Intelligent detection method and system for substation configuration files based on deep learning

Abstract

The present invention relates to the technical field of configuration file detection, and the present invention relates to a method and system for intelligent detection of substation configuration files based on deep learning, which calculates a first safety factor; obtains a first number of all links in a first externally associated IED linked list; obtains a second number of all links in a second externally associated IED linked list; calculates a second safety factor, which is related to the first number and the second number; calculates a target safety factor; in response to the target safety factor being less than a threshold, detects the communication parameters of all IEDs in the SCD configuration file; and when the target safety factor is greater than or equal to the threshold, detects the communication parameters of the target IED. When adding an IED, the present invention determines the scope of detecting the IED according to the probability of its impact on the logical communication relationship, so as to avoid erroneous judgments or omissions of key information that may be caused by manual monitoring, and ensure the quality of the detection work.

Description

Technical Field

[0001] The present invention generally relates to the technical field of configuration file detection. More specifically, the present invention relates to a method and system for intelligent detection of substation configuration files based on deep learning. Background Art

[0002] The engineering configuration file of the smart substation needs to go through multiple processes such as design, construction, equipment assembly, equipment debugging, and acceptance during the construction of the entire smart substation, and a large amount of file data will be generated in each link of the process. The configuration file volume is large. The manufacturer sets the value and debugs the equipment according to the SCD configuration file provided by the power design institute. The SCD configuration file management and control system is actually an information management system that can be traced back to the source. The configuration file describes the instance configuration and communication parameter information of all intelligent electronic devices IED, the connection information between IEDs, and the substation primary system structure and other configuration information, which is crucial to the configuration of the entire station.

[0003] When a new device is introduced into a substation, the SCD configuration file needs to be updated. At this time, the physical or logical connection mode between IED devices usually changes, such as signal lines, control logic, etc. Since there are many virtual terminals corresponding to IEDs, and the configuration of virtual terminals is often missed or inaccurately described, the phase configuration is wrong, and the system does not correspond, when some channels are missed or wrongly connected, the safe operation of the entire substation is seriously affected. Therefore, the normative check of the SCD configuration file usually includes the validity check of the IED communication parameter reference. It should be noted that the number of IEDs in the SCD configuration file is usually large, and a large substation may contain dozens or even hundreds of IEDs to meet the needs of power system safety, reliability and monitoring. Therefore, every time a new device is introduced into a substation, a lot of detection work is required to determine whether the logical communication relationship between the IEDs is correct. However, a large amount of detection work may cause the relevant inspectors to make wrong judgments or miss key information due to fatigue and other reasons, resulting in a decrease in the quality of detection work. Summary of the invention

[0004] In order to solve the technical problem that the large amount of inspection work mentioned above may cause relevant inspectors to misjudge or omit key information due to fatigue and other reasons, resulting in a decrease in the quality of the inspection work, the present invention provides solutions in the following aspects.

[0005] In a first aspect, the present invention proposes a substation configuration file intelligent detection method based on deep learning, the method comprising: calculating a first safety factor α :

[0006] ; wherein the increased preset number of IEDs is used as the target IED,m is the number of target IEDs. i target IEDs, a 1,i is the number of the sender IED of the target IED, a 2,i is the number of the target IED receiving IED, exp is the natural exponential e The exponential function with base , a 0 is a preset first standard quantity; visualize the configuration file after adding the target IED to obtain and obtain the first quantity of links in the first external associated IED list through the image recognition model; visualize the configuration file before adding the target IED to obtain the second external associated IED list, and obtain the second quantity of links in the second external associated IED list through the image recognition model; calculate the second safety factor β , the second safety factor β The absolute value of the difference between the first number and the second number and the number of target IEDs are related; Calculating a target safety factor H : H =(1- η )× α + η × β ,in η is the preset weight coefficient, 0< η <1; responds to target safety factor H If the value is less than the threshold, the communication parameters of all IEDs in the SCD configuration file are checked; in response to the target safety factor H If the value is greater than or equal to the threshold, the communication parameters of the target IED are detected.

[0007] In one embodiment, for any target IED, obtaining the number of the target IED, the number of sender IEDs of each target IED, and the number of target IED receiver IEDs includes: locating the target IED according to the device identifier corresponding to the target IED; obtaining the sender IED of the target IED according to the external reference list of the target IED, and obtaining the number of sender IEDs of the target IED; traversing the external reference list of each IED, taking the sender IED as the target IED as the receiver IED of the target IED, and obtaining the number of receiver IEDs of the target IED.

[0008] In one embodiment, it also includes obtaining the number of target IEDs, the number of receiving IEDs of each target IED, and the number of sending IEDs of each target IED through a deep learning model: identifying the label of each IED device through an image recognition model, obtaining the added IED, recording the added IED as the target IED, and counting the number of the target IEDs; for any target IED, obtaining the number of receiving IEDs in the first external associated IED list that receive the output data of the target IED through the image recognition model; for any target IED, obtaining the number of sending IEDs in the first external associated IED list that want to transmit data to the target IED through the image recognition model.

[0009] In one embodiment, the second safety factor is calculated β The formula for calculating the second safety factor is: β The formula is: ,in, n 1 is the first quantity, n 2 is the second number.

[0010] In one embodiment, the second standard quantity n 0 The value is the average value of the absolute value of the increase or decrease of the link in the external associated IED list corresponding to the SCD configuration file after each IED is added to the SCD configuration file.

[0011] In one embodiment, it also includes determining the first threshold and the target safety factor H and the size between the second threshold and the target safety factor, wherein the first threshold is greater than the second threshold: in response to the target safety factor H greater than a first threshold, detecting the communication parameters of the target IED; in response to the target safety factor H greater than the second threshold and less than or equal to the first threshold, detecting the communication parameters of all target IEDs, all sender IEDs of the target IEDs, and all receiver IEDs of the target IEDs; in response to the target safety factor H If the value is less than or equal to the second threshold, the files corresponding to all IEDs are detected.

[0012] In one embodiment, detecting the communication parameters of any IED in the SCD includes: determining whether the IED name, access point name, GOOSE control block name, and logical device instance of the communication part are consistent with the configuration corresponding to the IED part; in response to any one of the IED name, access point name, GOOSE control block name, and logical device instance of the communication part being inconsistent with the configuration corresponding to the IED part, confirming that the file corresponding to the IED is wrong.

[0013] In one embodiment, constructing an image recognition model includes: for any external associated IED link list, selecting each link through an identification box to obtain a training link list image, wherein, for any training link list image, each link indicates the transmission direction of data in the link with the direction of the arrow, and each IED is labeled with its corresponding device identifier; obtaining multiple training link list images through multiple external associated IED link lists, and training a neural network model with the multiple training link list images as training sets to obtain an image recognition model.

[0014] In a second aspect, a substation configuration file intelligent detection system based on deep learning includes a processor and a memory, the memory stores a computer program, and the processor executes the computer program to implement the substation configuration file intelligent detection method based on deep learning described in any one of the above invention contents.

[0015] The beneficial effects of the present invention are:

[0016] When at least one IED is added to the configuration file, the present invention determines whether the data of the IED is in a balanced flow state through the receiving IED and the sending IED of the data sent and received by each added IED, and also determines the probability of abnormality in the logical communication relationship between the IED and other IEDs through whether the total number of the receiving IED and the sending IED of the IED is normal. If the probability of abnormality in the logical communication relationship between the added IED and other IEDs is high, the communication parameters of all IEDs are detected; if the probability of abnormality in the logical communication relationship between the added IED and other IEDs is low, only the communication parameters of the IED added to the SCD configuration file are detected. Based on this, it is avoided that a large amount of detection work may cause the relevant inspectors to make wrong judgments or miss key information due to fatigue and other reasons, resulting in a decrease in the quality of the detection work. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0018] Figure 1 is a flowchart schematically illustrating the steps of a substation configuration file intelligent detection method based on deep learning according to an embodiment of the present invention;

[0019] Figure 2 is a structural diagram schematically showing a first external associated IED linked list according to an embodiment of the present invention;

[0020] Figure 3 is a structural diagram schematically showing a second external associated IED linked list according to an embodiment of the present invention;

[0021] Figure 4 is a structural diagram schematically showing a training linked list image according to an embodiment of the present invention;

[0022] Figure 5 It is a structural block diagram schematically showing a substation configuration file intelligent detection system based on deep learning according to an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0024] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0025] Figure 1 It is a flowchart schematically showing the steps of a substation configuration file intelligent detection method based on deep learning according to an embodiment of the present invention.

[0026] like Figure 1 As shown, the substation configuration file intelligent detection method based on deep learning includes steps S101 to S107.

[0027] Step S101: Calculate a first safety factor.

[0028] Among them, the first safety factor is calculated α The formula is:

[0029] ; wherein the increased preset number of IEDs is used as the target IED, m is the number of target IEDs. i target IEDs, a 1,i is the number of the sender IED of the target IED, a 2,i is the number of the target IED receiving IED, exp is the natural exponential e The exponential function with base , a 0 The preset first standard quantity.

[0030] It should be noted that in order to maintain the balance of data flow and avoid overloading of IED devices while other devices are underloaded, the number of corresponding sender IEDs and the number of corresponding receiver IEDs are usually close in the design of IED devices. The greater the difference between the number of corresponding sender IEDs and the number of corresponding receiver IEDs of an IED, the greater the probability of errors in the logical communication between the IED and other IEDs (for example, the communication link between the IED and other IEDs is missing or the data transmission direction in the communication link is wrong). Based on this, the first i Take a target IED as an example, The difference between the number of sender IEDs and the number of receiver IEDs of the target IED is represented. The larger the difference is, the greater the probability of an error relationship occurring in the logical communication between the IED and other IEDs.

[0031] Furthermore, the first standard quantity a 0 Indicates the normal value of the number of IEDs that have a communication relationship with the IED in the SCD configuration file after adding an IED. The number of IED devices that need to interact with it usually fluctuates between a certain value, and for different types of IED devices, the number of data interactions required with them varies slightly. Therefore, for the target IED, the greater the deviation of the number of IEDs that have a logical communication relationship with it from the normal value, the greater the probability of errors in the logical communication between it and other IEDs.

[0032] In one embodiment, the first standard quantity is calculated a 0 The method includes: for any IED addition in history, the sum of the number of sender IEDs and the number of receiver IEDs corresponding to all IED devices is averaged to obtain the average coefficient corresponding to the IED device added this time. The average coefficients corresponding to multiple IED additions in history are averaged to obtain the first standard quantity. a 0 Based on this, i Take a target IED as an example, The greater the deviation of the number of IEDs in communication with the target IED from the normal value, the greater the probability of error in the logical communication relationship between the target IED and others.

[0033] because and The first safety factor α Inversely proportional, so the larger the first safety factor, α The larger the value is, the smaller the probability that errors occur in the logical communication between at least one target IED and other IEDs among all the target IEDs.

[0034] In one embodiment, for any target IED, obtaining the number of the target IED, the number of sender IEDs of each target IED, and the number of receiver IEDs of the target IED includes: locating the target IED according to a device identifier corresponding to the target IED. Obtaining the sender IED of the target IED according to an external reference list of the target IED, and obtaining the number of sender IEDs of the target IED. Traversing the external reference list of each IED, taking the sender IED as the target IED as the receiver IED of the target IED, and obtaining the number of receiver IEDs of the target IED.

[0035] It should be noted that each IED device configuration has a unique device identifier, and the device identifier corresponding to the IED is usually stored in a structured format. Through the device identifier, operators can track the location of the IED device.

[0036] In another embodiment, the number of target IEDs, the number of receiving IEDs of each target IED, and the number of sending IEDs of each target IED are obtained through a deep learning model: the label of each IED device is identified through an image recognition model, the added IED is obtained, the added IED is recorded as the target IED, and the number of the target IEDs is counted; for any target IED, the number of receiving IEDs that receive output data of the target IED in the first external associated IED list is obtained through the image recognition model; for any target IED, the number of sending IEDs that want to transmit data to the target IED in the first external associated IED list is obtained through the image recognition model.

[0037] Step S102: Visualize the configuration file after adding the target IED to obtain and obtain the first number of links in the first external associated IED linked list through an image recognition model.

[0038] The image recognition model is used to identify the label of the IED to find the IED. The image recognition model is also used to select the links in the image to count the number of links in the image.

[0039] Furthermore, the external associated IED linked list is used to record the data input and output relationship between each IED and other IEDs to present the data flow in the SCD file. The lists corresponding to each IED in the external associated IED linked list include SV sending terminals, SV receiving terminals, GOOSE sending terminals and GOOSE receiving terminals.

[0040] Figure 2 FIG. 4 is a diagram schematically showing a structure of a first external associated IED linked list according to an embodiment of the present invention.

[0041] like Figure 2As shown, the first external associated IED linked list includes q + m The image recognition model selects each link through a rectangular first recognition frame. The number of the first recognition frames is counted to obtain the first number of links in the first external associated IED chain list.

[0042] Step S103: Visualize the configuration file before adding the target IED to obtain a second externally associated IED linked list, and obtain a second number of links in the second externally associated IED linked list through an image recognition model.

[0043] Figure 3 is a diagram schematically showing a structure of a second externally associated IED linked list according to an embodiment of the present invention.

[0044] like Figure 3 As shown, the second external associated IED list includes q The image recognition model selects each link through a rectangular second recognition frame. The number of the second recognition frames is counted to obtain the first and second numbers of the links in the second external associated IED chain list.

[0045] In one embodiment, constructing an image recognition model includes: for any external associated IED link list, selecting each link through an identification box to obtain a training link list image, wherein, for any training link list image, each link indicates the transmission direction of data in the link with the direction of the arrow, and each IED is labeled with its corresponding device identifier; obtaining multiple training link list images through multiple external associated IED link lists, and training a neural network model with the multiple training link list images as training sets to obtain an image recognition model.

[0046] Figure 4 Schematically illustrates a structural diagram of a training linked list image according to an embodiment of the present invention.

[0047] like Figure 4 As shown, for any two IEDs connected by a link, the IED indicated by the arrow in the link is the first IED, the other IED is the second IED, the second IED is the first IED sender IED, and the first IED is the second IED receiver IED.

[0048] Step S104: Calculate the second safety factor β , the second safety factor β It is related to both the absolute value of the difference between the first number and the second number and the number of target IEDs.

[0049] In one embodiment, the second safety factor is calculated β The formula is: ,in, n 1is the first quantity, n 2 is the second number. n 1 - n 2 Indicates the number of links added due to the addition of target IEDs. a 1,i is the number of the sender IED of the target IED, a 2,i is the number of the target IED receiving IED, so is the theoretical number of links that are connected to the target IED. Also, because installing a new IED may require rearranging the network topology or adjusting the location of existing equipment, that is, in the process of adding a target IED, the link between two non-target IEDs may disappear, and one of the two non-target IEDs may be connected to the target IED through a link. Therefore, the second safety factor β The bigger, n 1 - n 2 (the number of links added due to the addition of target IEDs) and The smaller the difference between (the theoretical number of links connected to the target IED), the smaller the number of links between two IEDs that are not the target IEDs that disappear due to the addition of the target IED.

[0050] It should be noted that, since modern IEDs are usually designed to be compatible with existing equipment and systems, in order to integrate the added IED into the existing network without causing large-scale system reconstruction, usually only a small number of links between other existing IEDs are adjusted. Based on this, the greater the difference between the number of added links and the number of links connected to the target IED, the more links between the IEDs of the two non-target IEDs disappear. At this time, among the links between the IEDs of the two non-target IEDs, there is a greater probability that there are links that disappear due to normal needs (such as the above-mentioned rearrangement of the network topology or adjustment of the location of existing equipment), that is, the greater the probability of errors in the logical communication between the added target IED and other IEDs.

[0051] In summary, the greater the second safety factor, β , among all the target IEDs, the probability of errors occurring in the logical communication between at least one target IED and other IEDs is smaller.

[0052] Step S105: Calculate the target safety factor.

[0053] Among them, the target safety factor H The calculation formula is:H =(1- η )× α + η × β , η is the preset weight coefficient, 0< η <1.

[0054] It should be noted that by α With the second safety factor β Perform a weighted sum to obtain the target safety factor, the first safety factor α The weight parameter is (1- η ), second safety factor β The weight parameter is the weight coefficient η . First safety factor α , Second safety factor β are proportional to the target safety factor. Therefore, the target safety factor H The larger the value is, the smaller the probability that errors occur in the logical communication between at least one target IED and other IEDs among all the target IEDs.

[0055] In one embodiment, η The empirical value of the weight coefficient is 0.1.

[0056] Step S106: Responding to the target safety factor H If the value is less than the threshold, the communication parameters of all IEDs in the SCD configuration file are checked.

[0057] Among them, when the target safety factor H When it is less than the threshold, it means that the probability of the logical communication relationship between IEDs in the SCD configuration file after adding the target IED is high. At this time, the files corresponding to all IEDs are detected, and the files corresponding to all IEDs are monitored first, and then the files corresponding to other IEDs are detected.

[0058] Step S107: Responding to the target safety factor H If the value is greater than or equal to the threshold, the communication parameters of the target IED are detected.

[0059] Among them, when the target safety factor HWhen it is less than the threshold, it means that the probability of an error in the logical communication relationship between IEDs in the SCD configuration file after the target IED is added is small. At this time, only the file corresponding to the target IED is detected. The reason is: the SCD file describes the topology, communication relationship and data model of each device in the power system. The configuration and communication path of each IED are independently defined. Usually, the logical relationship between other existing IEDs will not be directly affected by adding an IED. Therefore, the probability of an error in the logical communication relationship between IEDs in the SCD configuration file is small. At this time, only the file corresponding to the target IED is detected.

[0060] In another embodiment, the first threshold and the target safety factor are determined. H and the size between the second threshold and the target safety factor, wherein the first threshold is greater than the second threshold: in response to the target safety factor H greater than a first threshold, detecting the communication parameters of the target IED; in response to the target safety factor H greater than the second threshold and less than or equal to the first threshold, detecting the communication parameters of all target IEDs, all sender IEDs of the target IEDs, and all receiver IEDs of the target IEDs; in response to the target safety factor H If the value is less than or equal to the second threshold, the files corresponding to all IEDs are detected.

[0061] It should be noted that the detection of the communication parameters of any IED in the SCD includes: respectively determining whether the IED name, access point name, GOOSE control block name, and logical device instance of the communication part are consistent with the configuration corresponding to the IED part; in response to any one of the IED name, access point name, GOOSE control block name, and logical device instance of the communication part being inconsistent with the configuration corresponding to the IED part, confirming that the file corresponding to the IED is wrong.

[0062] Figure 5 It is a structural block diagram schematically showing the substation configuration file intelligent detection system based on deep learning according to this embodiment.

[0063] The present invention also provides a substation configuration file intelligent detection system based on deep learning. Figure 5 As shown, the system includes a processor and a memory, the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the deep learning-based intelligent detection method for substation configuration files according to the first aspect of the present invention is implemented.

[0064] The system also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface, whose configuration and functions are known in the art and thus will not be described in detail here.

[0065] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented using computer-readable / executable instructions that may be stored or otherwise maintained by such a computer-readable medium.

[0066] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0067] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A substation configuration file intelligent detection method based on deep learning, characterized in that: include: Calculate the first safety factor α: Wherein, the preset number of IEDs added is the target IED, m is the number of target IEDs, and for the i-th target IED, a 1,i is the number of the sender IED of the target IED, a 2,i is the number of receiving IEDs of the target IED, exp is an exponential function with the natural exponent e as the base, and a0 is a preset first standard number; Visualize the SCD configuration file after adding the target IED to obtain a first externally associated IED chain list, and obtain a first number of all links in the first externally associated IED chain list through an image recognition model; visualize the SCD configuration file before adding the target IED to obtain a second externally associated IED chain list, and obtain a second number of all links in the second externally associated IED chain list through an image recognition model; calculate a second safety factor β, the second safety factor β being related to a difference between the first number and the second number; Calculate the target safety factor H: H = (1-η) × α + η × β, where η is a preset weight coefficient, 0 < η < 1; in response to the target safety factor H being less than a threshold, detect the communication parameters of all IEDs in the SCD configuration file; in response to the target safety factor H being greater than or equal to the threshold, detect the communication parameters of all target IEDs.

2. The substation configuration file intelligent detection method based on deep learning according to claim 1 is characterized in that: For any target IED, obtaining the number of sender IEDs of each target IED and the number of receiver IEDs of the target IED includes: Locate the target IED according to the device identifier corresponding to the target IED; Obtaining a sender IED of the target IED according to an external reference list of the target IED, and obtaining the number of the sender IEDs of the target IED; The external reference list of each IED is traversed, the IED whose sender IED is the target IED is the receiver IED of the target IED, and the number of the receiver IEDs of the target IED is obtained.

3. The substation configuration file intelligent detection method based on deep learning according to claim 1 is characterized in that: The method also includes obtaining the number of target IEDs, the number of receiving IEDs of each target IED, and the number of sending IEDs of each target IED through a deep learning model: identifying a label of each IED device through an image recognition model, obtaining an added IED label to obtain the added IED, recording the added IED as a target IED, and counting the number of the target IEDs; For any target IED, the number of receiving IEDs in the first external associated IED linked list that receive output data of the target IED is obtained through the image recognition model; For any target IED, the number of sender IEDs transmitting data to the target IED in the first external associated IED linked list is obtained through the image recognition model.

4. The substation configuration file intelligent detection method based on deep learning according to claim 1 is characterized in that: The formula for calculating the second safety factor β is: Among them, n1 is the first number, and n2 is the second number.

5. The substation configuration file intelligent detection method based on deep learning according to claim 1 is characterized in that: The value of the first standard number a0 is: the average value of the number of IEDs added each time the IED is added to the SCD configuration file in the history of adding IEDs to the SCD configuration file.

6. The substation configuration file intelligent detection method based on deep learning according to claim 1 is characterized in that: It also includes determining the size between the first threshold and the target safety factor H, and determining the size between the second threshold and the target safety factor: In response to the target safety factor H being greater than a first threshold, the communication parameters of the target IED are detected; in response to the target safety factor H being greater than a second threshold and less than or equal to the first threshold, the communication parameters of all target IEDs, the sending IEDs of all target IEDs, and the receiving IEDs of all target IEDs are detected; in response to the target safety factor H being less than or equal to a second threshold, the communication parameters of all IEDs in the SCD configuration file are detected, wherein the first threshold is greater than the second threshold.

7. The substation configuration file intelligent detection method based on deep learning according to claim 1 is characterized in that: Checking the communication parameters of any IED in the SCD configuration file includes: For any IED, determine whether the IED name, access point name, GOOSE control block name, and logical device instance of the communication part of the SCD configuration file are consistent with the configuration corresponding to the IED part; For any IED, in response to any one of the IED name, access point name, GOOSE control block name, and logical device instance in the communication part of the SCD configuration file being inconsistent with the configuration corresponding to the IED part, it is confirmed that the logical communication relationship between the IEDs in the SCD configuration file is incorrect.

8. The substation configuration file intelligent detection method based on deep learning according to claim 3 is characterized in that: Building an image recognition model involves: For any external associated IED chain list, each link is selected through the identification box to obtain a training chain list image, wherein, for any training chain list image, each link indicates the transmission direction of data in the link with the direction of the arrow, and each IED is labeled with its corresponding device identifier; A plurality of training list images are obtained through a plurality of external associated IED lists, and a neural network model is trained with the plurality of training list images as training sets to obtain an image recognition model.

9. A substation configuration file intelligent detection system based on deep learning, comprising a processor and a memory, wherein the memory stores a computer program, characterized in that: The processor executes the computer program to implement the deep learning-based intelligent detection method for substation configuration files as described in any one of claims 1 to 8.

Images