A relay protection intelligent monitoring method and system

Abstract

The invention discloses a relay protection intelligent monitoring method and system, which relate to the technical field of relay protection, and comprise the following steps: obtaining working data of a relay protection device, the working data comprising power system monitoring data and supplementary monitoring data, the supplementary monitoring data being obtained by monitoring the relay protection device by setting corresponding sensors based on the need of data collection; marking abnormal working data from the working data, and annotating the abnormal working data to obtain abnormal position labels of the abnormal working data; the relay protection intelligent monitoring method and system, by analyzing the working data after the abnormality occurs, trains and establishes a relay protection device state judgment model, judges whether the protection action of the relay protection device is executed and whether it is successfully executed after the abnormality occurs, and reduces the loss caused by the failure to timely discover the protection action when the relay protection device should execute the protection action.

Description

Technical Field

[0001] The present invention relates to the technical field of relay protection, and in particular to a relay protection intelligent monitoring method and system. Background Art

[0002] Relay protection devices are used to monitor various faults and abnormal conditions in the power system to protect the safe and stable operation of the power system. They usually monitor the current, voltage, frequency and other parameters in the power system to detect the operating status of the system in real time. When the system encounters abnormal conditions such as overload, short circuit, ground fault, etc., the relay protection device will cut off the fault area in time according to the preset protection logic and parameters to protect the safe operation of the system.

[0003] The prior art with publication number CN115664009B discloses a relay protection online monitoring system based on big data, and the relay protection online monitoring system includes: a message acquisition unit, used to acquire the real-time data of the network message of the relay protection device; a message transcoding unit, used to convert each message parameter in the real-time data of the network message into a hexadecimal number to obtain the hexadecimal message data, and then combine every 4 adjacent numbers in the hexadecimal message data into a first transcoding, and then convert the first transcoding into a decimal second transcoding to obtain a second transcoding sequence; a probability prediction unit, used to input the second transcoding sequence into a pre-trained and converged protection action monitoring model to obtain the current protection action state probability of the relay protection device; a state determination unit, used to determine the protection action state of the relay protection device according to the current protection action state probability of the relay protection device, so as to perform online monitoring on the relay protection device.

[0004] However, in the prior art, the protected systems and circuits are usually monitored to ensure the safe operation of the system, and there is still a lack of monitoring means for the relay protection device itself. For example, when a fault occurs in the monitored system, the protection action of the relay protection device is triggered, but there is still a lack of monitoring means to determine whether the protection action of the relay protection device is triggered and whether it is successfully executed after being triggered. This may result in the protection action being unable to determine whether the protection action is effectively implemented, thereby missing the opportunity for remediation and expanding damage and losses. Summary of the invention

[0005] The object of the present invention is to provide a relay protection intelligent monitoring method and system to solve the above-mentioned deficiencies in the prior art.

[0006] In order to achieve the above object, the present invention provides the following technical solution: a relay protection intelligent monitoring method, comprising the following steps:

[0007] S1. Acquire working data of the relay protection device, wherein the working data includes power system monitoring data and supplementary monitoring data, wherein the power system monitoring data is the operating status data of the power system monitored by the relay protection device, and the supplementary monitoring data is obtained by monitoring the relay protection device by setting corresponding sensors based on the need for data collection;

[0008] S2. Marking abnormal work data from the work data, and annotating the abnormal work data to obtain abnormal position labels of the abnormal work data, and associating the abnormal position labels with corresponding abnormal work data, wherein data association can be achieved through SQL tools;

[0009] S3. Based on the working data, abnormal working data and abnormal position labels, a first machine learning model is trained to obtain an abnormality recognition model, wherein a normal data label is set to correspond to the working data without abnormality when training the first machine learning model;

[0010] S4, input the working data into the abnormal recognition model, identify and output the abnormal working data and the corresponding abnormal position label, and output the normal data label if no abnormal working data is identified;

[0011] S5. If abnormal working data is identified, first determine whether the relay protection device performs the corresponding protection action based on the supplementary monitoring data. If so, extract the working data, mark the working data after the protection action is successfully executed, obtain a normal protection label, and associate the normal protection label with the corresponding working data;

[0012] S6. Based on the working data after the abnormal working data is identified, the working data of the successful execution of the protection action, the normal protection label and the abnormal protection label, a second machine learning model is trained to obtain a relay protection device state judgment model, which is used to identify the working data of the successful execution of the protection action from the input working data after the abnormal working data is identified. If the working data of the successful execution of the protection action is identified, the normal protection label is output, otherwise, the abnormal protection label is output;

[0013] S7. Input the working data after the abnormal working data is identified into the state judgment model of the relay protection device to judge whether the protection of the relay protection device is normal.

[0014] Furthermore, S2 also includes the following steps: annotating the abnormal working data associated with the abnormal position label again to obtain the abnormal cause label, and associating the abnormal cause label with the corresponding abnormal position label, so that each abnormal working data corresponds to an abnormal position label and an abnormal cause label at the same time, which is used to indicate the abnormal position of each abnormal working data while indicating the specific abnormal cause of the abnormal position; because the abnormality of each abnormal position may be caused by multiple reasons, such as short circuit, open circuit, poor contact, etc. at the position, and the abnormal working data corresponding to different abnormal causes are also different, so they need to be analyzed and considered separately.

[0015] Furthermore, in the step S3, training the first machine learning model to obtain an abnormality recognition model further includes the following steps:

[0016] Get the corresponding exception location label and exception cause label when an exception occurs;

[0017] Based on the difference between the abnormal position label combination and the abnormal cause label, the abnormal data are classified to obtain multiple abnormal work data sets, wherein the abnormal position label combination has at least one abnormal position label. If the abnormal position labels and the corresponding abnormal cause labels of two abnormal work data are exactly the same, then the two abnormal work data belong to the same abnormal work data set, otherwise, they belong to two abnormal work data sets respectively; the features of the abnormal work data of each abnormal work data set are made as similar as possible, so as to facilitate the subsequent feature analysis of each abnormal work data set; for example, the abnormal position labels corresponding to two abnormal work data are both line one and line two, and the abnormal cause labels associated with line one and line two of one abnormal work data are both open circuit, and the abnormal cause label associated with line one of the other abnormal work data is short circuit and the abnormal cause label associated with line two is open circuit, then the two abnormal work data are classified into different abnormal work data sets;

[0018] Each abnormal working data set is analyzed separately to obtain the corresponding abnormal fitting function and abnormal decision boundary.

[0019] Furthermore, the step S4 further includes the following steps:

[0020] Based on the abnormal working data and abnormal decision boundary output by the abnormal recognition model, determine the abnormal decision boundary to which the abnormal working data belongs; that is, within which abnormal decision boundary the abnormal working data is located, the abnormal cause of the abnormal working data can be considered to be the abnormal position label combination and abnormal cause label corresponding to the abnormal decision boundary.

[0021] Output the abnormal cause label corresponding to the abnormal decision boundary to which the abnormal working data belongs and the abnormal position label associated with the abnormal cause label, that is, when outputting the abnormal cause label, the abnormal position label associated with the abnormal cause label is also output at the same time, so as to know the abnormal position of the abnormal cause corresponding to the abnormal cause label.

[0022] Furthermore, the method also includes establishing a visualization model of the power system and relay protection device for visually displaying the abnormal location and the cause of the abnormality. The visualization model can be a two-dimensional plane model or a three-dimensional solid model. The location of the abnormality can be conveniently and intuitively displayed based on the visualization model, wherein the visualization model can be manually drawn based on professional knowledge.

[0023] Furthermore, a visualization model of the power system and relay protection device is established to visualize the abnormal location and abnormal cause, which specifically includes the following steps:

[0024] Build a visual model;

[0025] The abnormal position label and the abnormal cause label are respectively associated with each position of the visualization model, so as to indicate the corresponding abnormal position on the visualization model and display the abnormal cause when the abnormal position label and the abnormal cause label are output. For example, the position of the visualization model corresponding to the abnormal position label can be represented by a color different from other positions, and the abnormal cause corresponding to the output abnormal cause label can be displayed near the position.

[0026] Furthermore, the method further includes predicting the abnormality based on the abnormal working data and the working data before the abnormal working data occurs, specifically including the following steps:

[0027] Obtain the abnormal location label and abnormal cause label corresponding to each abnormal working data, and obtain the abnormal location and abnormal cause;

[0028] Obtain working data of the abnormal position when it is working normally, and obtain normal comparison data;

[0029] Get the working data of the set time before the abnormal working data to get the abnormal preceding data;

[0030] Obtain the working data of the set time as mentioned above corresponding to the normal comparison data of the abnormal working data, and obtain the normal preceding data;

[0031] Based on the working data, abnormal preceding data, normal preceding data, abnormal prediction labels and normal prediction labels, a third machine learning model is trained to obtain an abnormal prediction model, which is used to determine whether the input working data is abnormal preceding data or normal preceding data, and output an abnormal prediction label if it is abnormal preceding data, and output a normal prediction label if it is normal preceding data.

[0032] A relay protection intelligent monitoring system, comprising a data acquisition module, a data annotation module, a power system anomaly identification module, a protection device anomaly identification module, a visualization model module, and an anomaly prediction module;

[0033] The data acquisition module is used to acquire the working data;

[0034] The data classification and labeling module is used to classify the working data into power system monitoring data and supplementary monitoring data according to different sources of the working data;

[0035] The data classification and annotation module is also used to annotate the work data and associate the annotated content with the corresponding work data;

[0036] The power system abnormality identification module is used to train a first machine learning model based on working data, abnormal working data and abnormal position labels to obtain an abnormality identification model;

[0037] The power system abnormality identification module is also used to call the abnormality identification model to identify abnormal working data in the input working data and locate the abnormal position;

[0038] The protection device abnormality identification module is used to train a second machine learning model based on the working data after the abnormal working data is identified, the working data of the successful execution of the protection action, the protection normal label and the protection abnormal label to obtain a relay protection device state judgment model;

[0039] The protection device abnormality identification module is also used to call the relay protection device state judgment model, identify the working data of the successful execution of the protection action based on the working data after the abnormal working data is identified, and output the protection normal label when the working data of the successful execution of the protection action is identified, and output the protection abnormal label when the working data of the successful execution of the protection action is not identified;

[0040] The visualization model module is used to set the visualization model of the power system and the relay protection device, and to visually display the visualization model and text;

[0041] The visualization model is provided with at least two different colors, and the visualization model module is further used to individually adjust the color of each position of the visualization model, and the color of each position of the visualization model corresponds to a normal state and an abnormal state respectively.

[0042] 1. Compared with the prior art, the intelligent monitoring method and system for relay protection provided by the present invention can automatically identify and separate abnormal working data and abnormal positions from working data by marking abnormal working data and associating them with corresponding abnormal positions, and training and establishing an abnormal recognition model.

[0043] 2. Compared with the prior art, the intelligent monitoring method and system for relay protection provided by the present invention analyzes the working data after the abnormality occurs, trains and establishes a state judgment model for the relay protection device, and judges whether the protection action of the relay protection device is executed and whether it is successfully executed after the abnormality occurs, thereby reducing the losses caused by the failure to execute the protection action or the failure to execute the protection action when the relay protection device should execute the protection action and the loss caused by the failure to execute the protection action in time.

[0044] 3. Compared with the prior art, the present invention provides an intelligent monitoring method and system for relay protection, which analyzes abnormal working data and its corresponding normal working data, and working data of a previous period, and trains an abnormal prediction model to predict the possibility of future abnormalities based on the working data, and take precautions against possible abnormalities in advance. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0046] Figure 1 A method step diagram provided for an embodiment of the present invention;

[0047] Figure 2 A system structure block diagram provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0049] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0050] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined. In addition, the terms "installed", "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal connection of two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0051] Example embodiments will be described more fully below with reference to the accompanying drawings, but the example embodiments may be embodied in different forms and should not be construed as limited to the embodiments set forth herein. On the contrary, the purpose of providing these embodiments is to make the present disclosure thorough and complete and to enable those skilled in the art to fully understand the scope of the present disclosure.

[0052] In the absence of conflict, the various embodiments of the present disclosure and the various features therein may be combined with each other.

[0053] As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0054] The terms used herein are only used to describe specific embodiments and are not intended to limit the present disclosure. As used herein, the singular forms "a", "an" and "the" are also intended to include the plural forms, unless the context clearly indicates otherwise. It will also be understood that when the terms "comprising" and / or "made of" are used in this specification, the presence of the features, wholes, steps, operations, elements and / or components is specified, but the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or groups thereof is not excluded.

[0055] The embodiments described herein may be described with reference to plan views and / or cross-sectional views by means of idealized schematic diagrams of the present disclosure. Therefore, the example illustrations may be modified according to manufacturing techniques and / or tolerances. Therefore, the embodiments are not limited to the embodiments shown in the accompanying drawings, but include modifications of the configurations formed based on the manufacturing process. Therefore, the regions illustrated in the accompanying drawings have schematic properties, and the shapes of the regions shown in the figures illustrate the specific shapes of the regions of the elements, but are not intended to be limiting.

[0056] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and the present disclosure, and will not be interpreted as having an idealized or overly formal meaning unless explicitly defined as such herein.

[0057] See also Figure 1-Figure 2 , a relay protection intelligent monitoring method, comprising the following steps:

[0058] S1. Obtain working data of the relay protection device, the working data including power system monitoring data and supplementary monitoring data, wherein the power system monitoring data is the operating status data of the power system monitored by the relay protection device, and the supplementary monitoring data is obtained by monitoring the relay protection device by setting corresponding sensors based on the needs of data collection.

[0059] S2. Mark abnormal work data from the work data, annotate the abnormal work data, obtain abnormal position labels of the abnormal work data, and associate the abnormal position labels with the corresponding abnormal work data, wherein the data association can be achieved through SQL tools;

[0060] The abnormal working data associated with the abnormal position label is annotated again to obtain the abnormal cause label, and the abnormal cause label is associated with the corresponding abnormal position label, that is, each abnormal working data corresponds to both the abnormal position label and the abnormal cause label, which is used to indicate the abnormal position of each abnormal working data while indicating the specific abnormal cause of the abnormal position; because the abnormality of each abnormal position may be caused by multiple reasons, such as short circuit, open circuit, poor contact, etc. at the position, and the abnormal working data corresponding to different abnormal causes are also different, so they need to be analyzed and considered separately.

[0061] S3. Based on the working data, abnormal working data and abnormal position labels, a first machine learning model is trained to obtain an abnormality recognition model, wherein a normal data label can also be set to correspond to the working data when there is no abnormality when training the first machine learning model; the present invention does not limit the specific machine learning model, and it can be selected as needed so that the output accuracy of the trained model meets the required accuracy, for example, it can be a CNN model, a random forest model, a time series model, etc.;

[0062] The following steps are also included:

[0063] Get the corresponding exception location label and exception cause label when an exception occurs;

[0064] Based on the difference between the abnormal position label combination and the abnormal cause label, the abnormal data are classified to obtain multiple abnormal work data sets, wherein the abnormal position label combination has at least one abnormal position label. If the abnormal position labels and the corresponding abnormal cause labels of two abnormal work data are exactly the same, then the two abnormal work data belong to the same abnormal work data set, otherwise, they belong to two abnormal work data sets respectively; the features of the abnormal work data of each abnormal work data set are made as similar as possible, so as to facilitate the subsequent feature analysis of each abnormal work data set; for example, the abnormal position labels corresponding to two abnormal work data are both line one and line two, and the abnormal cause labels associated with line one and line two of one abnormal work data are both open circuit, and the abnormal cause label associated with line one of the other abnormal work data is short circuit and the abnormal cause label associated with line two is open circuit, then the two abnormal work data are classified into different abnormal work data sets;

[0065] Each abnormal working data set is analyzed separately to obtain the corresponding abnormal fitting function and abnormal decision boundary.

[0066] S4, input the working data into the abnormal recognition model, identify and output the abnormal working data and the corresponding abnormal position label, and if no abnormal working data is identified, output the normal data label;

[0067] The following steps are also included:

[0068] Based on the abnormal working data and abnormal decision boundary output by the abnormal recognition model, determine the abnormal decision boundary to which the abnormal working data belongs; that is, within which abnormal decision boundary the abnormal working data is located, the abnormal cause of the abnormal working data can be considered to be the abnormal position label combination and abnormal cause label corresponding to the abnormal decision boundary.

[0069] Output the abnormal cause label corresponding to the abnormal decision boundary to which the abnormal working data belongs and the abnormal position label associated with the abnormal cause label, that is, when outputting the abnormal cause label, the abnormal position label associated with the abnormal cause label is also output at the same time, so as to know the abnormal position of the abnormal cause corresponding to the abnormal cause label.

[0070] S5. If abnormal working data is identified, first determine whether the relay protection device performs the corresponding protection action based on the supplementary monitoring data, wherein the supplementary monitoring data can be analyzed to directly determine whether the protection action is performed, and then determine whether the protection action is successfully performed through the following steps: if the corresponding protection action is performed, extract the working data, mark the working data after the protection action is successfully performed, obtain a normal protection tag, and associate the normal protection tag with the corresponding working data;

[0071] S6. Based on the working data after the abnormal working data is identified, the working data of the successful execution of the protection action, the normal protection label and the abnormal protection label, a second machine learning model is trained to obtain a relay protection device state judgment model, which is used to identify the working data of the successful execution of the protection action from the input working data after the abnormal working data is identified. If the working data of the successful execution of the protection action is identified, the normal protection label is output, otherwise, the abnormal protection label is output;

[0072] S7. Input the working data after the abnormal working data is identified into the state judgment model of the relay protection device to judge whether the protection of the relay protection device is normal.

[0073] S8. Predicting the abnormality based on the abnormal working data and the working data before the abnormal working data occurs, specifically comprising the following steps:

[0074] Obtain the abnormal location label and abnormal cause label corresponding to each abnormal working data, and obtain the abnormal location and abnormal cause;

[0075] Obtain working data of the abnormal position when it is working normally, and obtain normal comparison data;

[0076] Get the working data of the set time before the abnormal working data to get the abnormal preceding data;

[0077] Obtain working data of a set time period before normal comparison data corresponding to abnormal working data to obtain normal preceding data; wherein the normal preceding data under normal working state is selected as sample data to avoid increasing the error;

[0078] Based on the working data, abnormal preceding data, normal preceding data, abnormal prediction labels and normal prediction labels, a third machine learning model is trained to obtain an abnormal prediction model, which is used to determine whether the input working data is abnormal preceding data or normal preceding data, and output an abnormal prediction label if it is abnormal preceding data, and output a normal prediction label if it is normal preceding data.

[0079] S9. Establish a visualization model of the power system and relay protection device to visualize the abnormal location and abnormal cause. The visualization model can be a two-dimensional plane model or a three-dimensional model. The location of the abnormality can be displayed conveniently and intuitively based on the visualization model. The visualization model can be manually drawn based on professional knowledge. Specifically, the following steps are included:

[0080] Build a visual model;

[0081] The abnormal position label and the abnormal cause label are respectively associated with each position of the visualization model, so as to indicate the corresponding abnormal position on the visualization model and display the abnormal cause when the abnormal position label and the abnormal cause label are output. For example, the position of the visualization model corresponding to the abnormal position label can be represented by a color different from other positions, and the abnormal cause corresponding to the output abnormal cause label can be displayed near the position.

[0082] A relay protection intelligent monitoring system, comprising a data acquisition module, a data annotation module, a power system anomaly identification module, a protection device anomaly identification module, a visualization model module, and an anomaly prediction module;

[0083] The data acquisition module is used to acquire working data;

[0084] The data classification and labeling module is used to classify the working data into power system monitoring data and supplementary monitoring data according to the different sources of the working data;

[0085] The data classification and annotation module is also used to annotate the work data and associate the annotation content with the corresponding work data, such as marking the abnormal work data in the work data, marking the abnormal position corresponding to the abnormal work data, setting the abnormal position label corresponding to the abnormal position to be associated with the abnormal work data; annotating and setting the normal protection label to be associated with the work data where the protection action is successfully executed; annotating and setting the abnormal prediction label to be associated with the abnormal preceding data; annotating and setting the normal prediction label to be associated with the normal preceding data;

[0086] The power system anomaly identification module is used to train a first machine learning model based on working data, abnormal working data and abnormal position labels to obtain an anomaly identification model;

[0087] The power system anomaly identification module is also used to call the anomaly identification model to identify abnormal working data in the input working data and locate the abnormal position;

[0088] The protection device abnormality identification module is used to train the second machine learning model based on the working data after the abnormal working data is identified, the working data of the successful execution of the protection action, the protection normal label and the protection abnormal label to obtain the relay protection device state judgment model;

[0089] The protection device abnormality identification module is also used to call the relay protection device state judgment model, identify the working data of the successful execution of the protection action based on the working data after the abnormal working data is identified, and output the protection normal label when the working data of the successful execution of the protection action is identified, and output the protection abnormal label when the working data of the successful execution of the protection action is not identified;

[0090] The visualization model module is used to set the visualization model of the power system and relay protection device, and to visually display the visualization model and text;

[0091] The visualization model is provided with at least two different colors. The visualization model module is also used to individually adjust the color of each position of the visualization model. The color of each position of the visualization model corresponds to a normal state and an abnormal state, respectively.

[0092] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A relay protection intelligent monitoring method, characterized in that: The following steps are involved: S1. Acquire working data of the relay protection device, wherein the working data includes power system monitoring data and supplementary monitoring data, wherein the supplementary monitoring data is obtained by monitoring the relay protection device by setting corresponding sensors based on the need for data collection; S2. Mark abnormal working data from the working data, annotate the abnormal working data, obtain abnormal position labels of the abnormal working data, and associate the abnormal position labels with corresponding abnormal working data; S3. Based on the working data, abnormal working data and abnormal position labels, a first machine learning model is trained to obtain an abnormality recognition model; S4, inputting the working data into the abnormality recognition model, identifying and outputting the abnormal working data and the corresponding abnormal position label; S5. If abnormal working data is identified, first determine whether the relay protection device performs the corresponding protection action based on the supplementary monitoring data. If so, extract the working data, mark the working data after the protection action is successfully executed, obtain a normal protection label, and associate the normal protection label with the corresponding working data; S6. Based on the working data after the abnormal working data is identified, the working data of the successful execution of the protection action, the normal protection label and the abnormal protection label, a second machine learning model is trained to obtain a relay protection device state judgment model, which is used to identify the working data of the successful execution of the protection action from the input working data after the abnormal working data is identified. If the working data of the successful execution of the protection action is identified, the normal protection label is output, otherwise, the abnormal protection label is output; S7. Input the working data after the abnormal working data is identified into the state judgment model of the relay protection device to judge whether the protection of the relay protection device is normal.

2. The intelligent monitoring method for relay protection according to claim 1, characterized in that: The S2 further includes the following steps: annotating the abnormal working data associated with the abnormal position tag again to obtain an abnormal cause tag, and associating the abnormal cause tag with the corresponding abnormal position tag.

3. The intelligent monitoring method for relay protection according to claim 2, characterized in that: The step S3 in which the first machine learning model is trained to obtain an abnormality recognition model further includes the following steps: Get the corresponding exception location label and exception cause label when an exception occurs; Based on the difference in abnormal position label combinations and abnormal cause labels, the abnormal data is classified to obtain multiple abnormal work data sets; Each abnormal working data set is analyzed separately to obtain the corresponding abnormal fitting function and abnormal decision boundary.

4. The intelligent monitoring method for relay protection according to claim 3 is characterized in that: The step S4 further comprises the following steps: Based on the abnormal working data and the abnormal decision boundary output by the abnormal recognition model, determining the abnormal decision boundary to which the abnormal working data belongs; Output the abnormal cause label corresponding to the abnormal decision boundary to which the abnormal working data belongs and the abnormal position label associated with the abnormal cause label.

5. A relay protection intelligent monitoring method according to claim 4, characterized in that: The method further includes establishing a visualization model of the power system and the relay protection device for visually displaying the abnormal location and the abnormal cause.

6. A relay protection intelligent monitoring method according to claim 5, characterized in that: Establishing a visualization model of the power system and relay protection device to visualize the abnormal location and cause of the abnormality includes the following steps: Build a visual model; The abnormal position label and the abnormal cause label are respectively associated with each position of the visualization model, so as to indicate the corresponding abnormal position on the visualization model and display the abnormal cause when the abnormal position label and the abnormal cause label are output.

7. The intelligent monitoring method for relay protection according to claim 4, characterized in that: The method further includes predicting the abnormality based on the abnormal working data and the working data before the abnormal working data occurs, specifically including the following steps: Obtain the abnormal location label and abnormal cause label corresponding to each abnormal working data, and obtain the abnormal location and abnormal cause; Obtain working data of the abnormal position when it is working normally, and obtain normal comparison data; Get the working data of the set time before the abnormal working data to get the abnormal preceding data; Obtain the working data of the set time as mentioned above corresponding to the normal comparison data of the abnormal working data, and obtain the normal preceding data; Based on the working data, abnormal preceding data, normal preceding data, abnormal prediction labels and normal prediction labels, a third machine learning model is trained to obtain an abnormal prediction model, which is used to determine whether the input working data is abnormal preceding data or normal preceding data, and output an abnormal prediction label if it is abnormal preceding data, and output a normal prediction label if it is normal preceding data.

8. A relay protection intelligent monitoring system, used to execute a relay protection intelligent monitoring method according to any one of claims 1 to 7, characterized in that: It includes data acquisition module, data classification and annotation module, power system anomaly identification module, protection device anomaly identification module, visualization model module, and anomaly prediction module; The data acquisition module is used to acquire the working data; The data classification and labeling module is used to classify the working data into power system monitoring data and supplementary monitoring data according to different sources of the working data; The data classification and annotation module is also used to annotate the work data and associate the annotated content with the corresponding work data; The power system abnormality identification module is used to train a first machine learning model based on working data, abnormal working data and abnormal position labels to obtain an abnormality identification model; The power system abnormality identification module is also used to call the abnormality identification model to identify abnormal working data in the input working data and locate the abnormal position; The protection device abnormality identification module is used to train a second machine learning model based on the working data after the abnormal working data is identified, the working data of the successful execution of the protection action, the protection normal label and the protection abnormal label to obtain a relay protection device state judgment model; The protection device abnormality identification module is also used to call the relay protection device state judgment model, and identify the working data of successful execution of the protection action based on the working data after the abnormal working data is identified; The visualization model module is used to set the visualization model of the power system and the relay protection device, and to visually display the visualization model and text; The visualization model is provided with at least two different colors, and the visualization model module is further used to individually adjust the color of each position of the visualization model, and the color of each position of the visualization model corresponds to a normal state and an abnormal state respectively.

Images