Substation equipment anomaly detection method based on deep learning

Abstract

The invention relates to the technical field of anomaly detection, in particular to a substation equipment anomaly detection method based on deep learning. Comprising the following steps: determining whether the operation of a single secondary device is qualified based on a voltage distortion rate, marking the unqualified secondary device as an abnormal device, and determining whether the operation of the secondary device is qualified based on a frequency deviation for the secondary device whose voltage distortion rate is greater than a second preset voltage distortion rate and less than or equal to a first preset voltage distortion rate, unqualified secondary equipment is marked as abnormal equipment, and the secondary equipment is a relay protection device; based on the disqualification ratio, determining that the reason for disqualification of the test process is preset matching degree abnormity or secondary equipment detection abnormity; adjustment is correspondingly carried out according to reasons for disqualification of the test process; according to the invention, the equipment abnormity detection efficiency is effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of anomaly detection, and in particular to a substation equipment anomaly detection method based on deep learning. BACKGROUND

[0002] Secondary equipment is the nerve center and control core of the safe and stable operation of the substation, and plays an irreplaceable role in realizing accurate perception of the state of the power grid, rapid isolation of faults and intelligent regulation of the system. As a key component of secondary equipment, the relay protection device directly undertakes the important functions of real-time monitoring of the state of power equipment, rapid identification of faults and timely execution of protection actions, and its performance directly affects the reliability and safety of the overall operation of the substation. In actual operation, abnormalities in the relay protection device itself, such as misoperation, refusal to operate or setting value drift, may not only mask the real faults of primary equipment, but also cause incorrect protection behavior, thereby expanding the scope of the fault or even causing the system to shut down. Therefore, how to accurately analyze and quickly adjust the abnormal causes of the relay protection device to improve its detection accuracy and diagnostic efficiency for the abnormal state of the substation equipment has become a key technical problem that needs to be solved in the field of intelligent substation operation and maintenance.

[0003] Chinese Patent Publication No. CN112686530B discloses a relay protection operation reliability evaluation method, which includes the following steps: step 1, collecting basic evaluation indexes to establish a relay protection operation reliability evaluation index system; step 2, determining the weight of the basic evaluation indexes of the relay protection operation reliability; step 3, collecting relay protection operation statistical data, and calculating each basic evaluation index and the total score of the operation reliability in the relay protection operation reliability evaluation index system based on the weight value of step 2; effectively reducing the impact of incorrect operation of the relay protection device on the power grid system. It can be seen that the above technical solution has the following problems: the influence of the preset model factor on the substation equipment anomaly detection efficiency is not considered, which makes it difficult to accurately determine the reasons for the unqualified operation of the secondary equipment with a single influencing factor, thereby affecting the substation equipment anomaly detection efficiency. SUMMARY

[0004] Therefore, the present application provides a substation equipment anomaly detection method based on deep learning to overcome the problem in the prior art that the influence of the preset model factor on the substation equipment anomaly detection efficiency is not considered, which makes it difficult to accurately determine the reasons for the unqualified operation of the secondary equipment with a single influencing factor, thereby affecting the substation equipment anomaly detection efficiency.

[0005] To achieve the above-mentioned purpose, the present application provides a substation equipment anomaly detection method based on deep learning, which comprises: The operation of the single secondary device is determined to be qualified based on the voltage distortion rate, and the unqualified secondary device is marked as an abnormal device, wherein for the secondary device with the voltage distortion rate greater than the second preset voltage distortion rate and less than or equal to the first preset voltage distortion rate, the operation of the secondary device is determined to be qualified based on the frequency deviation, and the unqualified secondary device is marked as an abnormal device, and the secondary device is a relay protection device; The unqualified reason of the test process is determined to be a preset matching degree anomaly or a secondary device detection anomaly based on the unqualified proportion; When the unqualified reason is the preset matching degree anomaly, the preset matching degree is reduced based on the version update frequency; and the adjusted preset matching degree is corrected based on the effective proportion difference to reduce the adjusted preset matching degree. After the preset matching degree adjustment is completed, the unqualified proportion is re-detected, and if the unqualified proportion is greater than a preset unqualified proportion, the version update frequency is determined based on the test point number, and the pressure plate timeout time is adjusted based on the network load rate. A service tracking logical node is configured in an SCD file, and the consistency of a GOOSE trigger condition and an actual signal is automatically verified through MMS subscription dataset.

[0006] Further, the process of determining whether the operation of the single secondary device is qualified based on the voltage distortion rate includes: A standard current voltage is injected by a tester, and the deviation of the measurement value of the checking device from the standard value is recorded as the voltage distortion rate; Whether the operation of the secondary device is qualified is determined based on the comparison result of the voltage distortion rate with the first preset voltage distortion rate and the second preset voltage distortion rate; If the voltage distortion rate is greater than the first preset voltage distortion rate, it is determined that the operation of the secondary device is unqualified, and the secondary device is marked as an abnormal device; If the voltage distortion rate is greater than the second preset voltage distortion rate and less than or equal to the first preset voltage distortion rate, whether the operation of the secondary device is qualified is determined based on the frequency deviation; If the voltage distortion rate is less than or equal to the second preset voltage distortion rate, it is determined that the operation of the secondary device is qualified.

[0007] Further, in response to the first preset condition, the process of determining whether the operation of the secondary device is qualified based on the frequency deviation includes: A plurality of monitoring time windows are set in a single monitoring period, and a plurality of detection time points are set in each time window, and the frequency values measured by the relay protection devices at each detection time point are detected; The frequency values measured by the relay protection devices are recorded as current frequencies, the absolute values of the differences between the current frequencies and the rated frequency are recorded as frequency deviation values corresponding to each detection time point, and the average value of the frequency deviation values is recorded as the frequency deviation. If the frequency deviation is greater than the preset frequency deviation, it is determined that the operation of the secondary equipment is unqualified, and the secondary equipment is marked as an abnormal device; If the frequency deviation is less than or equal to the preset frequency deviation, it is determined that the operation of the secondary equipment is qualified. The first preset condition is that the voltage distortion rate is greater than the second preset voltage distortion rate and less than or equal to the first preset voltage distortion rate.

[0008] Further, the process of determining the unqualified reason of the test procedure based on the unqualified proportion includes: Obtain the number of abnormal devices and the total number of devices, and record the ratio of the number of abnormal devices to the total number of devices as the unqualified proportion; If the unqualified proportion is greater than the preset unqualified proportion, it is determined that the unqualified reason is a preset matching degree anomaly, and the preset matching degree is determined based on the version update frequency; If the unqualified proportion is less than or equal to the preset unqualified proportion, it is determined that the unqualified reason is a secondary equipment detection anomaly. If the unqualified proportion is 0, it is determined that the operation of the secondary equipment is qualified and the test is completed.

[0009] Further, in response to the second preset condition, the process of determining the preset matching degree includes: reducing the preset matching degree based on the version update frequency, and the reduction amplitude of the preset matching degree and the version update frequency are in a positive correlation relationship; The second preset condition is that the unqualified reason is determined to be a preset matching degree anomaly.

[0010] Further, in the adjustment process of the preset matching degree, the unqualified proportion is obtained, and the absolute value of the difference between the unqualified proportion and the preset unqualified proportion is recorded as the effective proportion difference; The adjusted preset matching degree is corrected based on the effective proportion difference to reduce the adjusted preset matching degree, and the correction amplitude and the effective proportion difference are in a positive correlation relationship.

[0011] Further, after the preset matching degree adjustment is completed, the unqualified proportion is re-detected, and if the unqualified proportion is greater than the preset unqualified proportion, the version update frequency is determined based on the number of test points.

[0012] Further, the reduction amplitude of the version update frequency is determined based on the number of test points, and the reduction amplitude of the version update frequency and the number of test points are in a positive correlation relationship.

[0013] Further, in the process of correcting the version update frequency, the press plate timeout time is adjusted based on the network load rate, and the increase amplitude of the press plate timeout time and the network load rate are in a positive correlation relationship.

[0014] Further, the service tracking logic node is configured in the SCD file, and the consistency of the GOOSE trigger condition and the actual signal is automatically verified through the MMS subscription dataset.

[0015] Compared with the prior art, the beneficial effects of the present application are that the present application determines whether the operation of the single secondary device is qualified by the voltage distortion rate, and determines whether the operation of the secondary device is qualified based on the frequency deviation for the secondary device with the voltage distortion rate greater than the second preset voltage distortion rate and less than or equal to the first preset voltage distortion rate; the operation of the secondary device can be determined based on multiple factors, so that the operation of the secondary device is more accurate; at the same time, the unqualified reason of the test process is determined to be a preset matching degree anomaly or a secondary device detection anomaly based on the unqualified proportion; and the parameters including the preset matching degree, the version update frequency and the press plate timeout time are adjusted corresponding to the unqualified reason, thereby effectively eliminating the influence of multiple factors in the test process on the device anomaly detection, thereby improving the efficiency of the substation device anomaly detection.

[0016] Further, the present application determines whether the operation of the secondary device is qualified based on the voltage distortion rate, which can quickly determine whether the operation of the secondary device is qualified, thereby improving the efficiency of determining whether the operation of the secondary device is qualified, and further enabling the secondary device to be marked when the operation of the secondary device is unqualified, thereby quickly analyzing the unqualified reason when the operation of the secondary device is unqualified, so that the accuracy and efficiency of the unqualified reason determination are improved.

[0017] Further, the present application determines whether the operation of the secondary device is qualified based on the frequency deviation under the first preset condition, which combines two different dimensions of measurement error for comprehensive determination, avoids the false positive rate and missed report risk of determining whether the operation of the secondary device is qualified according to a single parameter, and further makes the unqualified reason determination more persuasive and reliable.

[0018] Further, the present application determines that the unqualified reason of the test process is a preset matching degree anomaly or a secondary device detection anomaly based on the comparison result of the unqualified proportion and the preset unqualified proportion, which can more accurately analyze the reason for the operation of the secondary device being unqualified based on the unqualified proportion, thereby improving the accuracy of the unqualified reason determination, and thereby improving the efficiency of the substation device anomaly detection.

[0019] Further, the present application determines the preset matching degree based on the version update frequency, so that the determination of the preset matching degree is more reasonable, the preset matching degree reflects whether the SCD file and the device CID file are consistent, avoids the problem that the matching degree of the version update frequency and the consistency standard is poor in the actual environment, thereby affecting the accuracy of the unqualified reason determination, and thereby reducing the efficiency of the substation device anomaly detection.

[0020] Further, in the preset matching degree adjustment process, the preset matching degree after adjustment is corrected based on the effective proportion difference to reduce the preset matching degree after adjustment, the preset matching degree can be more accurately determined based on the effective proportion difference, and the adjustment accuracy of the preset matching degree is further ensured, so that the power substation equipment abnormality detection efficiency is further improved while the preset matching degree is accurately adjusted.

[0021] Further, the preset matching degree is adjusted, and then the unqualified proportion is re-detected. If the unqualified proportion is greater than the preset unqualified proportion, the version update frequency is determined based on the number of test points. The invalid adjustment of the preset matching degree is avoided, and the power substation equipment abnormality detection efficiency is further improved while the running unqualified reason of the test process is quickly excluded.

[0022] Further, the preset matching degree is adjusted, and then the unqualified proportion is re-detected. If the unqualified proportion is greater than the preset unqualified proportion, the version update frequency is determined based on the number of test points. The invalid adjustment of the preset matching degree is avoided, and the power substation equipment abnormality detection efficiency is further improved while the running unqualified reason of the test process is quickly excluded.

[0023] Further, in the correction and update detection frequency process, the press plate timeout time is adjusted based on the network load, so that the press plate timeout time is more in line with the actual situation of the application scenario, the influence of the network load on the press plate timeout time in the actual scenario is avoided, the power substation equipment abnormality detection efficiency is further improved while the running unqualified reason of the test process is quickly excluded.

[0024] Further, the service tracking logical node is configured in the SCD file, so that when an error operation or a safety accident occurs, the specific operator, operation time and operation content can be quickly and accurately traced back, accurate accountability is realized, and the safety of the system is greatly enhanced. BRIEF DESCRIPTION OF DRAWINGS

[0025] Fig. 1 The flowchart of the power substation equipment abnormality detection method based on deep learning of the present application; Fig. 2 The flowchart of the present application for determining whether the operation of the secondary equipment is qualified based on the voltage distortion rate; Fig. 3 The flowchart of the present application for determining whether the operation of the secondary equipment is qualified based on the frequency deviation; Fig. 4 The flowchart of the present application for determining the unqualified reason based on the unqualified proportion; DETAILED DESCRIPTION

[0026] In order to make the objects, technical schemes and advantages of the present application clearer, the following further describes the present application with reference to the embodiments; it should be understood that the specific embodiments described herein are only used to explain the present application and not to limit the present application.

[0027] The preferred embodiments of the present application are described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present application and are not intended to limit the protection scope of the present application.

[0028] It should be noted that in the description of the present application, the terms "upper", "lower", "left", "right", "inner", "outer" and the like indicating the direction or positional relationship of the terms are based on the direction or positional relationship shown in the drawings, which is only for the convenience of description and does not indicate or imply that the device or element must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application.

[0029] In addition, it should also be noted that in the description of the present application, unless otherwise explicitly specified and limited, the terms "mounting", "connecting", "connection" should be understood broadly, for example, it can be fixed connection, or detachable connection, or integrally connected; it can be mechanical connection, or electrical connection; it can be directly connected, or indirectly connected through an intermediate medium, or the internal communication of two elements. Those skilled in the art can understand the specific meaning of the above terms in the present application according to the specific circumstances.

[0030] Please refer to Figs. 1 to 4 The present application provides a substation equipment anomaly detection method based on deep learning, which comprises: determining whether the operation of a single secondary equipment is qualified based on the voltage distortion rate, and marking the unqualified secondary equipment as an abnormal equipment, wherein for the secondary equipment with a voltage distortion rate greater than a second preset voltage distortion rate and less than or equal to a first preset voltage distortion rate, it is determined whether the operation of the secondary equipment is qualified based on the frequency deviation, and the unqualified secondary equipment is marked as an abnormal equipment, and the secondary equipment is a relay protection device; determining the unqualified reason of the test process based on the unqualified proportion, which is a preset matching degree anomaly, or a secondary equipment detection anomaly; When the unqualified reason is a preset matching degree anomaly, the preset matching degree is reduced based on the version update frequency; and the adjusted preset matching degree is corrected based on the effective proportion difference to reduce the adjusted preset matching degree; After the preset matching degree adjustment is completed, the unqualified proportion is re-detected, and if the unqualified proportion is greater than a preset unqualified proportion, the version update frequency is determined based on the test point number, and the pressure plate timeout time is adjusted based on the network load rate; The service tracking logic node is configured in the SCD file, and the consistency of the GOOSE trigger condition and the actual signal is automatically verified through the MMS subscription dataset.

[0031] The application discloses a secondary equipment automatic test system applied to whole station test.

[0032] Specifically, the process of determining whether the operation of the single secondary equipment is qualified based on the voltage distortion rate comprises the following steps. The standard current voltage is injected through the tester, and the deviation of the measurement value of the checking device from the standard value is recorded as the voltage distortion rate. The operation of the secondary equipment is determined based on the comparison result of the voltage distortion rate with the first preset voltage distortion rate and the second preset voltage distortion rate. If the voltage distortion rate is greater than the first preset voltage distortion rate, it is determined that the operation of the secondary equipment is unqualified, and the secondary equipment is marked as an abnormal device. If the voltage distortion rate is greater than the second preset voltage distortion rate and less than or equal to the first preset voltage distortion rate, the operation of the secondary equipment is determined based on the frequency deviation. If the voltage distortion rate is less than or equal to the second preset voltage distortion rate, it is determined that the operation of the secondary equipment is qualified.

[0033] Specifically, in the embodiment, the tester is connected with the measured device, the three-phase voltage and three-phase current values are input on the interface parameter information by using the voltage current test unit, and are downloaded into the tester after running, so that the task of adding the voltage and current values to the protection device is completed, and the voltage distortion rate when the voltage and current values set by the protection device are consistent or inconsistent is checked in the panel menu of the protection device, wherein the first preset voltage distortion rate H0 is 5%, the second preset voltage distortion rate H1 is 2%, and the comparison result of the voltage distortion rate with the first preset voltage distortion rate and the second preset voltage distortion rate is as follows. If the voltage distortion rate is greater than the first preset voltage distortion rate H0, it is determined that the operation of the secondary equipment is unqualified, and a marking process is performed. If the voltage distortion rate is greater than the second preset voltage distortion rate H1 and less than or equal to the first preset voltage distortion rate, the operation of the secondary equipment is determined based on the frequency difference. If the voltage distortion rate is less than or equal to the second preset voltage distortion rate, it is determined that the operation of the secondary equipment is qualified.

[0034] The voltage distortion rate = [(measured value of the device - standard value of the tester) / standard value of the tester] * 100%. Specifically, in response to the first preset condition, the process of determining whether the operation of the secondary equipment is qualified based on the frequency deviation includes: In a single monitoring period, a plurality of monitoring time windows are set, and a plurality of detection time points are set in each time window, and the frequency values measured by the relay protection devices at each detection time point are detected; The frequency values measured by each of the relay protection devices are recorded as current frequencies, and the absolute values of the differences between each of the current frequencies and the rated frequency are recorded as frequency deviation values corresponding to each detection time point, and the average value of each frequency deviation value is recorded as the frequency deviation; If the frequency deviation is greater than the preset frequency deviation, it is determined that the operation of the secondary equipment is unqualified, and the secondary equipment is marked as an abnormal device; If the frequency deviation is less than or equal to the preset frequency deviation, it is determined that the operation of the secondary equipment is qualified; The first preset condition is that the voltage distortion rate is greater than the second preset voltage distortion rate and less than or equal to the first preset voltage distortion rate.

[0035] Specifically, in the embodiment of the present application, the length of a single time window is 1 min, and the number of detection time points in each time window is N=10, and the preset frequency deviation Q in the present application is 0.03 Hz; the comparison result based on the frequency deviation and the preset frequency deviation is as follows: If the frequency deviation is greater than the preset frequency deviation Q, it is determined that the operation of the secondary equipment is unqualified, and a marking process is performed; If the frequency deviation is less than or equal to the preset frequency deviation Q, it is determined that the operation of the secondary equipment is qualified. Specifically, the process of determining the unqualified reason based on the unqualified proportion in the test flow includes: The number of abnormal devices and the total number of devices are obtained, and the ratio of the number of abnormal devices to the total number of devices is recorded as the unqualified proportion; If the unqualified proportion is greater than the preset unqualified proportion, it is determined that the unqualified reason is a preset matching degree anomaly, and the preset matching degree is determined based on the version update frequency; If the unqualified proportion is less than or equal to the preset unqualified proportion, it is determined that the unqualified reason is a secondary equipment detection anomaly; If the unqualified proportion is 0, it is determined that the operation of the secondary equipment is qualified and the test is completed.

[0036] Specifically, in the embodiment of the present application, the preset unqualified proportion r0 is 0.78, and the comparison result based on the unqualified proportion and the preset unqualified proportion is as follows: If the unqualified proportion is greater than the preset unqualified proportion r0, it is determined that the unqualified reason is a test flow anomaly, and the preset matching degree is determined based on the version update frequency; If the unqualified proportion is less than or equal to a preset unqualified proportion r0, it is determined that the unqualified reason is a relay protection device abnormality, wherein if the unqualified proportion is 0, it is determined that the secondary equipment is qualified in operation and the test is completed.

[0037] When the unqualified reason is a secondary equipment detection abnormality, the virtual terminal connection between devices is verified, and the verification process includes: (1) a circuit breaker position signal and a locking signal of the intelligent terminal to the protection function; (2) a control command of the measurement and control function to the intelligent terminal; (3) a control command of the measurement and control function to the intelligent terminal; a circuit breaker, isolating switch position signal and alarm signal of the intelligent terminal to the measurement and control function; (4) a sampling value information of the merging unit to the protection and measurement and control functions; (5) an alarm signal of the interval merging unit to the test function; (6) an alarm signal of the TV merging unit to the measurement and control function.

[0038] After the virtual terminal connection is verified, the unqualified proportion is re-detected, and if the unqualified proportion is less than or equal to a preset unqualified proportion, it is determined that the unqualified reason is a test flow abnormality, and a preset matching degree is determined based on a version update frequency.

[0039] Specifically, in response to a second preset condition, the process of determining the preset matching degree includes: reducing the preset matching degree based on the version update frequency, and the reduction amplitude of the preset matching degree is in a positive correlation with the version update frequency; the second preset condition is that the unqualified reason is determined to be a test flow abnormality.

[0040] Specifically, in the embodiment of the application, the test terminal PC is connected to the device under test through a switch, and a calling button of an MMS communication command is provided in a toolbar on an integrated station test management interface, a graphical view interface and a list view interface, so that the MMS communication module can read the device data model file of the device under test; In the graphical view or the list view, the IED to be tested is determined, the device data model consistency detection function is selected, the device data model consistency detection interface is entered, the IED file generated after the device data model file of the device under test and the SCD are parsed is loaded on the interface, and the comparison test is performed; The background program tests and compares the two files, and feeds back the comparison result to the device data model consistency detection interface. In the comparison process of the two files, the consistency standard is adjusted to improve the consistency of the two files.

[0041] If the version update frequency is greater than a second preset version update frequency V0, the preset matching degree is reduced to 73% of the reference matching degree, and in the application, the second preset version update frequency V0 is 2 times per month; If the version update frequency is greater than the first preset version update frequency V1 and less than or equal to the second preset version update frequency, the preset matching degree is reduced to 84% of the reference matching degree, and in the embodiment of the application, the first preset version update frequency V1 = 1 time / month; If the version update frequency is less than or equal to the first preset version update frequency, the preset matching degree is reduced to 89% of the reference matching degree. The value of the reference matching degree can be adaptively set according to actual application requirements, and it can be understood that the higher the value of the reference matching degree, the greater the influence of the version update frequency on checking whether the SCD file and the device CID file are consistent, and the application provides a value Y0 = 90% of the reference matching degree.

[0042] Specifically, in the adjustment process of the preset matching degree, the application obtains the unqualified proportion, and takes the absolute value of the difference between the unqualified proportion and the preset unqualified proportion as an effective proportion difference; The adjusted preset matching degree is corrected based on the effective proportion difference to reduce the adjusted preset matching degree, and the correction amplitude is in a positive correlation with the effective proportion difference.

[0043] Specifically, in the embodiment of the application, if the effective proportion difference is greater than a second preset effective proportion difference W0, the preset matching degree is reduced to 91% of the adjusted preset matching degree, and in the application, the second preset effective proportion difference W0 = 0.06. If the effective proportion difference is greater than a first preset effective proportion difference W1 and less than or equal to the second preset effective proportion difference, the preset matching degree is reduced to 93% of the adjusted preset matching degree, and in the embodiment of the application, the first preset effective proportion difference W1 = 0.03. If the effective proportion difference is less than or equal to the first preset effective proportion difference, the preset matching degree is reduced to 96% of the adjusted preset matching degree.

[0044] Specifically, after the preset matching degree is adjusted, the application re-detects the unqualified proportion, and if the unqualified proportion is greater than the preset unqualified proportion, the version update frequency is determined based on the number of test points.

[0045] Specifically, the application determines the reduction amplitude of the update frequency based on the number of test points, and the reduction amplitude of the update frequency is in a positive correlation with the number of test points.

[0046] Specifically, in the embodiment of the application, if the number of test points is greater than a second preset number of test points E0, the version update frequency is reduced to 97% of the adjusted version update frequency, and in the application, the second preset number of test points E0 = 30. If the number of test points is greater than the first preset number of test points and less than or equal to the second preset number of test points, the version update frequency is reduced to 94% of the adjusted version update frequency, and in the application, the first preset number of test points E0=25. If the number of test points is less than or equal to the first preset number of test points, the version update frequency is reduced to 92% of the adjusted version update frequency.

[0047] Specifically, in the process of correcting the update detection frequency, the press plate timeout time is adjusted based on the network load rate, and the increase range of the press plate timeout time is positively correlated with the network load rate.

[0048] Specifically, in the embodiment of the application, the network load rate=(GOOSE message total flow+SV message total flow) / network available bandwidth*100%; if the network load rate is greater than the second preset network load rate, the increase range of the press plate timeout time is 10%, and in the application, the second preset network load rate G0=70%; If the network load rate is greater than the first preset network load rate and less than or equal to the second preset network load rate, the increase range of the press plate timeout time is 8%, and in the application, the first preset network load rate G1=30%; If the network load rate is less than or equal to the first preset network load rate, the increase range of the press plate timeout time is 5%.

[0049] Specifically, the application configures a service tracking logical node in an SCD file, and automatically verifies the consistency of the GOOSE trigger condition and the actual signal through MMS subscription data set.

[0050] Specifically, in the embodiment of the application, the test computer, the test instrument and the measured protection device form a closed loop through the switch, communication is completed with the test instrument by calling the test instrument control interface program, communication is completed with the digital protection device by calling the MMS communication program, automatic testing is performed according to the test template and the report template, and a standard test report is output after the testing is completed.

[0051] So far, the technical solutions of the application have been described in combination with the preferred embodiments shown in the drawings, but those skilled in the art can easily understand that the protection scope of the application is obviously not limited to these specific embodiments. Those skilled in the art can make equivalent changes or replacements to related technical features without departing from the principles of the application, and the technical solutions after the changes or replacements will fall within the protection scope of the application.

[0052] The above merely illustrates the preferred embodiments of the present application, and is not used to limit the present application; for those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A substation equipment anomaly detection method based on deep learning, characterized in that, The method comprises the following steps: determining whether the operation of a single secondary device is qualified based on the voltage distortion rate, and marking the unqualified secondary device as an abnormal device, wherein, for the secondary device with the voltage distortion rate greater than a second preset voltage distortion rate and less than or equal to a first preset voltage distortion rate, whether the operation of the secondary device is qualified is determined based on the frequency deviation, and the unqualified secondary device is marked as an abnormal device, and the secondary device is a relay protection device; determining the unqualified reason of the test process based on the unqualified proportion, wherein the unqualified reason is preset matching degree abnormality or secondary device detection abnormality; when the unqualified reason is preset matching degree abnormality, the preset matching degree is reduced based on the version update frequency, and the adjusted preset matching degree is corrected based on the effective proportion difference to reduce the adjusted preset matching degree; after the preset matching degree adjustment is completed, the unqualified proportion is re-detected, if the unqualified proportion is greater than a preset unqualified proportion, the version update frequency is determined based on the test point number, and the die plate timeout time is adjusted based on the network load rate; a service tracking logic node is configured in an SCD file, and the consistency between the GOOSE trigger condition and the actual signal is automatically verified through MMS subscription dataset. 2.The deep learning-based substation equipment anomaly detection method of claim 1, wherein, The process of determining whether the operation of a single secondary device is qualified based on the voltage distortion rate comprises the following steps: a standard current voltage is injected through a tester, and the deviation between the measured value of the checking device and the standard value is recorded as the voltage distortion rate; whether the operation of the secondary device is qualified is determined based on the comparison result of the voltage distortion rate with the first preset voltage distortion rate and the second preset voltage distortion rate; if the voltage distortion rate is greater than the first preset voltage distortion rate, it is determined that the operation of the secondary device is unqualified, and the secondary device is marked as an abnormal device; if the voltage distortion rate is greater than the second preset voltage distortion rate and less than or equal to the first preset voltage distortion rate, whether the operation of the secondary device is qualified is determined based on the frequency deviation; if the voltage distortion rate is less than or equal to the second preset voltage distortion rate, it is determined that the operation of the secondary device is qualified. 3.The substation equipment anomaly detection method based on deep learning according to claim 2, characterized in that, The process of determining whether the operation of a secondary device is qualified based on the frequency deviation in response to the first preset condition comprises the following steps: a plurality of monitoring time windows are set in a single monitoring period, and a plurality of detection time points are set in each time window, and the frequency value measured by the relay protection device at each detection time point is detected; the frequency value measured by each relay protection device is recorded as a current frequency, and the absolute value of the difference between each current frequency and the rated frequency is recorded as a frequency difference value corresponding to each detection time point, and the average value of each frequency difference value is recorded as a frequency deviation; if the frequency deviation is greater than a preset frequency deviation, it is determined that the operation of the secondary device is unqualified, and the secondary device is marked as an abnormal device; if the frequency deviation is less than or equal to the preset frequency deviation, it is determined that the operation of the secondary device is qualified; the first preset condition is that the voltage distortion rate is greater than the second preset voltage distortion rate and less than or equal to the first preset voltage distortion rate. 4.The substation equipment anomaly detection method based on deep learning according to claim 3, characterized in that, The process of determining the unqualified reason of the test process based on the unqualified proportion comprises the following steps: the number of abnormal devices and the total number of devices are obtained, and the ratio of the number of abnormal devices to the total number of devices is recorded as the unqualified proportion; If the unqualified proportion is greater than the preset unqualified proportion, it is determined that the unqualified reason is a preset matching degree anomaly, and the preset matching degree is determined based on the version update frequency; If the unqualified proportion is less than or equal to the preset unqualified proportion, it is determined that the unqualified reason is a secondary equipment detection anomaly; If the unqualified proportion is 0, it is determined that the secondary equipment is qualified and the test is completed. 5.The substation equipment anomaly detection method based on deep learning according to claim 4, characterized in that, In response to a second preset condition, the process of determining the preset matching degree includes: reducing the preset matching degree based on the version update frequency, and the reduction amplitude of the preset matching degree is positively correlated with the version update frequency; The second preset condition is that the unqualified reason is determined to be a preset matching degree anomaly. 6.The substation equipment anomaly detection method based on deep learning according to claim 5, wherein, In the adjustment process of the preset matching degree, the unqualified proportion is obtained, and the absolute value of the difference between the unqualified proportion and the preset unqualified proportion is recorded as the effective proportion difference; Based on the effective proportion difference, the adjusted preset matching degree is corrected to reduce the adjusted preset matching degree, and the correction amplitude is positively correlated with the effective proportion difference. 7.The substation equipment anomaly detection method based on deep learning according to claim 5, wherein, After the preset matching degree is adjusted, the unqualified proportion is re-detected, and if the unqualified proportion is greater than the preset unqualified proportion, the version update frequency is determined based on the number of test points. 8.The substation equipment anomaly detection method based on deep learning according to claim 7, characterized in that, The reduction amplitude of the version update frequency is determined based on the number of test points, and the reduction amplitude of the version update frequency is positively correlated with the number of test points. 9.The substation equipment anomaly detection method based on deep learning according to claim 7, wherein, In the process of correcting the version update frequency, the press plate timeout time is adjusted based on the network load rate, and the increase amplitude of the press plate timeout time is positively correlated with the network load rate. 10.The substation equipment anomaly detection method based on deep learning of the system of claim 1, wherein, The service tracking logical node is configured in the SCD file, and the consistency of the GOOSE trigger condition and the actual signal is automatically verified through MMS subscription dataset.

Images