System exception handling method, device, equipment, medium and product

Abstract

The invention discloses a system exception handling method, device and equipment, a medium and a product. The method comprises the steps of determining a current model file and a current program version corresponding to target equipment in the substation secondary system in response to an anomaly detection request for the substation secondary system; according to the current program version, determining a standard model file from a preset standard database, and based on a preset text analysis model, respectively determining a standard semantic vector and a current semantic vector corresponding to the standard model file and the current model file; and performing feature fine-grained matching interaction on the standard semantic vector and the current semantic vector to obtain a semantic generality vector, and determining a difference semantic vector and a corresponding exception level according to the semantic generality vector and the current semantic vector to perform exception processing. According to the technical scheme, the file of the transformer substation system can be analyzed and compared through the text analysis model so as to carry out anomaly detection on the system, and therefore targeted anomaly processing is facilitated.

Description

Technical Field

[0001] The present invention relates to the field of power grid technology, and in particular to a system abnormality processing method, device, equipment, medium and product. Background Art

[0002] With the rapid development of power systems, the automation level of substations is increasing day by day. As an important part of the power system, the operating status of the substation secondary system is directly related to the safe and stable operation of the power grid. In the substation secondary system, various automation equipment (such as relay protection devices, monitoring systems, communication equipment, etc.) need to work together to achieve effective control and monitoring of primary equipment (such as transformers, circuit breakers, etc.). In order to ensure the reliability and consistency of information interaction between these devices, the model files and software versions of the secondary system need to be strictly managed and verified. However, when the number of system files is large and the content is complex, manual verification is prone to omissions or misjudgments, which in turn affects the stability of the entire substation secondary system.

[0003] Therefore, how to use the text parsing model to parse and compare the files of the substation system to detect anomalies in the system, thereby facilitating targeted anomaly handling, is a problem that needs to be solved urgently. Summary of the invention

[0004] The present invention provides a system exception handling method, device, equipment, medium and product, which utilize a text parsing model to parse and compare files of a substation system to detect anomalies in the system, thereby facilitating targeted anomaly handling.

[0005] According to one aspect of the present invention, a system exception handling method is provided, comprising:

[0006] In response to a request for abnormality detection of a substation secondary system, determining a current model file and a current program version corresponding to a target device in the substation secondary system;

[0007] According to the current program version, determine the standard model file from the preset standard database, and determine the standard semantic vector and the current semantic vector corresponding to the standard model file and the current model file respectively based on the preset text parsing model;

[0008] The standard semantic vector and the current semantic vector are interactively matched with each other in fine-grained features to obtain a semantic commonality vector, and the difference semantic vector and the corresponding exception level are determined based on the semantic commonality vector and the current semantic vector for exception handling.

[0009] According to another aspect of the present invention, there is provided a system exception handling device, comprising:

[0010] A file determination module, for determining a current model file and a current program version corresponding to a target device in the secondary system of the substation in response to an abnormality detection request for the secondary system of the substation;

[0011] A vector determination module, used to determine a standard model file from a preset standard database according to the current program version, and to determine a standard semantic vector and a current semantic vector corresponding to the standard model file and the current model file respectively based on a preset text parsing model;

[0012] The processing module is used to perform feature fine-grained matching interaction on the standard semantic vector and the current semantic vector to obtain a semantic commonality vector, and determine the difference semantic vector and the corresponding exception level based on the semantic commonality vector and the current semantic vector for exception processing.

[0013] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0014] at least one processor; and

[0015] a memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the system exception handling method described in any embodiment of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the system exception handling method described in any embodiment of the present invention when executed.

[0018] According to another aspect of the present invention, a computer program product is provided. The computer program product includes a computer program. When the computer program is executed by a processor, the system exception handling method according to any embodiment of the present invention is implemented.

[0019] The technical solution of the embodiment of the present invention, in response to an abnormality detection request for a substation secondary system, determines the current model file and the current program version corresponding to the target device in the substation secondary system; determines the standard model file from a preset standard database according to the current program version, and determines the standard semantic vector and the current semantic vector corresponding to the standard model file and the current model file respectively based on a preset text parsing model; performs feature fine-grained matching interaction on the standard semantic vector and the current semantic vector to obtain a semantic commonality vector, and determines the difference semantic vector and the corresponding abnormality level according to the semantic commonality vector and the current semantic vector to perform abnormality processing. By using the text parsing model to parse and compare the files of the substation system, the system can be effectively detected for abnormalities, which helps to perform targeted abnormality processing.

[0020] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0022] Figure 1 is a flow chart of a system exception handling method provided in Embodiment 1 of the present invention;

[0023] Figure 2 is a flow chart of a system exception handling method provided by Embodiment 2 of the present invention;

[0024] Figure 3 It is a structural block diagram of a system exception handling device provided by Embodiment 3 of the present invention;

[0025] Figure 4 It is a schematic diagram of the structure of an electronic device provided in Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0027] It should be noted that the terms "first", "second", "target", "candidate", "alternative", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. The acquisition, storage, use, processing, etc. of data in the technical solution of this application comply with the relevant provisions of national laws and regulations.

[0028] Embodiment 1

[0029] Figure 1 is a flowchart of a system abnormality handling method provided by Embodiment 1 of the present invention; this embodiment is applicable to the case where abnormality detection is performed on a model file involved in a substation secondary system. The method can be executed by a system abnormality handling device. The system abnormality handling device can be implemented in the form of hardware and / or software. The system abnormality handling device can be configured in an electronic device, such as a substation secondary system, such as Figure 1 As shown, the system exception handling method includes:

[0030] S101 . In response to a request for abnormality detection on a substation secondary system, determine a current model file and a current program version corresponding to a target device in the substation secondary system.

[0031] Among them, the substation secondary system can include various automation equipment, such as relay protection devices, monitoring systems, communication equipment, etc. Through the collaborative work between different automation equipment, the substation secondary system can achieve effective control and monitoring of primary equipment (such as transformers, circuit breakers, etc.). Target equipment can include automation equipment and primary equipment. Anomaly detection request refers to a request to check the model files corresponding to each target device in the substation secondary system for anomaly detection.

[0032] The current model file refers to the model file currently used by the target device in the substation. The current model file may include an IED (Intelligent Electronic Device) configuration file and / or an SCD (Substation Configuration Description) file. The IED configuration file is an important document used to describe and configure intelligent electronic devices. These files usually contain information such as the functions, parameters, communication protocols, data point mapping, etc. of the device, and are the basis for ensuring that the IED works correctly in the substation automation system; the SCD file is an important file in the substation automation system for describing the configuration of the entire system. The SCD file follows the IEC 61850 standard (a global standard in the field of power system automation) and is used to define the configuration, communication parameters, and logical relationships between all intelligent electronic devices in the substation. The SCD file plays a key role in system design, debugging, operation, maintenance, and management. The current program version refers to the program version currently running on the target device in the substation, and the current program version corresponds to the target device one-to-one.

[0033] Optionally, the substation secondary system may periodically generate a request for verification of the current model file to detect anomalies, that is, detect an abnormality detection request for the substation secondary system. It may also be considered that an abnormality detection request for the substation secondary system is detected when a detection instruction issued by relevant personnel is detected, thereby performing an abnormality detection in response to the abnormality detection request for the substation secondary system.

[0034] S102. According to the current program version, determine the standard model file from the preset standard database, and based on the preset text parsing model, determine the standard semantic vector and the current semantic vector corresponding to the standard model file and the current model file, respectively.

[0035] Optionally, the current program version may be parsed to determine the model information corresponding to the current program version, and further combined with the correspondence between the model information and the model file, a standard model file that complies with the current program version of each target device may be determined from a preset standard database.

[0036] Optionally, based on a preset text parsing model, the standard semantic vector and the current semantic vector corresponding to the standard model file and the current model file are respectively determined, including: based on a pre-trained word embedding model and sentence embedding model, determining the standard word vector and the standard sentence vector corresponding to the standard model file, and determining the current word vector and the current sentence vector corresponding to the current model file; based on a preset semantic model, semantically encoding the standard word vector and the standard sentence vector to obtain a standard semantic vector, and semantically encoding the current word vector and the current sentence vector to obtain a current semantic vector.

[0037] Among them, the preset text parsing model can be a pre-trained word embedding model, a sentence embedding model, and a preset semantic model. The pre-trained word embedding model can specifically be a Word2Vec (word to vector) or GloVe (GlobalVectors for Word Representation) model. The sentence embedding model can be a BERT (Bidirectional Encoder Representations from Transformers) or a RoBERTa model (Robustly OptimizedBERT Pretraining Approach). The semantic model can be an LSTM (Long Short Term Memory), GRU (Gated Recurrent Unit), or Transformer model.

[0038] Exemplarily, a pre-trained word embedding model and a pre-trained sentence embedding model can be used to convert word sequences and the entire file content into word vectors and sentence vectors, the segmented word sequence is input into the word embedding model to generate a vector representation of each word, and the pre-trained sentence embedding model is used to convert the entire file content into a sentence vector, and then the average value of all word vectors in the file is used as the vector representation of the file, and the word vectors are weighted averaged according to the importance of the words (such as TF-IDF (Term Frequency-Inverse Document Frequency)) to generate a vector representation of the file; a deep learning model (i.e., a preset semantic model) is used to further semantically encode the word vectors or sentence vectors; the pre-trained semantic model is used to semantically encode the file content; and the word vectors or sentence vectors are input into the selected semantic model to obtain a standard semantic vector and a current semantic vector.

[0039] S103, performing feature fine-grained matching interaction on the standard semantic vector and the current semantic vector to obtain a semantic commonality vector, and determining a difference semantic vector and a corresponding abnormality level based on the semantic commonality vector and the current semantic vector to perform abnormality processing.

[0040] Among them, the standard semantic vector refers to the vector obtained by semantically parsing the standard model file, and the current semantic vector refers to the vector obtained by semantically parsing the current model file. The semantic commonality vector refers to the vector determined by the standard semantic vector and the current semantic vector that meet the similarity conditions. The difference semantic vector refers to the deviation vector between the semantic commonality vector and the current semantic vector. The exception level refers to the level of severity of the exception determined according to the size of the deviation vector. The exception level can be, for example, an error exception, a warning exception, and a prompt exception.

[0041] Optionally, the standard semantic vector and the current semantic vector can be respectively subjected to fine-grained feature decoupling to obtain a standard sub-feature collection and a current sub-feature collection. Furthermore, the standard sub-feature collection and the current sub-feature collection can be subjected to optimal feature sub-component pairing screening and interactive fusion processing to obtain a semantic commonality vector.

[0042] Optionally, perform fine-grained feature matching interaction on the standard semantic vector and the current semantic vector to obtain a semantic commonality vector, including: perform fine-grained feature decoupling on the standard semantic vector and the current semantic vector to respectively obtain a standard sub-feature collection corresponding to the standard semantic vector and a current sub-feature collection corresponding to the current semantic vector; determine at least two groups of candidate feature pairs based on the standard sub-feature collection and the current sub-feature collection, and determine a hyperbolic space distance measurement factor for each group of candidate feature pairs to determine a target feature pair from the candidate feature pairs; determine an interactive fusion vector of the target feature pair, and perform a cascade operation on the interactive fusion vector to obtain a semantic commonality vector.

[0043] Among them, feature fine-grained decoupling refers to the operation of decomposing the semantic encoding representation vector into multiple sub-component feature vectors through fine-grained matching technology. A candidate feature pair can be composed of a pair of sub-feature combinations, specifically a standard sub-feature and a current sub-feature. The target feature pair refers to the candidate feature pair corresponding to the optimal pairing position.

[0044] Exemplarily, the current semantic vector is subjected to fine-grained feature decoupling to obtain the current sub-feature set corresponding to the current semantic vector, which can be expressed by the following formula:

[0045] Decouple(v1)={v 11 ,v 12 ,...,v 1i ,...,v 1n}

[0046] Exemplarily, the standard semantic vector is subjected to fine-grained feature decoupling to obtain a standard sub-feature set corresponding to the standard semantic vector, which can be expressed by the following formula:

[0047] Decouple(v2) = {v 21 ,v 22 ,...,v 2i ,...,v 2n}

[0048] Among them, v1 is the current semantic vector corresponding to the current model file, v2 is the standard semantic vector corresponding to the standard model file, Decouple is the fine-grained feature decoupling operation, and v 11 ,v 12 ,v 1i ,v1n are the first, second, i-th and n-th current sub-feature vectors in the current sub-feature collection, respectively, v 21 ,v 22 ,v 2i ,v 2n The first, second, i-th and n-th standard sub-feature vectors in the standard sub-feature collection respectively.

[0049] Exemplarily, the hyperbolic space distance metric factor d of each set of candidate feature pairs can be determined based on the following formula: P (v 1i ,v 2j ):

[0050]

[0051] Among them, (v 1i ,v 2j ) represents the candidate feature pair, v 1i represents the i-th current sub-feature in the current sub-feature collection, v 2j represents the jth standard sub-feature in the standard sub-feature set, ‖·‖ represents the vector norm, arccosh is the inverse hyperbolic cosine function, d P (v 1i ,v 2j ) is v 1i and v 2j The hyperbolic space distance metric between them.

[0052] Optionally, after determining the hyperbolic space distance measurement factor of each group of candidate feature pairs, the minimum hyperbolic space distance measurement factor can be taken as the optimal pairing position value, and the candidate feature pair corresponding to the optimal pairing position value can be further taken as the target feature pair.

[0053] For example, according to the hyperbolic space distance measurement factor of each set of candidate feature pairs, d P (v 1i ,v 2j ) The j parameter corresponding to the minimum value is taken as the optimal pairing position value k:

[0054]

[0055] Optionally, determining the interactive fusion vector of the target feature pair includes: performing subtraction operations, addition operations and multiplication operations on the target feature pair according to preset subtraction weights, preset addition weights and preset multiplication weights, respectively, to obtain a sum matrix, a deviation matrix and a product matrix of the target feature pair; and determining the vector matrix obtained by adding the sum matrix, the deviation matrix and the product matrix as the interactive fusion vector of the target feature pair.

[0056] Exemplarily, the interactive fusion vector v of the target feature pair can be determined based on the following formula: pi :

[0057]

[0058] Among them, v 1i represents the i-th current sub-feature in the current sub-feature collection, v 2k Represents the kth standard sub-feature in the standard sub-feature collection. ⊙ and They are positional subtraction, positional dot multiplication and positional addition, i.e., subtraction operation, addition operation and multiplication operation. α, β and γ are preset subtraction weights, preset addition weights and preset multiplication weights respectively.

[0059] Optionally, after determining the interactive fusion vector of the target feature pair, a cascade operation may be performed on the interactive fusion vector based on the following formula to obtain a semantic commonality vector:

[0060] v f =[v p1 ;v p2 ;...v pn ]

[0061] Among them, [·;·;·] is the cascade operation of vectors, v f represents the semantic commonality vector, v pi is the i-th interaction fusion vector.

[0062] Optionally, an embodiment of the present invention may adopt the above-mentioned feature fine-grained matching interaction scheme based on the hyperbolic space distance measurement factor to determine the semantic commonality vector, or may adopt the feature fine-grained matching interaction scheme based on similarity to determine the semantic commonality vector. It may also adopt both schemes to respectively determine the semantic commonality vector, and then take the average value of the semantic commonality vectors determined by the two schemes as the final semantic commonality vector obtained by performing feature fine-grained matching interaction.

[0063] Optionally, a similarity-based feature fine-grained matching interaction scheme can be used to determine the semantic commonality vector, and accordingly, a feature fine-grained matching interaction is performed on the standard semantic vector and the current semantic vector to obtain the semantic commonality vector, including: determining the word similarity between each standard word vector in the standard semantic vector and each current word vector in the current semantic vector, and determining the sentence similarity between each standard sentence vector in the standard semantic vector and each current sentence vector in the current semantic vector; determining the target word vector pair and the target sentence vector based on the word similarity, the sentence similarity and a preset similarity threshold, and performing aggregation processing on the target word vector pair and the target sentence vector to obtain the semantic commonality vector.

[0064] Optionally, the standard word vector and the current word vector may be aligned word by word, and the similarity between each set of word vector pairs, i.e., word similarity, may be determined; the standard sentence vector and the current sentence vector may be aligned sentence by sentence, and the similarity between each set of sentence vector pairs, i.e., sentence similarity, may be determined.

[0065] Optionally, a word vector whose word similarity is greater than a preset similarity threshold may be determined as a target word vector, and a sentence vector whose sentence similarity is greater than a preset similarity threshold may be determined as a target sentence vector.

[0066] Optionally, the average value of the target word vectors or the average value of the target sentence vectors can be directly determined as the semantic commonality vector; or weighted averaging can be performed according to the importance of each target word vector and target sentence vector (such as TF-IDF) to obtain the semantic commonality vector, which is not limited to the present invention.

[0067] Optionally, a difference semantic vector and a corresponding abnormality level are determined based on the semantic commonality vector and the current semantic vector, including: determining the difference semantic vector based on a difference component extractor of a preset differential network according to the semantic commonality vector and the current semantic vector; and determining the abnormality level corresponding to the difference semantic vector based on a preset difference grading model for abnormality processing.

[0068] Among them, the difference semantic vector corresponds to the difference caused by different model parameter settings or missing functional modules.

[0069] Optionally, the semantic commonality vector and the current semantic vector can be input into a difference component extractor of a preset differential network to determine a deviation vector between the semantic commonality vector and the current semantic vector, i.e., a difference semantic vector. Further, based on the difference semantic vector and a preset difference grading model, the difference size of the difference semantic vector is evaluated to determine the abnormality level corresponding to the difference semantic vector.

[0070] The technical solution of the embodiment of the present invention, in response to an abnormality detection request for a substation secondary system, determines the current model file and the current program version corresponding to the target device in the substation secondary system; determines the standard model file from a preset standard database according to the current program version, and determines the standard semantic vector and the current semantic vector corresponding to the standard model file and the current model file respectively based on a preset text parsing model; performs feature fine-grained matching interaction on the standard semantic vector and the current semantic vector to obtain a semantic commonality vector, and determines the difference semantic vector and the corresponding abnormality level according to the semantic commonality vector and the current semantic vector to perform abnormality processing. By using the text parsing model to parse and compare the files of the substation system, the system can be effectively detected for abnormalities, which helps to perform targeted abnormality processing.

[0071] Embodiment 2

[0072] Figure 2 is a flowchart of a system abnormality processing method provided in Embodiment 2 of the present invention; based on the above embodiments, this embodiment provides a preferred example of checking a model file to evaluate whether a substation secondary system is abnormal, specifically, Figure 2 As shown, the method includes the following process:

[0073] Optionally, text parsing is performed on the currently used model file and the standard model file respectively to obtain the text semantic encoding representation vector of the currently used model file (i.e., the current semantic vector) and the text semantic encoding representation vector of the standard model file (i.e., the standard semantic vector), respectively. Furthermore, the text semantic encoding representation vector of the currently used model file and the text semantic encoding representation vector of the standard model file can be interactively matched with each other in terms of features to obtain the semantic commonality encoding representation vector of the currently used model file-standard model file (i.e., the semantic commonality vector).

[0074] Optionally, an evaluation result of a level label used to represent the difference can be generated based on the differential component semantics between the currently used model file-standard model file semantic commonality component encoding representation vector (i.e., semantic commonality vector) and the currently used model file text semantic encoding representation vector (i.e., current semantic vector), thereby obtaining an exception level for exception handling.

[0075] Optionally, the semantic commonality vector may be determined by the following method based on a hyperbolic space distance metric factor:

[0076] 1. Calculate the hyperbolic space distance measurement factor between any two of the set of the current model file text semantic sub-component feature vectors and the set of the standard model file text semantic sub-component feature vectors;

[0077] Specifically, the hyperbolic space distance measurement factor can be determined based on the following process (1)-(7):

[0078] (1) extracting the current model file text semantic sub-component feature vector at a predetermined position (k is the optimal pairing position value) from the set of the current model file text semantic sub-component feature vectors to obtain a predetermined current model file text semantic sub-component feature vector;

[0079] (2) Extracting a standard model file text semantic sub-component feature vector at a predetermined position from a set of standard model file text semantic sub-component feature vectors to obtain a predetermined standard model file text semantic sub-component feature vector.

[0080] (3) calculating the square of the one-norm of the difference vector between the predetermined semantic subcomponent feature vector of the text of the currently used model file and the predetermined semantic subcomponent feature vector of the text of the standard model file to obtain the one-norm representation of the semantic difference between the semantic subcomponent of the currently used model file and the standard model file;

[0081] (4) Calculate the positional difference between the constant 1 and the square of the one-norm of the feature vector of the text semantic sub-component of the current model file to obtain the one-norm representation of the text semantic sub-component of the current model file.

[0082] (5) Calculate the positional difference between the constant 1 and the square of the one-norm of the standard model document text semantic sub-component feature vector to obtain the one-norm representation of the standard model document text semantic sub-component.

[0083] (6) After calculating the multiplication between the norm representation of the text semantic subcomponent of the current model file and the norm representation of the text semantic subcomponent of the standard model file to obtain the multiplication representation, the norm representation of the semantic difference between the semantic subcomponent of the current model file and the standard model file is divided by the multiplication representation to obtain the semantic implicit association representation of the semantic subcomponent of the current model file and the standard model file.

[0084] (7) The semantic-norm semantic implicit association representation of the current model file-standard model file is multiplied by a constant 2 and then added to a constant 1, and the obtained feature representation is passed through an inverse hyperbolic cosine function to obtain a hyperbolic space distance measurement factor.

[0085] 2. Based on the hyperbolic space distance measurement factor, the set of text semantic sub-component feature vectors of the currently used model file and the set of text semantic sub-component feature vectors of the standard model file are screened for the best feature sub-component pairing to obtain the optimal {currently used model file text semantic sub-component feature vector, standard model file text semantic sub-component feature vector} pairing set;

[0086] Specifically, the position of the standard model file text semantic sub-component feature vector corresponding to the minimum value of multiple hyperbolic space distance measurement factors can be selected as the optimal pairing position value of the currently used model file text semantic sub-component feature vector;

[0087] Specifically, the {currently used model file text semantic sub-component feature vector, standard model file text semantic sub-component feature vector} corresponding to the optimal pairing position value can be paired to obtain the optimal {currently used model file text semantic sub-component feature vector, standard model file text semantic sub-component feature vector} pairing.

[0088] 3. Pair each optimal {currently used model file text semantic sub-component feature vector, standard model file text semantic sub-component feature vector} in the optimal {currently used model file text semantic sub-component feature vector, standard model file text semantic sub-component feature vector} pair and input it into the feature multi-dimensional interaction module to obtain a set of currently used model file-standard model file semantic optimal sub-component feature pairing interaction fusion feature vectors;

[0089] Specifically, the current model file text semantic subcomponent feature vector and the standard model file text semantic subcomponent feature vector in the optimal {current model file text semantic subcomponent feature vector, standard model file text semantic subcomponent feature vector} pairing can be respectively subjected to positional difference, positional dot multiplication and positional summation to obtain the current model file-standard model file semantic subcomponent differential feature vector, the current model file-standard model file semantic subcomponent dot product feature vector and the current model file-standard model file semantic subcomponent sum feature vector.

[0090] Specifically, the position-weighted sum of the semantic sub-component difference feature vector of the current model file - standard model file, the semantic sub-component dot product feature vector of the current model file - standard model file and the semantic sub-component sum feature vector of the current model file - standard model file can be calculated to obtain the interactive fusion feature vector of the optimal semantic sub-component feature pairing of the current model file - standard model file.

[0091] 4. Concatenate the set of interactive fusion feature vectors of the semantic optimal sub-component feature pairing of the current model file and the standard model file to obtain the semantic common component encoding representation vector of the current model file and the standard model file. The interactive process includes but is not limited to addition, multiplication or other forms of combination operations.

[0092] The technical solution of the present invention can realize more intelligent automatic calibration of the substation secondary system model, thereby avoiding the low efficiency and low accuracy problems brought about by the traditional model calibration method, ensuring that potential faults can be discovered and handled in a timely manner, thereby improving the stability of the substation secondary system.

[0093] Embodiment 3

[0094] Figure 3: is a structural block diagram of a system exception handling device provided in Embodiment 3 of the present invention; this embodiment can be applied to the case where anomaly detection is performed on a model file involved in a secondary system of a substation; the system exception handling device provided in the embodiment of the present invention can execute the system exception handling method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method; the system exception handling device can be implemented in the form of hardware and / or software, and configured in an electronic device with a system exception handling function, such as a secondary system of a substation, such as Figure 3 As shown, the system exception handling device specifically includes:

[0095] The file determination module 301 is used to determine the current model file and the current program version corresponding to the target device in the secondary system of the substation in response to the abnormality detection request for the secondary system of the substation;

[0096] A vector determination module 302 is used to determine a standard model file from a preset standard database according to the current program version, and to determine a standard semantic vector and a current semantic vector corresponding to the standard model file and the current model file respectively based on a preset text parsing model;

[0097] The processing module 303 is used to perform feature fine-grained matching interaction on the standard semantic vector and the current semantic vector to obtain a semantic commonality vector, and determine a difference semantic vector and a corresponding abnormality level according to the semantic commonality vector and the current semantic vector to perform abnormality processing.

[0098] The technical solution of the embodiment of the present invention, in response to an abnormality detection request for a substation secondary system, determines the current model file and the current program version corresponding to the target device in the substation secondary system; determines the standard model file from a preset standard database according to the current program version, and determines the standard semantic vector and the current semantic vector corresponding to the standard model file and the current model file respectively based on a preset text parsing model; performs feature fine-grained matching interaction on the standard semantic vector and the current semantic vector to obtain a semantic commonality vector, and determines the difference semantic vector and the corresponding abnormality level according to the semantic commonality vector and the current semantic vector to perform abnormality processing. By using the text parsing model to parse and compare the files of the substation system, the system can be effectively detected for abnormalities, which helps to perform targeted abnormality processing.

[0099] Furthermore, the processing module 303 may include:

[0100] A collection determination unit, used for performing feature fine-grained decoupling of the standard semantic vector and the current semantic vector, so as to obtain a standard sub-feature collection corresponding to the standard semantic vector and a current sub-feature collection corresponding to the current semantic vector respectively;

[0101] A feature pair determination unit, used to determine at least two groups of candidate feature pairs according to the standard sub-feature set and the current sub-feature set, and determine a hyperbolic space distance measurement factor of each group of candidate feature pairs, so as to determine a target feature pair from the candidate feature pairs;

[0102] The vector determination unit is used to determine the interactive fusion vector of the target feature pair and perform a cascade operation on the interactive fusion vector to obtain a semantic commonality vector.

[0103] Furthermore, the vector determination unit is specifically used for:

[0104] According to the preset subtraction weight, the preset addition weight and the preset multiplication weight, subtraction operation, addition operation and multiplication operation are performed on the target feature pair respectively to obtain the sum matrix, deviation matrix and product matrix of the target feature pair;

[0105] The vector matrix obtained by adding the sum matrix, the deviation matrix and the product matrix is ​​determined as the interaction fusion vector of the target feature pair.

[0106] Furthermore, the processing module 303 is specifically used for:

[0107] Determine the word similarity between each standard word vector in the standard semantic vector and each current word vector in the current semantic vector, and determine the sentence similarity between each standard sentence vector in the standard semantic vector and each current sentence vector in the current semantic vector;

[0108] According to the word similarity, sentence similarity and the preset similarity threshold, the target word vector pair and the target sentence vector are determined, and the target word vector pair and the target sentence vector are aggregated to obtain the semantic commonality vector.

[0109] Furthermore, the vector determination module 302 is specifically used for:

[0110] Based on the pre-trained word embedding model and sentence embedding model, determine the standard word vector and standard sentence vector corresponding to the standard model file, and determine the current word vector and current sentence vector corresponding to the current model file;

[0111] Based on the preset semantic model, the standard word vector and the standard sentence vector are semantically encoded to obtain the standard semantic vector, and the current word vector and the current sentence vector are semantically encoded to obtain the current semantic vector.

[0112] Furthermore, the processing module 303 is also used for:

[0113] According to the semantic commonality vector and the current semantic vector, a difference semantic vector is determined based on a difference component extractor of a preset difference network;

[0114] Based on the preset difference grading model, the exception level corresponding to the difference semantic vector is determined for exception handling.

[0115] Embodiment 4

[0116] Figure 4 It is a schematic diagram of the structure of an electronic device provided in Embodiment 4 of the present invention. Figure 4 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0117] like Figure 4 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0118] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0119] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a system exception handling method.

[0120] In some embodiments, the system exception handling method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the system exception handling method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to execute the system exception handling method in any other appropriate manner (e.g., by means of firmware).

[0121] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0122] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0123] In the context of the present invention, a computer readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device, or equipment. A computer readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. Alternatively, a computer readable storage medium may be a machine readable signal medium. A more specific example of a machine readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0124] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0125] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0126] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0127] In one embodiment, the embodiment of the present invention further includes a computer program product, the computer program product includes a computer program, and the computer program implements the system exception handling method of any embodiment of the present invention when executed by a processor.

[0128] The computer program product may be implemented in a computer program code for performing the operation of the present invention written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​and conventional procedural programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0129] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0130] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A system exception handling method, characterized in that: include: In response to a request for abnormality detection of a substation secondary system, determining a current model file and a current program version corresponding to a target device in the substation secondary system; According to the current program version, determine the standard model file from the preset standard database, and determine the standard semantic vector and the current semantic vector corresponding to the standard model file and the current model file respectively based on the preset text parsing model; The standard semantic vector and the current semantic vector are interactively matched with each other in fine-grained features to obtain a semantic commonality vector, and the difference semantic vector and the corresponding exception level are determined based on the semantic commonality vector and the current semantic vector for exception handling.

2. The method according to claim 1, characterized in that: Perform fine-grained feature matching interaction on the standard semantic vector and the current semantic vector to obtain a semantic commonality vector, including: Perform feature fine-grained decoupling of the standard semantic vector and the current semantic vector to obtain a standard sub-feature collection corresponding to the standard semantic vector and a current sub-feature collection corresponding to the current semantic vector respectively; Determine at least two groups of candidate feature pairs according to the standard sub-feature set and the current sub-feature set, and determine the hyperbolic space distance measurement factor of each group of candidate feature pairs to determine the target feature pair from the candidate feature pairs; Determine the interactive fusion vector of the target feature pair, and perform a cascade operation on the interactive fusion vector to obtain a semantic commonality vector.

3. The method according to claim 2, characterized in that Determine the interactive fusion vector of the target feature pair, including: According to the preset subtraction weight, the preset addition weight and the preset multiplication weight, subtraction operation, addition operation and multiplication operation are performed on the target feature pair respectively to obtain the sum matrix, deviation matrix and product matrix of the target feature pair; The vector matrix obtained by adding the sum matrix, the deviation matrix and the product matrix is ​​determined as the interaction fusion vector of the target feature pair.

4. The method according to claim 1, characterized in that Perform fine-grained feature matching interaction on the standard semantic vector and the current semantic vector to obtain a semantic commonality vector, including: Determine the word similarity between each standard word vector in the standard semantic vector and each current word vector in the current semantic vector, and determine the sentence similarity between each standard sentence vector in the standard semantic vector and each current sentence vector in the current semantic vector; According to the word similarity, sentence similarity and the preset similarity threshold, the target word vector pair and the target sentence vector are determined, and the target word vector pair and the target sentence vector are aggregated to obtain the semantic commonality vector.

5. The method according to claim 1, characterized in that Based on the preset text parsing model, the standard semantic vector and the current semantic vector corresponding to the standard model file and the current model file are determined respectively, including: Based on the pre-trained word embedding model and sentence embedding model, determine the standard word vector and standard sentence vector corresponding to the standard model file, and determine the current word vector and current sentence vector corresponding to the current model file; Based on the preset semantic model, the standard word vector and the standard sentence vector are semantically encoded to obtain the standard semantic vector, and the current word vector and the current sentence vector are semantically encoded to obtain the current semantic vector.

6. The method according to claim 1, characterized in that According to the semantic commonality vector and the current semantic vector, the difference semantic vector and the corresponding abnormal level are determined, including: According to the semantic commonality vector and the current semantic vector, a difference semantic vector is determined based on a difference component extractor of a preset difference network; Based on the preset difference grading model, the exception level corresponding to the difference semantic vector is determined for exception handling.

7. A system abnormality handling device, characterized in that: include: A file determination module, for determining a current model file and a current program version corresponding to a target device in the secondary system of the substation in response to an abnormality detection request for the secondary system of the substation; A vector determination module, used to determine a standard model file from a preset standard database according to the current program version, and to determine a standard semantic vector and a current semantic vector corresponding to the standard model file and the current model file respectively based on a preset text parsing model; The processing module is used to perform feature fine-grained matching interaction on the standard semantic vector and the current semantic vector to obtain a semantic commonality vector, and determine the difference semantic vector and the corresponding exception level based on the semantic commonality vector and the current semantic vector for exception processing.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the system exception handling method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the system exception handling method according to any one of claims 1 to 6 when executed.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the computer program implements the system exception handling method according to any one of claims 1 to 6.

Images