Power distribution network data topology restoration method and device, electronic equipment and storage medium

Abstract

The invention provides a power distribution network data topology restoration method and device, electronic equipment and a storage medium, and relates to the technical field of power systems. The method comprises the steps that a first participation device obtains multiple pieces of data flow information among services involved in a first service domain of the power distribution network within a preset time period; extracting address information of each service from the plurality of pieces of data stream information, and determining a plurality of nodes and an initial node attribute of each node based on the address information of each service; determining a plurality of communication links based on the plurality of pieces of data flow information, determining a plurality of edges among the plurality of nodes based on the plurality of communication links, and determining an initial edge attribute of each edge based on the plurality of pieces of data flow information under each communication link to obtain an initial data topological graph of the first service domain; and receiving global node attributes and global edge attributes sent by the central coordination equipment, and updating the initial data topological graph based on the global node attributes and the global edge attributes to obtain a global data topological graph of the first service domain.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a method, device, electronic equipment and storage medium for restoring distribution network data topology. Background Art

[0002] With the development of smart grid technology and new power system services, the topological structure of distribution networks is becoming increasingly complex. The corresponding network topology description and expression are also playing an increasingly important role in distribution network business management, data analysis, security risk monitoring and identification, etc. However, the network topology in related technologies has the problem of low accuracy and cannot reflect the actual business operation of the distribution network. Summary of the invention

[0003] Embodiments of the present invention provide a distribution network data topology restoration method, device, electronic device and storage medium, which are used to solve the problem of low accuracy of distribution network topology.

[0004] In a first aspect, an embodiment of the present invention provides a method for restoring distribution network data topology, which is applied to a first participating device, and the method includes: obtaining multiple data flow information between various services involved in a first business domain of the distribution network within a preset time period; the first business domain is the business domain to which the first participating device belongs; extracting address information of each service from the multiple data flow information, and based on the address information of each service, determining multiple nodes and initial node attributes of each node; wherein each of the services is a node, and the initial node attributes include the address information of the service; determining multiple communication links based on the multiple data flow information, and determining multiple edges between the multiple nodes based on the multiple communication links, And based on the multiple data flow information under each of the communication links, the initial edge attributes of each edge are determined to obtain the initial data topology map of the first business domain; the initial node attributes and the initial edge attributes are sent to the central coordination device; wherein the central coordination device is used to perform cross-domain aggregation processing on the initial node attributes and initial edge attributes of each business domain based on the federated learning framework to obtain the global node attributes of each node and the global edge attributes of each edge; the global node attributes and the global edge attributes sent by the central coordination device are received, and the initial data topology map is updated based on the global node attributes and the global edge attributes to obtain the global data topology map of the first business domain.

[0005] In some embodiments, determining multiple communication links based on the multiple data flow information includes: determining multiple first communication links based on source address information and destination address information in the multiple data flow information; determining a stability index value of each of the first communication links based on the data flow information under each of the first communication links; and determining the multiple communication links from the multiple first communication links based on the stability index values.

[0006] In some embodiments, the stability index value includes a standard deviation of the number of flows and an average flow rate of the first communication link; determining the multiple communication links from the multiple first communication links based on the stability index value includes: determining the multiple first communication links whose standard deviation of the number of flows is less than or equal to a first threshold and whose average flow rate is greater than or equal to a second threshold as the multiple communication links.

[0007] In some embodiments, the initial edge attributes of each edge are determined based on the multiple data flow information under each of the communication links, including: for each of the communication links, based on the multiple data flow information under the communication link, the flow attributes of the multiple data flows under the communication link are determined; the flow attributes include the timestamp, traffic size and / or flow intensity of the data flow; the attributes of the data flows with overlapping timestamps under the communication link are merged to obtain the flow attributes of the multiple independent data flows under the communication link; the flow attributes of the multiple independent data flows under the communication link are determined as the initial edge attributes of the corresponding edge of the communication link.

[0008] In a second aspect, an embodiment of the present invention provides a distribution network data topology restoration method, which is applied to a central coordination device, and the method includes: receiving the initial node attributes of each node and the initial edge attributes of each edge in the initial data topology map of each business domain of the distribution network sent by multiple first participating devices; wherein the initial data topology map is constructed by the first participating device based on the data flow information between the various services involved in the business domain to which it belongs; based on the federated learning framework, cross-domain aggregation processing is performed on the initial node attributes and initial edge attributes in each business domain to obtain the global node attributes of each node in each business domain and the global edge attributes of each edge; the global node attributes of each node in each business domain and the global edge attributes of each edge are sent to the first participating device of the corresponding business domain, so that each of the first participating devices updates the initial data topology map based on the global node attributes and the global edge attributes to obtain the global data topology map of the corresponding business domain.

[0009] In a third aspect, an embodiment of the present invention provides a distribution network data topology restoration device, which is deployed on a first participating device, and the device includes: a first acquisition module, used to obtain multiple data flow information between various services involved in a first business domain of the distribution network within a preset time period; the first business domain is the business domain to which the first participating device belongs; a first determination module, used to extract address information of each service from the multiple data flow information, and based on the address information of each service, determine multiple nodes and initial node attributes of each of the nodes; wherein each of the services is a node, and the initial node attributes include the address information of the service; a second determination module, used to determine multiple communication links based on the multiple data flow information, and determine the multiple nodes based on the multiple communication links. A plurality of edges, and based on the multiple data flow information under each of the communication links, the initial edge attributes of each edge are determined to obtain the initial data topology map of the first business domain; a first sending module is used to send the initial node attributes and the initial edge attributes to the central coordination device; wherein the central coordination device is used to perform cross-domain aggregation processing on the initial node attributes and the initial edge attributes of each business domain based on the federated learning framework, to obtain the global node attributes of each node and the global edge attributes of each edge; a first receiving module is used to receive the global node attributes and the global edge attributes sent by the central coordination device, and update the initial data topology map based on the global node attributes and the global edge attributes to obtain the global data topology map of the first business domain.

[0010] In a fourth aspect, an embodiment of the present invention provides a distribution network data topology restoration device, which is deployed on a central coordination device, and the device includes: a second receiving module, used to receive the initial node attributes of each node and the initial edge attributes of each edge in the initial data topology map of each business domain of the distribution network sent by multiple first participating devices; wherein the initial data topology map is constructed by the first participating device based on the data flow information between the various services involved in the business domain to which it belongs; an aggregation module, used to perform cross-domain aggregation processing on the initial node attributes and initial edge attributes of each business domain based on a federated learning framework, and obtain the global node attributes of each node in each of the business domains and the global edge attributes of each edge; a second sending module, used to send the global node attributes of each node in each of the business domains and the global edge attributes of each edge to the first participating device of the corresponding business domain, so that each of the first participating devices updates the initial data topology map based on the global node attributes and the global edge attributes to obtain the global data topology map of the corresponding business domain.

[0011] In a fifth aspect, an embodiment of the present invention provides an electronic device, including a processor and a memory storing a computer program, wherein the processor implements the above-mentioned distribution network data topology restoration method when executing the program.

[0012] In a sixth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the above-mentioned distribution network data topology restoration method is implemented.

[0013] In a seventh aspect, an embodiment of the present invention provides a computer program product, including a computer program, which implements the above-mentioned distribution network data topology restoration method when executed by a processor.

[0014] The data topology restoration method, device, electronic device and storage medium provided by the embodiment of the present invention are as follows: the first participating device obtains multiple data flow information between various services involved in the first business domain of the distribution network within a preset time period; extracts the address information of each service from the multiple data flow information, and determines multiple nodes and the initial node attributes of each node based on the address information of each service; determines multiple communication links based on the multiple data flow information, determines multiple edges between multiple nodes based on the multiple communication links, and determines the initial edge attributes of each edge based on the multiple data flow information under each communication link, and obtains the initial data topology map of the first business domain; receives the global node attributes and global edge attributes sent by the central coordination device, and updates the initial data topology map based on the global node attributes and global edge attributes to obtain the global data topology map of the first business domain. The present invention performs data topology restoration through the data flow information between systems within the domain, and can reflect the actual business operation of each system based on the business flow information, so as to achieve high-precision intra-domain topology restoration. In addition, through the federated learning framework, the topology data of different business domains are cross-domain aggregated, which can improve the security and privacy of the data and achieve comprehensive and accurate restoration of the distribution network data. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0016] Figure 1 One of the flow charts of the method for restoring the distribution network data topology provided in an embodiment of the present invention; Figure 2 A second flow chart of a method for restoring distribution network data topology provided in an embodiment of the present invention; Figure 3 One of the structural schematic diagrams of the distribution network data topology restoration device provided in an embodiment of the present invention; Figure 4The second structural diagram of the distribution network data topology restoration device provided by the embodiment of the present invention; Figure 5 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] It should be noted that with the development of smart grid technology and new power system services, the topological structure of the distribution network is becoming increasingly complex, and the corresponding network topology description and expression are playing an increasingly important role in distribution network business management, data analysis, security risk monitoring and identification, etc. However, relying on network topology can no longer meet the actual needs of the above applications. It is also necessary to establish an accurate data topology to fully and detailedly display the connection relationship between each node, network traffic, business flow and other information, to assist the efficient development of new smart distribution network management and security applications.

[0019] With the expansion of the distribution network and the rapid growth of data volume, traditional centralized data processing methods are unable to efficiently cope with the real-time processing needs of large-scale data. Secondly, data privacy and security issues are becoming increasingly prominent, and centralized data processing models are prone to privacy leaks.

[0020] In order to solve the above problems, the present invention provides a distribution network data topology restoration method, device, electronic device and storage medium.

[0021] Figure 1 This is a flow chart of a method for restoring the topology of distribution network data provided by an embodiment of the present invention. It should be noted that the method for restoring the topology of distribution network data provided by an embodiment of the present invention is applied to the first participating device. Figure 1 As shown, the method may include the following steps.

[0022] Step 101: obtain multiple data flow information between various services involved in a first service domain of a distribution network within a preset time period; the first service domain is the service domain to which the first participating device belongs.

[0023] The preset time period may be a continuous time period including the current moment and the moment before the current moment, and the length of the preset time period may be determined based on actual needs, for example, may be 1 hour.

[0024] In some embodiments, the correspondence between the first service domain and the first participating device can be predetermined, and the first participating device is used to perform topological restoration on the distribution network data of the first service domain. That is, the distribution network can include multiple service domains, and the data topological restoration corresponding to each service domain is performed by different devices.

[0025] Among them, the business domain refers to the different businesses supported in the distribution network, such as electric vehicle business, distributed photovoltaic business, virtual battery business, etc. Each business is counted as a business domain, and each business involves different application services. The various services involved in the first business domain refer to the multiple application services involved in the operation of the first business. Multiple data flow information refers to the data flow information of the business interaction between the various services in the first business domain.

[0026] In some embodiments, each data flow information includes relevant information of a data flow, for example, each data flow information may include source service address information, target service address information, start time, end time, byte size of each data packet, protocol type, service identifier, etc. of the data flow. The source service address information and the target service address information may include IP address information and port number, respectively, and may also include MAC address information.

[0027] In some embodiments, multiple data stream information can be obtained through an API interface, a network scanning tool, etc., and the obtained data can be preprocessed to obtain multiple data stream information. Among them, the preprocessing may include: timestamp standardization processing, converting the timestamps in the data stream information into an international standard time format to eliminate the problem of inconsistent time formats between different systems; data cleaning, eliminating irrelevant information and noise data, as well as data that does not conform to the format specification; outlier detection, using statistical analysis methods such as Z-score to identify and process outliers in the data.

[0028] Step 102, extracting address information of each service from multiple data stream information, and determining multiple nodes and initial node attributes of each node based on the address information of each service; wherein each service is a node, and the initial node attributes include the address information of the service.

[0029] In some embodiments, the address information of each service may include IP address information, port number, MAC address information, etc. Since the service address information contained in the data flow information between different services may be different, the data flow information between some services includes IP address information and port number, and the data flow information between some services also includes MAC address information. Therefore, when extracting the address information of each service from multiple data flow information, the IP address information, port number, and MAC address information in each data flow information can be extracted to obtain the extracted address information of each service.

[0030] Since each pair of IP address information and port number can locate a service, when multiple nodes are determined based on the address information of each service, each pair of different IP address information and port number can be regarded as a node based on the extracted address information. Communication nodes with the same IP address information but different port numbers are regarded as different nodes.

[0031] In some embodiments, the address information of the service corresponding to each node can be directly determined as the initial node attribute of each node. The data flow information can also include information such as service identification and name, and the service identification and name of each node can also be used as the initial node attribute.

[0032] Step 103, determine multiple communication links based on multiple data flow information, determine multiple edges between multiple nodes based on the multiple communication links, and determine the initial edge attributes of each edge based on the multiple data flow information under each communication link to obtain an initial data topology graph of the first business domain.

[0033] The communication link is used to represent the channel for data interaction between services. Each communication link can be represented based on the source service address information and the destination service address information in the data flow information. Different pairs of source service address information and destination service address information serve as different communication links.

[0034] In some embodiments, in order to ensure the effectiveness of the communication link, the implementation process of determining multiple communication links based on multiple data flow information may include: determining multiple first communication links based on the source service address information and the destination service address information in the multiple data flow information; determining the stability index value of each first communication link based on the data flow information under each first communication link; and determining multiple communication links from the multiple first communication links based on the stability index values.

[0035] As an example, based on the source service address information and the destination service address information in each data flow information, a communication pair of {(source service IP address information, port number), (destination service IP address information, port number)} can be determined, thereby obtaining multiple different communication pairs of {(source service IP address information, port number), (destination service IP address information, port number)}, each communication pair represents a first communication link, and the same communication pair represents the same first communication link. The data flow information under each first communication link refers to the data flow information whose source service address information and destination service address information are consistent with the communication pair corresponding to the first communication link.

[0036] The stability index value of the first communication link is used to characterize the stability of the first communication link, for example, it may be the average flow of the first communication link within a preset time period, or it may be the standard deviation of the number of flows of the first communication link within a preset time period.

[0037] In some embodiments, the stability index value of the first communication link includes the standard deviation of the number of flows and the average flow of the first communication link; based on the stability index value, multiple communication links are determined from multiple first communication links, including: multiple first communication links whose standard deviation of the number of flows is less than or equal to a first threshold and whose average flow is greater than or equal to a second threshold are determined as multiple communication links.

[0038] Among them, the implementation process of determining the average flow of each first communication link based on the data flow information under each first communication link may include: for each first communication link, based on the data flow information under the first communication link, determining the flow size of each data flow under the first communication link, and determining the total flow size under the first communication link; dividing the total flow size by the length of the preset time period to obtain the average flow of the first communication link, as shown in the following formula (1).

[0039] (1); in, is the average flow of the i-th first communication link; is the duration of the preset time period; is the number of data flows under the i-th first communication link; is the flow size of the jth data flow under the ith first communication link. The flow size of each data flow may be the sum of sizes of all data packets contained in the data flow.

[0040] The implementation process of determining the standard deviation of the number of flows of each first communication link based on the data flow information under each first communication link may include: for each first communication link, based on the data flow information under the first communication link, determining the number of data flows at each observation time point under the first communication link, and determining the average number of flows of the first communication link; and determining the standard deviation of the number of flows of the first communication link according to the number of data flows at each observation time point and the average number of flows, as shown in the following formula (2).

[0041] (2); in, is the standard deviation of the number of flows of the i-th first communication link; is the number of data flows of the i-th first communication link at the j-th observation time point; N is the number of observation time points within the preset time period; is the number of data flows of the i-th first communication link at the q-th observation time point.

[0042] In some embodiments, the implementation process of determining multiple edges between multiple nodes based on multiple communication links includes: for each communication link, determining a first node corresponding to the source service address information of the communication link, determining a second node corresponding to the destination service address information of the communication link, and establishing an edge from the first node to the second node. In other words, each communication link represents an edge.

[0043] It should be noted that the initial data topology graph of the first service domain includes the multiple nodes and the multiple edges between the nodes determined in the above steps, as well as the initial node attributes of each node and the initial edge attributes of each edge.

[0044] In some embodiments, extracting address information of each service from multiple data stream information in step 102, and determining multiple nodes based on the address information of each service, and determining multiple communication links based on multiple data stream information in step 103 can all be implemented by a preset topological semantic model, which is used to traverse and extract data from multiple data stream information, obtain address information of each service, and determine multiple nodes based on the address information, and output a node set. The topological semantic model is also used to determine multiple communication pairs based on the extracted source service address information and port number, and the destination service address information and port number, and determine multiple communication links based on the stability index value of each communication pair.

[0045] In some embodiments, based on multiple data flow information under each communication link, the implementation process of determining the initial edge attribute of each edge may include the following steps.

[0046] Step S1, for each communication link, based on multiple data flow information under the communication link, determine the flow attributes of each of the multiple data flows under the communication link; the flow attributes include the timestamp, flow size and / or flow strength of the data flow.

[0047] The timestamp of the data flow may be the start time and the end time of the data flow. The flow size of the data flow may be the sum of the number of bytes of each data packet in the data flow. The flow intensity of the data flow may be the ratio of the flow size of the data flow to the delay of the data flow.

[0048] In some embodiments, the flow attributes of each data flow may also include a protocol type, the number of flows, and the like.

[0049] Step S2, merging the attributes of the data streams that overlap in time under the communication link to obtain the flow attributes of multiple independent data streams under the communication link.

[0050] In some embodiments, the timestamp of each data stream under the communication link can be compared with the timestamps of other data streams under the communication link in turn to determine whether there are multiple data streams with overlapping time under the communication link. , then the data stream Di and the data stream Dj overlap in time; is the start time of data stream Di, is the end time of the data stream Dj, is the start time of the data stream Dj, is the end time of data stream Di.

[0051] The process of merging the attributes of data stream Di and data stream Dj includes: merging data stream Di and data stream Dj into one data stream; integrating the timestamps of data stream Di and data stream Dj, and obtaining the timestamp of the merged data stream as ; Accumulate the flow sizes of data stream Di and data stream Dj to obtain the flow size of the merged data stream; determine the ratio of the flow size of the merged data stream to the duration corresponding to the merged timestamp as the flow intensity of the merged data stream; Integrate the protocol types of data stream Di and data stream Dj. If the protocol types of the two data streams are the same, the protocol type of the merged data stream remains unchanged. If the protocol types of the two data streams are different, the protocol types of the two data streams are both included in the protocol type of the merged data stream, that is, the protocol type of data stream Di is A and the protocol type of data stream Dj is B, then the protocol types of the merged data stream are A and B.

[0052] The independent data stream includes both the data stream obtained after merging and the data stream that does not overlap with other data streams in time among the multiple data stream information under the communication link. That is to say, the independent data stream obtained after merging refers to the data stream that does not overlap in time. If data stream D1 overlaps with data stream D2 in time, data stream D1 overlaps with data stream D3 in time, and data stream D3 overlaps with data stream D4 in time, then the data streams D1, D2, D3 and D4 can be merged to obtain a merged data stream and the stream attributes of the merged data stream.

[0053] Step S3, determining the flow attributes of multiple independent data flows under the communication link as the initial edge attributes of the corresponding edges of the communication link.

[0054] In some embodiments, the edge attributes of each edge can be initialized, and the initialized edge attributes of each edge are an empty list. For each edge corresponding to a communication link, the timestamp of each data stream under the communication link is compared with the timestamp of other data streams under the communication link in turn. If there is no time overlap between the data stream and the other data streams, the flow attributes of the data stream are written into the corresponding edge attribute list; if there is time overlap between the data stream and the other data streams, the attributes of the data stream and the data stream with which it has time overlap are merged, and the flow attributes of the merged data stream are written into the corresponding edge attribute list. The final edge attribute list is the initial edge attributes of the corresponding edge.

[0055] In some embodiments, the initial edge attributes may include flow attributes of multiple independent data flows, and each independent data flow may be distinguished by a data flow identifier.

[0056] Step 104, sending the initial node attributes and the initial edge attributes to the central coordination device; wherein the central coordination device is used to aggregate the initial node attributes and the initial edge attributes of each business domain based on the federated learning framework to obtain the global node attributes of each node and the global edge attributes of each edge.

[0057] In order to improve the privacy and security of original data between different participating devices, an embodiment of the present invention introduces a federated learning framework, in which a central coordinating device performs cross-domain topology splicing based on the initial node attributes and initial edge attributes of each business domain, and sends the obtained global node attributes of each node and the global edge attributes of each edge to the first participating device. The first participating device then updates the initial data topology map based on the global node attributes and the global edge attributes to obtain the global data topology map of the first business domain, so as to improve the accuracy and comprehensiveness of data topology restoration.

[0058] In some embodiments, the central coordination device can compare the IP address information and port numbers of nodes between different business domains based on the initial node attributes of different nodes in each business domain, and determine multiple shared node groups, each shared node group includes multiple shared nodes, and a shared node refers to a node that exists in multiple business domains, that is, a node with consistent IP address information and port numbers between different business domains. For example, the IP address information and port number of node 1 in business domain A are consistent with the IP address information and port number of node 2 in business domain B, then node 1 in business domain A and node 2 in business domain B are shared nodes; for each shared node group, the initial node attributes of the multiple shared nodes in the shared node group are taken as the union, and the result after taking the union is used as the global node attribute of each shared node in the shared node group; for other nodes that are not shared nodes, since these nodes do not exist across domains, there is no need to perform cross-domain aggregation on the node attributes of these nodes, and the initial node attributes of these nodes are directly used as global node attributes.

[0059] In some embodiments, the central coordination device can compare the communication links corresponding to each edge between different business domains based on the initial edge attributes of different edges in each business domain, and determine multiple shared edge groups, each shared edge group includes multiple shared edges, and a shared edge refers to an edge that exists in multiple business domains, that is, an edge between different business domains where the source service address information is consistent with the destination service address information, such as edge 1 from node 1 in business domain A to node 2, and edge 2 from node 3 to node 4 in business domain B. If the IP address information and port number of node 1 are consistent with the IP address information and port number of node 3, and the nodes The IP address information and port number of point 2 are also consistent with the IP address information and port number of node 4, then edge 1 in business domain A and edge 2 in business domain B are shared edges; for each shared edge group, the initial edge attributes of multiple shared edges in the shared edge group are taken as the union, and the time-overlapping attributes of the result obtained after taking the union are merged to obtain the merged edge attributes as the global edge attributes of each shared edge in the shared edge group; for other edges that are not shared edges, since these edges do not exist across domains, there is no need to perform cross-domain aggregation on the edge attributes of these edges, and the initial edge attributes of these edges are directly used as global edge attributes.

[0060] Step 105: Receive global node attributes and global edge attributes sent by the central coordination device, and update the initial data topology map based on the global node attributes and global edge attributes to obtain the global data topology map of the first service domain.

[0061] That is to say, the first participating device replaces the initial node attributes of each node in the initial data topology diagram of the first business domain with global node attributes, and replaces the initial edge attributes of each edge with global edge attributes, to obtain an updated global data topology diagram of the first business domain. The global data topology diagram of the first business domain can not only reflect the node connection and business flow under the business domain, but also reflect the business interaction relationship between cross-regions, so that the global data topology diagram of the first business domain can better reflect the actual business operation of each system and improve the accuracy and comprehensiveness of topology restoration.

[0062] According to the distribution network data topology restoration method of the embodiment of the present invention, the first participating device obtains multiple data flow information between various services involved in the first business domain of the distribution network within a preset time period; extracts the address information of each service from the multiple data flow information, and determines multiple nodes and the initial node attributes of each node based on the address information of each service; determines multiple communication links based on multiple data flow information, determines multiple edges between multiple nodes based on multiple communication links, and determines the initial edge attributes of each edge based on multiple data flow information under each communication link, and obtains the initial data topology map of the first business domain; receives the global node attributes and global edge attributes sent by the central coordination device, and updates the initial data topology map based on the global node attributes and global edge attributes to obtain the global data topology map of the first business domain. The present invention performs data topology restoration through data flow information between systems within the domain, and can reflect the actual business operation of each system based on business flow information, and realizes high-precision intra-domain topology restoration. In addition, through the federated learning framework, the topology data of different business domains are cross-domain aggregated, which can improve the security and privacy of the data and realize the comprehensive and accurate restoration of the distribution network data.

[0063] In order to implement the above embodiment, the embodiment of the present invention also provides a distribution network data topology restoration method applied to a central coordination device.

[0064] Figure 2 The second flow chart of the method for restoring the topology of the distribution network data provided by the embodiment of the present invention. It should be noted that the method for restoring the topology of the distribution network data provided by the embodiment of the present invention is applied to the central coordination device. Figure 2 As shown, the method may include the following steps.

[0065] Step 201, receiving the initial node attributes of each node and the initial edge attributes of each edge in the initial data topology map of each business domain of the distribution network sent by multiple first participating devices; wherein the initial data topology map is constructed by the first participating device based on the data flow information between each service involved in the business domain to which it belongs.

[0066] That is to say, the architecture of the distribution network data topology restoration in an embodiment of the present invention includes multiple first participating devices and a central coordination device, each first parameter device is used to restore the data topology within its business domain, and the central coordination device is used to perform cross-domain aggregation of node attributes and edge attributes.

[0067] In some embodiments, after each first participating device obtains the initial data topology map of the area to which it belongs, it sends the initial node attributes of each node and the initial edge attributes of each edge in the initial data topology map to the central coordination device. The central coordination device receives the initial node attributes and initial edge attributes of the corresponding business domain sent by each first participating device.

[0068] Step 202, based on the federated learning framework, cross-domain aggregation processing is performed on the initial node attributes and initial edge attributes in each business domain to obtain the global node attributes of each node and the global edge attributes of each edge in each business domain.

[0069] It can be understood that there will be overlaps in services within different business domains, forming connections across business domains. The nodes corresponding to these overlapping services usually have common characteristics, and there will also be overlaps in the data flows between overlapping services. The edges corresponding to the overlaps of these data flows usually also have common characteristics. Therefore, in order to make the node attributes and edge attributes more complete and ensure data isolation between devices with different parameters, the central coordination device can perform cross-domain aggregation of node attributes and edge attributes based on the federated learning framework to obtain the global node attributes of each node in each business domain and the global edge attributes of each edge.

[0070] Among them, the global node attributes refer to the complete node attributes after the central coordination device performs cross-domain aggregation based on the initial node attributes of different business domains, and the global edge attributes refer to the complete edge attributes after the central coordination device performs cross-domain aggregation based on the initial edge attributes of different business domains.

[0071] In some embodiments, the central coordination device can compare the IP address information and port numbers of nodes between different business domains based on the initial node attributes of different nodes in each business domain, and determine multiple shared node groups, each shared node group includes multiple shared nodes, and a shared node refers to a node that exists in multiple business domains, that is, a node with consistent IP address information and port numbers between different business domains. For example, the IP address information and port number of node 1 in business domain A are consistent with the IP address information and port number of node 2 in business domain B, then node 1 in business domain A and node 2 in business domain B are shared nodes; for each shared node group, the initial node attributes of the multiple shared nodes in the shared node group are taken as the union, and the result after taking the union is used as the global node attribute of each shared node in the shared node group; for other nodes that are not shared nodes, since these nodes do not exist across domains, there is no need to perform cross-domain aggregation on the node attributes of these nodes, and the initial node attributes of these nodes are directly used as the global node attributes, and finally the global node attributes of each node in each business domain are obtained.

[0072] In some embodiments, the central coordination device can compare the communication links corresponding to each edge between different business domains based on the initial edge attributes of different edges in each business domain, and determine multiple shared edge groups, each shared edge group includes multiple shared edges, and a shared edge refers to an edge that exists in multiple business domains, that is, an edge between different business domains where the source service address information is consistent with the destination service address information, such as edge 1 from node 1 in business domain A to node 2, and edge 2 from node 3 to node 4 in business domain B. If the IP address information and port number of node 1 are consistent with the IP address information and port number of node 3, and the IP address information and port number of node 2 are consistent, The port number is consistent with the IP address information and port number of node 4, then edge 1 in business domain A and edge 2 in business domain B are shared edges; for each shared edge group, the initial edge attributes of multiple shared edges in the shared edge group are taken as the union, and the time-overlapping attributes of the result obtained after taking the union are merged to obtain the merged edge attributes as the global edge attributes of each shared edge in the shared edge group; for other edges that are not shared edges, since these edges do not exist across domains, there is no need to perform cross-domain aggregation on the edge attributes of these edges, and the initial edge attributes of these edges are directly used as the global edge attributes, and finally the global edge attributes of each edge in each business domain are obtained.

[0073] Step 203, the global node attributes of each node in each business domain and the global edge attributes of each edge are sent to the first participating device of the corresponding business domain, so that each first participating device updates the initial data topology map based on the global node attributes and the global edge attributes to obtain the global data topology map of the corresponding business domain.

[0074] As an example, the global node attributes of each node in business domain A and the global edge attributes of each edge are sent to the first participating device corresponding to business domain A. The first participating device corresponding to business domain A updates the initial data topology map of business domain A based on the received global node attributes and global edge attributes to obtain the global data topology map of business domain A.

[0075] According to the distribution network data topology restoration method of the embodiment of the present invention, the central coordination device receives the initial node attributes of each node in the initial data topology map of each business domain of the distribution network sent by multiple first participating devices, as well as the initial edge attributes of each edge; based on the federated learning framework, the initial node attributes and initial edge attributes in each business domain are respectively cross-domain aggregated to obtain the global node attributes of each node in each business domain and the global edge attributes of each edge; the global node attributes of each node in each business domain and the global edge attributes of each edge are sent to the first participating device of the corresponding business domain, so that each first participating device updates the initial data topology map based on the global node attributes and the global edge attributes to obtain the global data topology map of the corresponding business domain. The present invention performs data topology restoration through data flow information between systems within the domain, and can reflect the actual business operation of each system based on business flow information, thereby realizing high-precision intra-domain topology restoration. In addition, through the federated learning framework, cross-domain aggregation of topological data of different business domains can improve data security and privacy, and realize comprehensive and accurate restoration of distribution network data.

[0076] The following describes a distribution network data topology restoration device provided by an embodiment of the present invention. The distribution network data topology restoration device described below and the distribution network data topology restoration method described above can correspond to each other.

[0077] Figure 3 This is one of the structural diagrams of the distribution network data topology restoration device provided in the embodiment of the present invention. It should be noted that the distribution network data topology original device in the embodiment of the present invention is deployed in the first participating device. Figure 3 As shown, the device includes: a first acquisition module 310 , a first determination module 320 , a second determination module 330 , a first sending module 340 and a first receiving module 350 .

[0078] The first acquisition module 310 is used to acquire multiple data flow information between various services involved in the first business domain of the distribution network within a preset time period; the first business domain is the business domain to which the first participating device belongs; A first determination module 320 is used to extract address information of each service from the multiple data stream information, and determine multiple nodes and initial node attributes of each node based on the address information of each service; wherein each service is a node, and the initial node attributes include the address information of the service; A second determination module 330 is used to determine multiple communication links based on the multiple data flow information, determine multiple edges between multiple nodes based on the multiple communication links, and determine the initial edge attribute of each edge based on the multiple data flow information under each communication link, so as to obtain an initial data topology graph of the first service domain; A first sending module 340 is used to send the initial node attributes and the initial edge attributes to the central coordination device; wherein the central coordination device is used to perform cross-domain aggregation processing on the initial node attributes and the initial edge attributes of each business domain based on the federated learning framework to obtain the global node attributes of each node and the global edge attributes of each edge; The first receiving module 350 is used to receive the global node attributes and global edge attributes sent by the central coordination device, and update the initial data topology map based on the global node attributes and global edge attributes to obtain the global data topology map of the first service domain.

[0079] In some embodiments, the second determination module 330 is specifically used to: determine multiple first communication links based on source service address information and destination service address information in multiple data flow information; determine the stability index value of each first communication link based on the data flow information under each first communication link; and determine multiple communication links from multiple first communication links based on the stability index value.

[0080] In some embodiments, the stability index value includes the standard deviation of the number of flows and the average flow of the first communication link; the second determination module 330 is also used to: determine multiple first communication links whose standard deviation of the number of flows is less than or equal to the first threshold and whose average flow is greater than or equal to the second threshold as multiple communication links.

[0081] In some embodiments, the second determination module 330 is also used to: for each communication link, based on multiple data flow information under the communication link, determine the flow attributes of each of the multiple data flows under the communication link; the flow attributes include the timestamp, traffic size and / or flow intensity of the data flow; merge the attributes of the data flows that overlap in time under the communication link to obtain the flow attributes of multiple independent data flows under the communication link; determine the flow attributes of the multiple independent data flows under the communication link as the initial edge attributes of the corresponding edge of the communication link.

[0082] Figure 4 This is a second structural diagram of the distribution network data topology restoration device provided by the embodiment of the present invention. It should be noted that the distribution network data topology original device of the embodiment of the present invention is deployed in the central coordination device. Figure 4 As shown, the device includes: a second receiving module 410 , an aggregation module 420 and a second sending module 430 .

[0083] The second receiving module 410 is used to receive the initial node attributes of each node and the initial edge attributes of each edge in the initial data topology diagram of each service domain of the distribution network sent by multiple first participating devices; wherein the initial data topology diagram is constructed by the first participating device based on the data flow information between each service involved in the service domain to which it belongs; Aggregation module 420, used to perform cross-domain aggregation processing on initial node attributes and initial edge attributes of each business domain based on the federated learning framework, to obtain global node attributes of each node and global edge attributes of each edge in each business domain; The second sending module 430 is used to send the global node attributes of each node in each business domain and the global edge attributes of each edge to the first participating device of the corresponding business domain, so that each first participating device updates the initial data topology map based on the global node attributes and the global edge attributes to obtain the global data topology map of the corresponding business domain.

[0084] Figure 5 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 5As shown, the electronic device may include: a processor 510, a communication interface 520, a memory 530 and a communication bus 540, wherein the processor 510, the communication interface 520 and the memory 530 communicate with each other through the communication bus 540. The processor 510 may call the computer program in the memory 530 to execute the steps of the distribution network data topology restoration method provided in the above-mentioned embodiments, including: obtaining multiple data flow information between various services involved in the first business domain of the distribution network within a preset time period; the first business domain is the business domain to which the first participating device belongs; extracting the address information of each service from the multiple data flow information, and determining multiple nodes and initial node attributes of each of the nodes based on the address information of each service; wherein each of the services is a node, and the initial node attributes include the address information of the service; determining multiple communication links based on the multiple data flow information, and determining the multiple nodes based on the multiple communication links. , and based on the multiple data flow information under each of the communication links, determine the initial edge attributes of each edge to obtain the initial data topology map of the first business domain; send the initial node attributes and the initial edge attributes to the central coordination device; wherein the central coordination device is used to perform cross-domain aggregation processing on the initial node attributes and initial edge attributes of each business domain based on the federated learning framework, and obtain the global node attributes of each node and the global edge attributes of each edge; receive the global node attributes and the global edge attributes sent by the central coordination device, and update the initial data topology map based on the global node attributes and the global edge attributes to obtain the global data topology map of the first business domain. And / or, The steps of executing the distribution network data topology restoration method provided by the above embodiments include: receiving the initial node attributes of each node and the initial edge attributes of each edge in the initial data topology map of each business domain of the distribution network sent by multiple first participating devices; wherein the initial data topology map is constructed by the first participating device based on the data flow information between the various services involved in the business domain to which it belongs; based on the federated learning framework, cross-domain aggregation processing is performed on the initial node attributes and initial edge attributes in each business domain respectively to obtain the global node attributes of each node in each business domain and the global edge attributes of each edge; the global node attributes of each node in each business domain and the global edge attributes of each edge are sent to the first participating device of the corresponding business domain, so that each of the first participating devices updates the initial data topology map based on the global node attributes and the global edge attributes to obtain the global data topology map of the corresponding business domain.

[0085] In addition, the logic instructions in the above-mentioned memory 530 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0086] On the other hand, an embodiment of the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the steps of the distribution network data topology restoration method provided in the above embodiments.

[0087] On the other hand, an embodiment of the present invention further provides a processor-readable storage medium, wherein the processor-readable storage medium stores a computer program, and the computer program is used to enable the processor to execute the steps of the distribution network data topology restoration method provided in the above embodiments.

[0088] The processor-readable storage medium can be any available medium or data storage device that can be accessed by the processor, including but not limited to magnetic storage (such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO), etc.), optical storage (such as CD, DVD, BD, HVD, etc.), and semiconductor storage (such as ROM, EPROM, EEPROM, non-volatile memory (NANDFLASH), solid-state drive (SSD)), etc.

[0089] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0090] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0091] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for restoring distribution network data topology, characterized in that: Applied to a first participating device, the method includes: Acquire multiple data flow information between various services involved in a first service domain of the distribution network within a preset time period; the first service domain is the service domain to which the first participating device belongs; Extracting address information of each service from the plurality of data stream information, and determining a plurality of nodes and initial node attributes of each of the nodes based on the address information of each service; wherein each of the services is a node, and the initial node attributes include the address information of the service; Determine multiple communication links based on the multiple data flow information, determine multiple edges between the multiple nodes based on the multiple communication links, and determine the initial edge attribute of each edge based on the multiple data flow information under each of the communication links, to obtain an initial data topology graph of the first service domain; Sending the initial node attributes and the initial edge attributes to a central coordination device; wherein the central coordination device is used to perform cross-domain aggregation processing on the initial node attributes and initial edge attributes of each business domain based on a federated learning framework to obtain the global node attributes of each node and the global edge attributes of each edge; The global node attributes and the global edge attributes sent by the central coordination device are received, and the initial data topology map is updated based on the global node attributes and the global edge attributes to obtain the global data topology map of the first service domain.

2. The method according to claim 1, characterized in that The determining of a plurality of communication links based on the plurality of data flow information comprises: Determine a plurality of first communication links according to the source service address information and the destination service address information in the plurality of data flow information; Determine a stability index value of each of the first communication links based on data flow information under each of the first communication links; The plurality of communication links are determined from the plurality of first communication links based on the stability index value.

3. The method according to claim 2, characterized in that The stability index value includes a standard deviation of the number of flows and an average flow rate of the first communication link; and determining the plurality of communication links from the plurality of first communication links based on the stability index value includes: A plurality of first communication links whose standard deviation of the number of flows is less than or equal to a first threshold and whose average flow is greater than or equal to a second threshold are determined as the plurality of communication links.

4. The method according to any one of claims 1 to 3, characterized in that The determining of the initial edge attribute of each edge based on the multiple data flow information under each of the communication links includes: For each of the communication links, based on the information of multiple data flows under the communication link, determine the flow attributes of each of the multiple data flows under the communication link; the flow attributes include the timestamp, flow size and / or flow strength of the data flow; Merging attributes of data streams that overlap in time under the communication link to obtain stream attributes of multiple independent data streams under the communication link; The flow attributes of the multiple independent data flows under the communication link are determined as the initial edge attributes of the corresponding edges of the communication link.

5. A method for restoring distribution network data topology, characterized in that: Applied to a central coordination device, the method comprises: Receive the initial node attributes of each node and the initial edge attributes of each edge in the initial data topology graph of each service domain of the distribution network sent by multiple first participating devices; wherein the initial data topology graph is constructed by the first participating device based on the data flow information between various services involved in the service domain to which it belongs; Based on the federated learning framework, cross-domain aggregation processing is performed on the initial node attributes and initial edge attributes in each of the business domains to obtain the global node attributes of each node in each of the business domains and the global edge attributes of each edge; The global node attributes of each node in each of the business domains and the global edge attributes of each edge are sent to the first participating device of the corresponding business domain, so that each of the first participating devices updates the initial data topology map based on the global node attributes and the global edge attributes to obtain the global data topology map of the corresponding business domain.

6. A distribution network data topology restoration device, characterized in that: Deployed on a first participating device, the apparatus comprises: A first acquisition module is used to acquire multiple data flow information between various services involved in a first service domain of the distribution network within a preset time period; the first service domain is the service domain to which the first participating device belongs; A first determination module, configured to extract address information of each service from the plurality of data stream information, and determine a plurality of nodes and initial node attributes of each of the nodes based on the address information of each service; wherein each of the services is a node, and the initial node attributes include the address information of the service; A second determination module is used to determine multiple communication links based on the multiple data flow information, determine multiple edges between the multiple nodes based on the multiple communication links, and determine the initial edge attribute of each edge based on the multiple data flow information under each of the communication links, so as to obtain an initial data topology graph of the first service domain; A first sending module is used to send the initial node attributes and the initial edge attributes to a central coordination device; wherein the central coordination device is used to perform cross-domain aggregation processing on the initial node attributes and initial edge attributes of each business domain based on a federated learning framework to obtain the global node attributes of each node and the global edge attributes of each edge; The first receiving module is used to receive the global node attributes and the global edge attributes sent by the central coordination device, and update the initial data topology map based on the global node attributes and the global edge attributes to obtain the global data topology map of the first service domain.

7. A distribution network data topology restoration device, characterized in that: Deployed in a central coordination device, the device includes: A second receiving module is used to receive the initial node attributes of each node and the initial edge attributes of each edge in the initial data topology diagram of each service domain of the distribution network sent by multiple first participating devices; wherein the initial data topology diagram is constructed by the first participating device based on the data flow information between various services involved in the service domain to which it belongs; An aggregation module, used to perform cross-domain aggregation processing on the initial node attributes and initial edge attributes of each business domain based on a federated learning framework, to obtain global node attributes of each node in each business domain and global edge attributes of each edge; The second sending module is used to send the global node attributes of each node in each of the business domains and the global edge attributes of each edge to the first participating device of the corresponding business domain, so that each of the first participating devices updates the initial data topology map based on the global node attributes and the global edge attributes to obtain the global data topology map of the corresponding business domain.

8. An electronic device comprising a processor and a memory storing a computer program, characterized in that: When the processor executes the computer program, the distribution network data topology restoration method described in any one of claims 1 to 5 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for restoring the distribution network data topology according to any one of claims 1 to 5 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for restoring the distribution network data topology according to any one of claims 1 to 5 is implemented.

Images