step 22, use the set of operating state information of the distribution network equipment to determine the set of operating state information of the defective environment.

[0035] In the embodiments of the present application, the set of operating state information of the defective environment includes several groups of operating state information of the defective environment (including but not limited to equipment damage state information, equipment contamination state information, ambient temperature state information, ambient humidity state information, etc.) that have a sequential order.

[0036] In some possible examples, the use of the set of operating state information of the distribution network equipment to determine the set of operating state information of the defective environment described in step 22 can be implemented through step 221 and step 222.

[0037] Step 221, for each piece of operating state information of the distribution network equipment in the set of operating state information of the distribution network equipment, determine the first defective environment information set, the second defective environment information set, and the third defective environment information set through the operating state information optimization network.

[0038] In the embodiments of the present application, the operating state information optimization network is used to filter noise information, and the importance indices (ranging from 0 to 1) of the first defective environment information set, the second defective environment information set, and the third defective environment information set are different.

[0039] Step 222: Using the first defect environment information set, the second defect environment information set, and the third defect environment information set pointed to by each piece of distribution network equipment operation status information, determine the defect environment operation status information pointed to by each piece of distribution network equipment operation status information.

[0040] Designed in this way, by implementing Step 221 - Step 222, it is possible to determine the defect environment information set based on different importance indices, thereby ensuring the integrity and accuracy of the defect environment operation status information.

[0041] Step 23: Combining the set of distribution network equipment operation status information, determine the set of distribution network equipment status description distributions through the first status description mining subnet (which can be understood as a feature extraction subnet or a feature extraction module, etc.) covered by the distribution network equipment anomaly recognition network (such as a neural network model, including but not limited to CNN or LSTM, etc.).

[0042] In the embodiment of the present application, the set of distribution network equipment status description distributions includes several distribution network equipment status description distributions (which can be understood as the status feature maps of the distribution network equipment).

[0043] Step 24: Combining the set of defect environment operation status information, determine the set of defect environment status description distributions through the second status description mining subnet covered by the distribution network equipment anomaly recognition network.

[0044] In the embodiment of the present application, the set of defect environment status description distributions includes several defect environment status description distributions (which can be understood as the status feature maps of the defect environment / anomaly environment).

[0045] Step 25: Combining the set of distribution network equipment status description distributions and the set of defect environment status description distributions, determine the defect theme quantization possibility (this defect theme quantization possibility can be understood as the defect theme probability, for example, the defect theme quantization possibility A_0.6 can represent that the recognition probability of the defect theme "insulator damage" pointed to by the distribution network equipment operation status information is 0.6) pointed to by the set of distribution network equipment operation status information through the equipment defect recognition subnet (this equipment defect recognition subnet can be understood as a classification subnet, including but not limited to classifiers, support vector machines, etc.) covered by the distribution network equipment anomaly recognition network.

[0046] In some possible embodiments, the combination of the distribution set of the distribution network equipment state descriptions and the distribution set of the defect environment state descriptions described in step 25, and determining the quantization possibility of the defect theme pointed to by the distribution network equipment operation state information set through the equipment defect identification subnet included in the distribution network equipment anomaly identification network can be implemented through the technical solutions described in steps 251 to 254.

[0047] Step 251: Combine the distribution set of the distribution network equipment state descriptions, and determine a number of first array-type equipment state contents through the first equipment line structure local attention subnet included in the distribution network equipment anomaly identification network.

[0048] In the embodiments of the present application, each first array-type equipment state content (which can be understood as a feature vector, for example) is matched with a distribution network equipment state description distribution. Further, the equipment line structure local attention subnet can be understood as a spatial attention model or a spatial domain attention model.

[0049] Under some independently implementable design ideas, the combination of the distribution set of the distribution network equipment state descriptions described in step 251 and determining a number of first array-type equipment state contents through the first equipment line structure local attention subnet included in the distribution network equipment anomaly identification network may include the following: for each distribution network equipment state description distribution in the distribution set of the distribution network equipment state descriptions, determine the first individual refined state description distribution through the individual description refinement model included in the first equipment line structure local attention subnet; wherein, the first equipment line structure local attention subnet belongs to the distribution network equipment anomaly identification network; for each distribution network equipment state description distribution in the distribution set of the distribution network equipment state descriptions, determine the first group refined state description distribution through the group description refinement model included in the first equipment line structure local attention subnet; for each distribution network equipment state description distribution in the distribution set of the distribution network equipment state descriptions, combine the first individual refined state description distribution and the first group refined state description distribution, and determine the first global state description distribution through the moving average processing model included in the first equipment line structure local attention subnet; for each distribution network equipment state description distribution in the distribution set of the distribution network equipment state descriptions, combine the first global state description distribution and the distribution network equipment state description distribution, and determine the first array-type equipment state content through the first group description refinement model included in the first equipment line structure local attention subnet.

[0050] In the embodiments of the present application, the individual simplification corresponds to the max pooling process, and the group simplification corresponds to the average pooling process. Similar understandings can be made for the related simplification models, which will not be elaborated here. With such a design, the accuracy and feature recognition of the first array-type device status content can be ensured.

[0051] Step 252: Combine the defective environment state description distribution set, and determine a number of second array-type device status contents through the second device line structure local attention subnet covered by the distribution network device anomaly recognition network.

[0052] In the embodiments of the present application, each second array-type device status content is matched with a defective environment state description distribution.

[0053] For some independently implementable design ideas, the

[0054] Step 253: Perform a merging operation on the number of first array-type device status contents and the number of second array-type device status contents to obtain a number of target array-type device status contents.

[0055] In the embodiments of the present application, each target array-type device status content covers a first array-type device status content and a second array-type device status content. In addition, the merging operation can be understood as a splicing operation.

[0056] Step 254: Combine the number of target array-type device status contents, and determine the quantization possibility of the defect theme pointed to by the distribution network device operation status information set through the device defect recognition subnet covered by the distribution network device anomaly recognition network.

[0057] For some independently implementable design ideas, the combination of the number of target array-type device status contents described in Step 254 to determine the quantization possibility of the defect theme pointed to by the distribution network device operation status information set through the device defect recognition subnet covered by the distribution network device anomaly recognition network can be achieved through the following technical solution: Combine the number of target array-type device status contents, and determine the global array-type device status content through the device real-time demand local attention subnet covered by the distribution network device anomaly recognition network; wherein, the global array-type device status content is determined by using the number of target array-type device status contents and a number of time dimension importance indexes, and each target array-type device status content is matched with a time dimension importance index; Combine the global array-type device status content, and determine the quantization possibility of the defect theme pointed to by the distribution network device operation status information set through the device defect recognition subnet covered by the distribution network device anomaly recognition network.

[0058] In the embodiments of the present application, the device real-time demand local attention subnet can be understood as a time attention model or a time dimension attention model, and the globalized array-type device state content can be understood as a fused state feature vector, and the time dimension importance index can be understood as a time series weight. Thus, the quantization possibility of the defect theme can be determined based on the timeliness level, so as to ensure that the anomaly recognition will not be delayed as much as possible.

[0059] In some other possible embodiments, combining the above-mentioned several target array-type device state contents and determining the globalized array-type device state content through the device real-time demand local attention subnet covered by the distribution network device anomaly recognition network described in the above steps may include the following contents: combining the above-mentioned several target array-type device state contents and determining several first local array-type device state contents through the first local attention model covered by the device real-time demand local attention subnet; wherein, the device real-time demand local attention subnet belongs to the distribution network device anomaly recognition network; combining the above-mentioned several first local array-type device state contents and determining several second local array-type device state contents through the second local attention model covered by the device real-time demand local attention subnet; determining several time dimension importance indexes by using the above-mentioned several second local array-type device state contents; wherein, each time dimension importance index is matched with a target array-type device state content; determining the globalized array-type device state content by using the above-mentioned several target array-type device state contents and several time dimension importance indexes.

[0060] With such a design, through the divide-and-conquer idea and phased processing, the globalized array-type device state content can be determined completely and quickly.

[0061] It can be understood that through step 251-step 254, the corresponding feature vector can be determined by means of the device line structure local attention subnet, and the quantization possibility of the defect theme can be accurately determined based on the fusion of the feature vectors.

[0062] Under some independently implementable design concepts, the combination of the distribution set of the power distribution network equipment state description and the distribution set of the defect environment state description, and the determination of the quantization possibility of the defect theme pointed to by the power distribution network equipment operation state information set through the equipment defect identification subnet covered by the power distribution network equipment anomaly identification network can be achieved through the following technical solutions: Combine the distribution set of the power distribution network equipment state description, and determine a number of first array-type equipment state contents through the first group description reduction model covered by the power distribution network equipment anomaly identification network; wherein, each first array-type equipment state content is matched with a power distribution network equipment state description distribution; Combine the distribution set of the defect environment state description, and determine a number of second array-type equipment state contents through the second group description reduction model covered by the power distribution network equipment anomaly identification network; wherein, each second array-type equipment state content is matched with a defect environment state description distribution; Perform a merging operation on the number of first array-type equipment state contents and the number of second array-type equipment state contents to obtain a number of target array-type equipment state contents; wherein, each target array-type equipment state content covers a first array-type equipment state content and a second array-type equipment state content; Combine the number of target array-type equipment state contents, and determine the globalized array-type equipment state content through the equipment real-time demand local attention subnet covered by the power distribution network equipment anomaly identification network; wherein, the globalized array-type equipment state content is determined by using the number of target array-type equipment state contents and a number of time dimension importance indexes, and each target array-type equipment state content is matched with a time dimension importance index; Combine the globalized array-type equipment state content, and determine the quantization possibility of the defect theme pointed to by the power distribution network equipment operation state information set through the equipment defect identification subnet covered by the power distribution network equipment anomaly identification network.

[0063] Step 26, use the quantization possibility of the defect theme to determine the anomaly identification situation of the power distribution network equipment operation state information set.

[0064] In the embodiment of the present application, the anomaly identification situation can be determined by comparing the quantization possibility of the defect theme with a set probability. For example, if the probability 0.9 corresponding to the quantization possibility of the defect theme B_0.9 is greater than the set probability 0.8, it can be determined that the anomaly identification situation of the power distribution network equipment operation state information set is the defect theme "transformer breather defect" corresponding to the quantization possibility of the defect theme B_0.9.

[0065] Exemplarily, the anomaly identification of the equipment body defects in the present application mainly includes: transformer, CT / PT oil leakage identification, meter breakage identification, insulator breakage identification, transformer breather defect identification, equipment surface pollution identification, etc., but not limited thereto.

[0066] In summary, when applied to the embodiments of the present application, first, a set of operation status information of distribution network equipment is collected, then a set of operation status information of the defect environment is determined according to the set of operation status information of the distribution network equipment. Based on the set of operation status information of the distribution network equipment, a set of distribution of device status descriptions of the distribution network equipment is determined through the first status description mining subnet included in the distribution network equipment anomaly recognition network. And based on the set of operation status information of the defect environment, a set of distribution of defect environment status descriptions is determined through the second status description mining subnet included in the distribution network equipment anomaly recognition network. Then, based on the set of distribution of device status descriptions of the distribution network equipment and the set of distribution of defect environment status descriptions, the quantization possibility of the defect theme pointed to by the operation status information of the distribution network equipment is determined through the device defect recognition subnet included in the distribution network equipment anomaly recognition network. Subsequently, the anomaly recognition situation of the set of operation status information of the distribution network equipment is determined according to the quantization possibility of the defect theme.

[0067] Through the above technical solution, after obtaining the set of operation status information of the distribution network equipment, the set of operation status information of the defect environment can be further determined. In this way, different operation status information can be targeted mined and recognized through different status description mining subnets to accurately and completely obtain different distributions of status descriptions. Further, by integrating different distributions of status descriptions and invoking the device defect recognition subnet, the quantization possibility of the defect theme pointed to by the operation status information of the distribution network equipment can be accurately located. In this way, the anomaly recognition situation corresponding to the set of operation status information of the distribution network equipment can be accurately located according to the quantization possibility of the defect theme, so as to analyze from both the device operation level and the defect environment level to improve the accuracy and credibility of the anomaly recognition of the distribution network equipment.

[0068] In several embodiments provided in the embodiments of the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device and method embodiments described above are only illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architectures, functions, and operations of devices, methods, and computer program products according to multiple embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0069] In addition, each functional module in various embodiments of the present application may be integrated together to form an independent part, or each module may exist alone, or two or more modules may be integrated to form an independent part.

[0070] If the above-mentioned function is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, the abnormal recognition system 10 of a distribution network device, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes. It should be noted that in this article, the term "including", "comprising", or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article, or device including a series of elements not only includes those elements but also includes other elements not explicitly listed, or further includes elements inherent to such a process, method, article, or device. Without further limitations, an element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, method, article, or device including the said element.

[0071] The foregoing are only the preferred embodiments of the present application and are not used to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. A method for identifying anomalies in distribution network equipment, characterized in that, Applied to a distribution network equipment anomaly recognition system, the method at least includes: Collect a set of operation status information of distribution network equipment; wherein, the set of operation status information of distribution network equipment contains several groups of operation status information of distribution network equipment with a sequential order; Use the set of operation status information of distribution network equipment to determine a set of operation status information of defective environment; wherein, the set of operation status information of defective environment contains several groups of operation status information of defective environment with a sequential order; Combined with the set of operation status information of distribution network equipment, determine a set of distribution of equipment status descriptions of distribution network equipment through a first status description mining subnet covered by the distribution network equipment anomaly recognition network; wherein, the set of distribution of equipment status descriptions of distribution network equipment includes several distributions of equipment status descriptions of distribution network equipment; Combined with the set of operation status information of defective environment, determine a set of distribution of status descriptions of defective environment through a second status description mining subnet covered by the distribution network equipment anomaly recognition network; wherein, the set of distribution of status descriptions of defective environment includes several distributions of status descriptions of defective environment; Combined with the set of distribution of equipment status descriptions of distribution network equipment and the set of distribution of status descriptions of defective environment, determine the quantification possibility of defective topics pointed to by the set of operation status information of distribution network equipment through an equipment defect recognition subnet covered by the distribution network equipment anomaly recognition network; Use the quantification possibility of defective topics to determine the anomaly recognition situation of the set of operation status information of distribution network equipment; The step of combining the set of distribution of equipment status descriptions of distribution network equipment and the set of distribution of status descriptions of defective environment, and determining the quantification possibility of defective topics pointed to by the set of operation status information of distribution network equipment through an equipment defect recognition subnet covered by the distribution network equipment anomaly recognition network includes: Combined with the set of distribution of equipment status descriptions of distribution network equipment, determine several first array-type equipment status contents through a first local attention subnet of equipment line structure covered by the distribution network equipment anomaly recognition network; wherein, each first array-type equipment status content matches a distribution of equipment status descriptions of distribution network equipment; Combined with the set of distribution of status descriptions of defective environment, determine several second array-type equipment status contents through a second local attention subnet of equipment line structure covered by the distribution network equipment anomaly recognition network; wherein, each second array-type equipment status content matches a distribution of status descriptions of defective environment; Perform a merging operation on the several first array-type equipment status contents and the several second array-type equipment status contents to obtain several target array-type equipment status contents; wherein, each target array-type equipment status content covers a first array-type equipment status content and a second array-type equipment status content; Combined with the several target array-type equipment status contents, determine the quantification possibility of defective topics pointed to by the set of operation status information of distribution network equipment through the equipment defect recognition subnet covered by the distribution network equipment anomaly recognition network.

2. The method according to claim 1, wherein Combined with the distribution set of the distribution network equipment status descriptions, a number of first array-type equipment status contents are determined through the first equipment line structure local attention subnet covered by the distribution network equipment anomaly recognition network, including: For each distribution network equipment status description distribution in the distribution set of the distribution network equipment status descriptions, a first individualized simplified status description distribution is determined through the individualized description simplification model covered by the first equipment line structure local attention subnet; wherein, the first equipment line structure local attention subnet belongs to the distribution network equipment anomaly recognition network; For each distribution network equipment status description distribution in the distribution set of the distribution network equipment status descriptions, a first group simplified status description distribution is determined through the group description simplification model covered by the first equipment line structure local attention subnet; For each distribution network equipment status description distribution in the distribution set of the distribution network equipment status descriptions, combined with the first individualized simplified status description distribution and the first group simplified status description distribution, a first globalized status description distribution is determined through the moving average processing model covered by the first equipment line structure local attention subnet; For each distribution network equipment status description distribution in the distribution set of the distribution network equipment status descriptions, combined with the first globalized status description distribution and the distribution network equipment status description distribution, a first array-type equipment status content is determined through the first group description simplification model covered by the first equipment line structure local attention subnet.

3. The method according to claim 1, wherein Combined with the distribution set of the defect environment status descriptions, a number of second array-type equipment status contents are determined through the second equipment line structure local attention subnet covered by the distribution network equipment anomaly recognition network, including: For each defect environment status description distribution in the distribution set of the defect environment status descriptions, a second individualized simplified status description distribution is determined through the individualized description simplification model covered by the second equipment line structure local attention subnet; wherein, the second equipment line structure local attention subnet belongs to the distribution network equipment anomaly recognition network; For each defect environment status description distribution in the distribution set of the defect environment status descriptions, a second group simplified status description distribution is determined through the group description simplification model covered by the second equipment line structure local attention subnet; For each defect environment status description distribution in the distribution set of the defect environment status descriptions, combined with the second individualized simplified status description distribution and the second group simplified status description distribution, a second globalized status description distribution is determined through the moving average processing model covered by the second equipment line structure local attention subnet; For each defect environment status description distribution in the distribution set of the defect environment status descriptions, combined with the second globalized status description distribution and the defect environment status description distribution, a second array-type equipment status content is determined through the second group description simplification model covered by the second equipment line structure local attention subnet.

4. The method according to claim 1, wherein Combining the above-mentioned several target array-type device status contents, determining the quantization possibility of the defect theme pointed to by the set of distribution network device operation status information through the device defect identification subnet included in the distribution network device anomaly identification network, includes: Combining the above-mentioned several target array-type device status contents, determining the globalized array-type device status content through the device real-time demand local attention subnet included in the distribution network device anomaly identification network; wherein, the globalized array-type device status content is determined by using the above-mentioned several target array-type device status contents and several time dimension importance indexes, and each target array-type device status content is matched with a time dimension importance index; Combining the globalized array-type device status content, determining the quantization possibility of the defect theme pointed to by the set of distribution network device operation status information through the device defect identification subnet included in the distribution network device anomaly identification network.

5. The method according to claim 4, characterized in that, The step of combining the above-mentioned several target array-type device status contents and determining the globalized array-type device status content through the device real-time demand local attention subnet included in the distribution network device anomaly identification network, includes: Combining the above-mentioned several target array-type device status contents, determining several first local array-type device status contents through the first local attention model included in the device real-time demand local attention subnet; wherein, the device real-time demand local attention subnet belongs to the distribution network device anomaly identification network; Combining the above-mentioned several first local array-type device status contents, determining several second local array-type device status contents through the second local attention model included in the device real-time demand local attention subnet; Determining several time dimension importance indexes by using the above-mentioned several second local array-type device status contents; wherein, each time dimension importance index is matched with a target array-type device status content; Determining the globalized array-type device status content by using the above-mentioned several target array-type device status contents and several time dimension importance indexes.

6. The method according to claim 1, wherein The step of combining the set of distribution network device status description distributions and the set of defect environment status description distributions, and determining the quantization possibility of the defect theme pointed to by the set of distribution network device operation status information through the device defect identification subnet included in the distribution network device anomaly identification network, includes: Combining the set of distribution network device status description distributions, determining several first array-type device status contents through the first group description reduction model included in the distribution network device anomaly identification network; wherein, each first array-type device status content is matched with a distribution network device status description distribution; Combining the set of defect environment status description distributions, determining several second array-type device status contents through the second group description reduction model included in the distribution network device anomaly identification network; wherein, each second array-type device status content is matched with a defect environment status description distribution; Perform a merging operation on the several first array-type device status contents and the several second array-type device status contents to obtain several target array-type device status contents; wherein each target array-type device status content covers a first array-type device status content and a second array-type device status content; Combine the several target array-type device status contents, and determine a globalized array-type device status content through the device real-time demand local attention subnet included in the distribution network device anomaly recognition network; wherein the globalized array-type device status content is determined by using the several target array-type device status contents and several time dimension importance indexes, and each target array-type device status content is matched with a time dimension importance index; Combine the globalized array-type device status content, and determine the quantization possibility of the defect theme pointed to by the distribution network device operation status information set through the device defect recognition subnet included in the distribution network device anomaly recognition network.

7. The method according to any one of claims 1 to 6, characterized in that, The determining the defect environment operation status information set by using the distribution network device operation status information set includes: For each distribution network device operation status information in the distribution network device operation status information set, determine a first defect environment information set, a second defect environment information set, and a third defect environment information set through the operation status information optimization network; Use the first defect environment information set, the second defect environment information set, and the third defect environment information set pointed to by each distribution network device operation status information to determine the defect environment operation status information pointed to by each distribution network device operation status information.

8. An abnormal recognition system for distribution network equipment, characterized in that, It includes a processor, a network module, and a memory; the processor communicates with the memory through the network module, and the processor reads and runs a computer program from the memory to execute the method according to any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, A computer program is stored thereon, and the computer program realizes the method according to any one of claims 1-7 when running.

Images