A digital security chip processing system and method

Abstract

The present application discloses a digital security chip processing system and method, which uses deep learning-based artificial intelligence technology to perform semantic analysis on the security protection demand information of the digital power grid system, mines the contextual semantic features of the security protection demand information, and extracts a set of type labels of security chips from the database, and performs cross-domain query matching between the security protection demand features and the security chip type feature set to achieve intelligent recommendation of security chips, and then performs node deployment of the digital power grid system and encryption processing of target data based on the selected security chip. In this way, the efficiency and accuracy of security chip selection can be effectively improved, the complexity of manual configuration can be reduced, and intelligent matching between security protection requirements and security chip types can be achieved.

Description

Technical Field

[0001] The present application relates to the field of intelligent encryption, and more specifically, to a digital security chip processing system and method. Background Art

[0002] With the rapid development of information technology, especially the continuous progress in emerging technology fields such as the Internet of Things (IoT), big data and cloud computing, the digital power grid system, as a core component of the modern energy Internet, faces unprecedented information security challenges. In order to ensure the reliable operation and information security of the power grid system, various security technologies and solutions have emerged. Among them, data encryption technology, as one of the important means to ensure data security, has been widely used in various fields. However, traditional data software encryption methods usually rely on key management systems, but due to the complex key generation, distribution, storage and update processes, the risk of key leakage or theft is high.

[0003] As a special integrated circuit that integrates multiple security functions, digital security chips have higher security and reliability than traditional data software encryption methods. They can provide high-intensity data encryption, identity authentication, secure storage and other services, and are an important means to improve the security of digital power grid systems. However, in the application process of digital security chips, engineers are usually required to rely on their manual experience to select security chips suitable for specific application scenarios. This process is not only inefficient, but also difficult to guarantee the accuracy and optimality of the results, especially when faced with a large and constantly updated chip library. Therefore, a digital security chip processing method is expected to achieve intelligent selection of digital security chips. Summary of the invention

[0004] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a system and method based on digital security chip processing, which uses artificial intelligence technology based on deep learning to perform semantic analysis on the security protection demand information of the digital power grid system, mines the contextual semantic features of the security protection demand information, and extracts a set of type labels of the security chip from the database, and performs cross-domain query matching between the security protection demand features and the security chip type feature set to achieve intelligent recommendation of security chips, and then performs node deployment of the digital power grid system and encryption processing of target data based on the selected security chip. In this way, the efficiency and accuracy of security chip selection can be effectively improved, the complexity of manual configuration can be reduced, and intelligent matching between security protection requirements and security chip types can be achieved.

[0005] According to one aspect of the present application, a digital security chip processing method is provided, which includes:

[0006] Obtain security protection requirement information of the digital power grid system and extract a set of type labels of security chips from the database;

[0007] Performing one-hot encoding on the type label of each security chip in the set of type labels of the security chip to obtain a set of one-hot encoding vectors of the security chip type label;

[0008] Semantically encoding the security protection requirement information to obtain a semantically encoded feature vector of the security protection requirement information;

[0009] Performing cross-domain query matching analysis on a set of the semantic encoding feature vector of the security protection requirement information and the one-hot encoding vector of the security chip type label to obtain a requirement-chip type cross-domain query matching feature vector;

[0010] Generate a recommendation result of a security chip based on the requirement-chip type cross-domain query matching feature vector;

[0011] Integrating the selected security chip into key nodes of the digital grid system;

[0012] Utilizing the encryption algorithm of the security chip to encrypt the target data;

[0013] Among them, performing cross-domain query matching analysis on the set of the semantic encoding feature vector of the security protection requirement information and the one-hot encoding vector of the security chip type label to obtain the requirement-chip type cross-domain query matching feature vector includes:

[0014] The semantically encoded feature vector of the security protection requirement information is used as a query feature vector, and a feature matrix obtained by arranging a set of the one-hot encoding vectors of the security chip type label is used as an initial key matrix. Based on the feature distribution difference between the query feature vector and each row vector in the initial key matrix, the initial key matrix is ​​locally associated with the enhancement to obtain an updated key matrix;

[0015] A cross-domain attention query based on a self-attention mechanism is performed on the query feature vector and the update key matrix to obtain the requirement-chip type cross-domain query matching feature vector.

[0016] According to another aspect of the present application, a digital security chip processing system is provided, comprising:

[0017] An information acquisition module, used to obtain the security protection requirements of the digital power grid system and extract a set of type labels of security chips from the database;

[0018] A security chip type label one-hot encoding module, used for performing one-hot encoding on the type label of each security chip in the set of security chip type labels to obtain a set of security chip type label one-hot encoding vectors;

[0019] A security protection requirement information semantic encoding module, used for semantically encoding the security protection requirement information to obtain a security protection requirement information semantic encoding feature vector;

[0020] A cross-domain query matching analysis module, used to perform a cross-domain query matching analysis on a set of the semantic encoding feature vector of the security protection requirement information and the one-hot encoding vector of the security chip type label to obtain a requirement-chip type cross-domain query matching feature vector;

[0021] A recommendation result generation module, used to generate a recommendation result of a security chip based on the requirement-chip type cross-domain query matching feature vector;

[0022] A chip integration module, used to integrate the selected security chip into a key node of a digital power grid system;

[0023] The encryption module is used to encrypt the target data using the encryption algorithm of the security chip.

[0024] Compared with the prior art, the digital security chip processing system and method provided by the present application uses deep learning-based artificial intelligence technology to perform semantic analysis on the security protection demand information of the digital power grid system, mines the contextual semantic features of the security protection demand information, and extracts a set of type labels of security chips from the database. By performing cross-domain query matching between the security protection demand features and the security chip type feature set, intelligent recommendation of security chips is achieved, and then node deployment of the digital power grid system and encryption processing of target data are performed based on the selected security chip. In this way, the efficiency and accuracy of security chip selection can be effectively improved, the complexity of manual configuration can be reduced, and intelligent matching between security protection requirements and security chip types can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0026] Figure 1 is a flowchart of a digital security chip processing method according to an embodiment of the present application;

[0027] Figure 2 A schematic diagram of data flow based on a digital security chip processing method according to an embodiment of the present application;

[0028] Figure 3 is a flowchart of sub-step S4 of the digital security chip processing method according to an embodiment of the present application;

[0029] Figure 4 is a block diagram of a digital security chip processing system according to an embodiment of the present application. DETAILED DESCRIPTION

[0030] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0031] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0032] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0033] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0034] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0035] Traditional data software encryption methods usually rely on key management systems, but due to the complex key generation, distribution, storage and update processes, there is a high risk of key leakage or theft. As a special integrated circuit that integrates multiple security functions, digital security chips have higher security and reliability than traditional data software encryption methods. They can provide high-intensity data encryption, identity authentication, secure storage and other services, and are an important means to improve the security of digital power grid systems. However, in the application process of digital security chips, engineers are usually required to rely on their manual experience to select security chips suitable for specific application scenarios. This process is not only inefficient, but also difficult to guarantee the accuracy and optimality of the results, especially when faced with a large and constantly updated chip library. Therefore, a digital security chip processing method is expected to achieve intelligent selection of digital security chips.

[0036] In the technical solution of the present application, a processing method based on a digital security chip is proposed. Figure 1 Flow chart of a digital security chip processing method according to an embodiment of the present application. Figure 2 Schematic diagram of data flow based on the digital security chip processing method according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to the embodiment of the present application, the digital security chip processing method includes the following steps: S1, obtaining security protection requirement information of the digital power grid system, and extracting a set of type labels of the security chip from the database; S2, performing one-hot encoding on the type labels of each security chip in the set of type labels of the security chip to obtain a set of one-hot encoded vectors of the security chip type labels; S3, performing semantic encoding on the security protection requirement information to obtain a semantic encoding feature vector of the security protection requirement information; S4, performing cross-domain query matching analysis on the set of the semantic encoding feature vector of the security protection requirement information and the one-hot encoded vector of the security chip type label to obtain a demand-chip type cross-domain query matching feature vector; S5, generating a recommendation result for the security chip based on the demand-chip type cross-domain query matching feature vector; S6, integrating the selected security chip into the key node of the digital power grid system; S7, encrypting the target data using the encryption algorithm of the security chip.

[0037] In particular, the S1 obtains the security protection requirement information of the digital power grid system and extracts a set of type labels of security chips from the database. It should be understood that as a complex and critical infrastructure, the digital power grid system has diverse security protection requirements. By obtaining the security protection requirement information of the digital power grid system, its security requirements in different scenarios can be clarified. At the same time, since different security chips usually have different functional characteristics, performance parameters and applicable scenarios, by extracting a set of type labels of security chips from the database, a rich selection basis can be provided for the security protection of the power grid system.

[0038] In particular, S2 performs one-hot encoding on the type labels of each security chip in the set of type labels of the security chip to obtain a set of one-hot encoding vectors of the security chip type labels. In order to realize the intelligent selection of security chips using deep learning technology, first of all, it is necessary to semantically understand the security protection requirement information and the type labels of each security chip to convert non-numerical data into a form that can be understood and processed by the machine learning model. Considering that the type label of the security chip is discrete classification information, therefore, in the technical solution of the present application, the one-hot encoding technology is used to process the type labels of each security chip in the set of type labels of the security chip, and each type label is converted into a unique and mutually independent binary numerical vector, while retaining the distinction and integrity between the original labels, thereby obtaining a set of one-hot encoding vectors of the security chip type labels. It is worth mentioning that one-hot encoding is an encoding method that converts categorical variables into numerical form, which is commonly used in machine learning and data processing. Its main purpose is to represent each category as a binary vector so that the model can effectively process these category information.

[0039] In particular, the S3 performs semantic encoding on the security protection requirement information to obtain a semantic encoding feature vector of the security protection requirement information. In a specific embodiment of the present application, the security protection requirement information is input into a semantic encoder based on the BERT model for contextual semantic encoding to obtain a semantic encoding feature vector of the security protection requirement information. Considering that the security protection requirement information usually contains unstructured data such as text descriptions and parameter requirements, it contains rich semantic information. Therefore, in order to fully explore its contextual semantic features, the present application uses a semantic encoder based on the BERT model to semantically encode the security protection requirement information, so as to utilize the bidirectional semantic encoding capability of the BERT model to fully capture the semantic meaning and contextual semantic association of each word and phrase in the text, and to mine the deep semantic features of the security protection requirement information, so as to generate a semantic encoding feature vector of the security protection requirement information, and keep the same data form as the security chip type label, so as to provide a data basis for subsequent cross-domain query matching analysis. It is worth mentioning that BERT is a pre-trained language model based on the Transformer architecture, which has the characteristics of bidirectional semantic encoding. The core feature of BERT is its bidirectional context understanding capability. Traditional language models are usually unidirectional (from left to right or from right to left), while BERT encodes by considering all words in the context at the same time. This means that when BERT generates the representation of a word, it considers all the words before and after the word at the same time, thereby capturing richer semantic information.

[0040] In particular, the S4 performs a cross-domain query matching analysis on the set of the semantically encoded feature vectors of the security protection requirement information and the one-hot encoded vectors of the security chip type labels to obtain a requirement-chip type cross-domain query matching feature vector. In order to improve the matching accuracy between the security chip type and the security protection requirement, the present application proposes a cross-domain feature query matching method, which is based on the Transformer architecture and gating mechanism, and performs a matching analysis on the set of the semantically encoded feature vectors of the security protection requirement information and the one-hot encoded vectors of the security chip type labels to extract the correlation characteristics between the set of security chip type labels and the security protection requirement. In a specific example of the present application, Figure 3 As shown, the S4 includes: S41, taking the semantically encoded feature vector of the security protection requirement information as the query feature vector, and taking the feature matrix obtained by arranging the set of the unique-hot encoding vectors of the security chip type label as the initial key matrix, based on the feature distribution difference between the query feature vector and each row vector in the initial key matrix, locally strengthening the initial key matrix to obtain an updated key matrix; S42, performing a cross-domain attention query based on the self-attention mechanism on the query feature vector and the updated key matrix to obtain the requirement-chip type cross-domain query matching feature vector.

[0041] Specifically, the S41 uses the semantically encoded feature vector of the security protection requirement information as the query feature vector, and uses the feature matrix obtained by arranging the set of unique-hot encoding vectors of the security chip type label as the initial key matrix, and based on the feature distribution differences between the query feature vector and each row vector in the initial key matrix, performs local association enhancement on the initial key matrix to obtain an updated key matrix. Specifically, first, calculate the feature distribution difference energy coefficient between the query feature vector and each row vector in the initial key matrix to obtain a set of feature distribution difference energy coefficients. By calculating the feature distribution difference energy coefficient between the query feature vector and each row vector in the initial key matrix, the local association strength contained in each row vector is revealed. Secondly, based on the set of feature distribution difference energy coefficients, the initial key matrix is ​​matrix updated to obtain the updated key matrix. That is, based on the gating mechanism, for the row vector corresponding to the feature distribution difference energy coefficient less than a predetermined threshold, its mean vector with the query feature vector is used to update its feature representation, thereby further strengthening the security chip type label feature representation that is more relevant to security protection requirements, so as to improve the pertinence and matching quality of the key matrix and generate an updated key matrix.

[0042] Among them, calculating the feature distribution difference energy coefficient between the query feature vector and each row vector in the initial key matrix to obtain a set of feature distribution difference energy coefficients includes: extracting the semantic difference features between the query feature vector and each row vector in the initial key matrix respectively to obtain a set of semantic association difference feature vectors; calculating the energy coefficient of each semantic association difference feature vector in the set of semantic association difference feature vectors to obtain the set of feature distribution difference energy coefficients. Specifically, the process of extracting the semantic difference features between the query feature vector and each row vector in the initial key matrix respectively to obtain a set of semantic association difference feature vectors includes: calculating the position point addition, position point subtraction and position point multiplication between the query feature vector and the row vector to obtain a first matching fusion result, a second matching fusion result and a third matching fusion result; after cascading the first matching fusion result, the second matching fusion result and the third matching fusion result into a multi-scale semantic association matching vector, performing one-dimensional convolution processing and maximum pooling processing on them to obtain the semantic association difference feature vector. Specifically, the process of calculating the energy coefficient of each semantic association difference feature vector in the set of semantic association difference feature vectors to obtain the set of feature distribution difference energy coefficients includes: weighted summing the square of the difference between the maximum eigenvalue of the semantic association difference feature vector and its eigenmean, the feature variance of the semantic association difference feature vector, and the bias term to obtain a weighted sum value, and dividing the sum of the feature variance of the semantic association difference feature vector and the bias term by the weighted sum value to obtain the feature distribution difference energy coefficient.

[0043] More specifically, based on the set of feature distribution difference energy coefficients, the initial key matrix is ​​updated to obtain an updated key matrix, including: comparing each feature distribution difference energy coefficient in the set of feature distribution difference energy coefficients with a predetermined threshold, and in response to the feature distribution difference energy coefficient being less than the predetermined threshold, using the mean vector between the row vector corresponding to the feature distribution difference energy coefficient and the query feature vector as the updated row vector to obtain the updated key matrix.

[0044] Specifically, the S42 performs a cross-domain attention query based on the self-attention mechanism on the query feature vector and the updated key matrix to obtain the demand-chip type cross-domain query matching feature vector. In the technical solution of the present application, a query vector and a value vector are constructed based on the query feature vector, and each row vector in the updated key matrix is ​​used as a key vector. Through the self-attention calculation of the Transformer structure, a cross-domain information query matching is performed to obtain a cross-domain query attention vector between the security protection demand feature and each security chip type label feature. Finally, by calculating the mean of each cross-domain query attention vector by position, all security chip type label information is integrated to generate a comprehensive demand-chip type cross-domain query matching feature vector. In this way, the demand-chip type cross-domain query matching feature vector contains the association matching strength information between each security chip type label and the security protection demand. Specifically, first, the query feature vector is processed using a value embedding matrix to obtain a value vector; then, the query feature vector is used as the query vector and each row vector in the update key matrix is ​​used as a key vector, and the query vector, the key vector and the value vector are input into a cross-domain attention query module based on a Transformer structure to obtain a sequence of demand-chip type cross-domain query attention vectors; then, the positional mean vector of the sequence of demand-chip type cross-domain query attention vectors is calculated to obtain the demand-chip type cross-domain query matching feature vector.

[0045] In summary, in the above embodiment, a cross-domain query matching analysis is performed on the set of the semantic encoding feature vector of the security protection requirement information and the unique hot encoding vector of the security chip type label to obtain a requirement-chip type cross-domain query matching feature vector, including: processing the query feature vector and the key matrix with the following cross-domain query matching formula to obtain the requirement-chip type cross-domain query matching feature vector, wherein the cross-domain query matching formula is:

[0046]

[0047]

[0048]

[0049]

[0050]

[0051]

[0052]

[0053] in, represents the initial bond matrix, , , and are the first, second, and and row vectors, i.e., the unique hot encoding vectors of each security chip type label, The value of is the number of vectors in the set of one-hot encoding vectors of the security chip type label, represents dot product, Indicates point addition, Indicates point reduction, Indicates cascade operation, represents the maximum pooling operation, represents the maximum pooling function, represents a one-dimensional convolution operation, Indicates semantic association difference feature vectors, , and Respectively represent the The feature variance, maximum eigenvalue and feature mean of the semantic association difference feature vector, and are different bias terms, is the first in the set of characteristic distribution difference energy coefficients The characteristic distribution difference energy coefficient, is the predetermined threshold, represents the updated key matrix, represents the first row vector, represents the query feature vector, is the value embedding matrix, is a value vector, is the normalized exponential function, The value of is the length of the row vector in the update key matrix, represents the transpose of a vector, It is Demand-Chip type cross-domain query attention vector, The feature vector representing the demand-chip type cross-domain query matching.

[0054] It is worth mentioning that in other examples of the present application, the following steps can be used to perform cross-domain query matching analysis on the set of semantically encoded feature vectors of the security protection requirement information and the one-hot encoded vectors of the security chip type label to obtain a cross-domain query matching feature vector of the requirement-chip type, for example: first, collect security protection requirement information and extract relevant features; and collect security chip types and extract corresponding labels; then, semantically encode the security protection requirement information to generate a feature vector, and one-hot encode the security chip type to generate a label vector; then, use natural language processing (NLP) technology (such as Word2Vec, BERT, etc.) to semantically encode the security protection requirement information to obtain a feature vector; convert the security chip type label into a one-hot encoding to obtain a binary vector representation; then, select a suitable matching algorithm, such as cosine similarity, Euclidean distance, etc., to calculate the similarity between feature vectors; calculate the similarity between each security protection requirement feature vector and the one-hot encoded vectors of all security chip types to obtain a matching score; finally, generate a cross-domain query matching feature vector of the requirement-chip type based on the similarity score.

[0055] In particular, the S5 generates a recommendation result of a security chip based on the requirement-chip type cross-domain query matching feature vector. In a specific example of the present application, the requirement-chip type cross-domain query matching feature vector is input into a classifier-based security chip recommender to obtain the recommendation result, and the recommendation result is used to represent the type label of the recommended security chip. Through classification processing, the recommendation result of the security chip type is obtained, thereby realizing intelligent matching between the security protection requirements of the digital power grid system and the security chip type.

[0056] In a preferred example, inputting the requirement-chip type cross-domain query matching feature vector into a classifier-based security chip recommender to obtain a recommendation result includes:

[0057] Determine a maximum eigenvalue of the demand-chip type cross-domain query matching eigenvalue and a minimum eigenvalue of the demand-chip type cross-domain query matching eigenvalue vector;

[0058] Calculate the demand-chip type cross-domain query matching mean and demand-chip type cross-domain query matching standard deviation of the feature set of the demand-chip type cross-domain query matching feature vector, and calculate the quotient of the demand-chip type cross-domain query matching mean and the demand-chip type cross-domain query matching standard deviation to obtain a demand-chip type cross-domain query matching statistical normalization value;

[0059] Calculate the inverse of each eigenvalue of the demand-chip type cross-domain query matching eigenvalue vector, multiply it by the difference between the maximum eigenvalue of the demand-chip type cross-domain query matching and the minimum eigenvalue of the demand-chip type cross-domain query matching, and then subtract it from the demand-chip type cross-domain query matching statistical normalization value to obtain a demand-chip type cross-domain query matching distribution approximation vector;

[0060] Calculate an exponential function with a natural constant as the base and each eigenvalue of the demand-chip type cross-domain query matching distribution approximation vector as the exponent to obtain a demand-chip type cross-domain query matching distribution class approximation vector;

[0061] Performing dot-addition on the demand-chip type cross-domain query matching distribution class approximation vector and the demand-chip type cross-domain query matching statistical normalized value, and calculating the base-two logarithm of the absolute value of each eigenvalue of the dot-added vector to obtain an optimized demand-chip type cross-domain query matching eigenvector; and

[0062] The optimized requirement-chip type cross-domain query matching feature vector is input into a classifier-based security chip recommender to obtain a recommendation result.

[0063] Among them, the demand-chip type cross-domain query matching feature vector is recorded as The optimization expression is:

[0064]

[0065]

[0066]

[0067]

[0068] in, is the demand-chip type cross-domain query matching feature vector, and are respectively the maximum eigenvalue of the demand-chip type cross-domain query matching eigenvalue and the minimum eigenvalue of the demand-chip type cross-domain query matching eigenvalue vector, and is the difference between the maximum eigenvalue of the demand-chip type cross-domain query matching and the minimum eigenvalue of the demand-chip type cross-domain query matching, is the demand-chip type cross-domain query matching average, is the demand-chip type cross-domain query matching standard deviation, Statistical normalization value for cross-domain query matching of demand-chip type, is the demand-chip type cross-domain query matching distribution class approximation vector, is the logarithm to base 2, () is the exponential operation, It's a little bit more. is the dot product, It's a point reduction. is the optimized demand-chip type cross-domain query matching feature vector.

[0069] Here, in the preferred example, since the set of the semantic encoding feature vector of the security protection requirement information and the set of the unique-hot encoding vector of the security chip type label respectively represent the encoded semantic features of the set of security protection requirement information and the type label of the security chip, when they are input into the cross-domain feature gated query network with enhanced local association of the key matrix, the local association difference of the key matrix attributed to the difference in the distribution of the encoded semantic features will also cause the driving query difference across the text semantic space, so that the requirement-chip type cross-domain query matching feature vector will also have a diversified set expression distribution of the cross-domain query aggregation feature. Therefore, it is expected to improve the balance between the regression mapping accuracy and completeness when the requirement-chip type cross-domain query matching feature vector is input into the classifier-based security chip recommender for classification regression, thereby improving the semantic accuracy of the recommendation results.

[0070] Therefore, by performing random statistical normalization on the diverse feature set of the demand-chip type cross-domain query matching feature vector, a standardized continuous probability density distribution is approximated for the response hypothesis test of the confidence space constructed based on the overall feature value of the demand-chip type cross-domain query matching feature vector relative to each feature value of the demand-chip type cross-domain query matching feature vector, thereby establishing the target reachability of the diverse feature distribution of the demand-chip type cross-domain query matching feature vector to a unified regression target, so as to achieve a balanced executable performance between mapping accuracy and mapping completeness in the classification regression process based on the diverse feature distribution of the demand-chip type cross-domain query matching feature vector, and improve the semantic accuracy of the recommendation results obtained by the demand-chip type cross-domain query matching feature input to the classifier-based security chip recommender.

[0071] In particular, S6 and S7 integrate the selected security chip into the key nodes of the digital grid system; and use the encryption algorithm of the security chip to encrypt the target data. Here, by integrating the selected security chip into the key nodes of the digital grid system, the target data is encrypted using the encryption algorithm of the security chip to achieve the security protection function of the digital grid system and ensure the security of data transmission and storage. In the digital grid system, key nodes generally include key facilities such as substations, power plants, and dispatching centers. By deploying the recommended security chips on these key nodes, it is possible to effectively prevent data from being stolen or tampered with during transmission, improve the overall security performance of the system, and thus ensure the stable operation of the power system and the protection of user privacy.

[0072] In summary, the digital security chip processing method according to the embodiment of the present application is explained, which uses artificial intelligence technology based on deep learning to perform semantic analysis on the security protection demand information of the digital power grid system, mines the contextual semantic features of the security protection demand information, and extracts a set of type labels of the security chip from the database, and performs cross-domain query matching between the security protection demand features and the security chip type feature set to achieve intelligent recommendation of security chips, and then performs node deployment of the digital power grid system and encryption processing of target data based on the selected security chip. In this way, the efficiency and accuracy of security chip selection can be effectively improved, the complexity of manual configuration can be reduced, and intelligent matching between security protection requirements and security chip types can be achieved.

[0073] Furthermore, a digital security chip processing system is also provided.

[0074] Figure 4 FIG. 1 is a block diagram of a digital security chip processing system according to an embodiment of the present application. Figure 4As shown, according to the embodiment of the present application, the digital security chip processing system 300 includes: an information acquisition module 310, which is used to obtain the security protection requirements of the digital power grid system and extract a set of type labels of the security chip from the database; a security chip type label unique hot encoding module 320, which is used to uniquely encode the type labels of each security chip in the set of type labels of the security chip to obtain a set of unique hot encoding vectors of the security chip type label; a security protection requirement information semantic encoding module 330, which is used to semantically encode the security protection requirement information to obtain a semantic encoding feature vector of the security protection requirement information; a cross-domain query matching analysis module 340, which is used to perform cross-domain query matching analysis on the set of the semantic encoding feature vector of the security protection requirement information and the unique hot encoding vector of the security chip type label to obtain a demand-chip type cross-domain query matching feature vector; a recommendation result generation module 350, which is used to generate a recommendation result of the security chip based on the demand-chip type cross-domain query matching feature vector; a chip integration module 360, which is used to integrate the selected security chip into the key node of the digital power grid system; an encryption module 370, which is used to encrypt the target data using the encryption algorithm of the security chip.

[0075] As described above, the digital security chip processing system 300 according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with a digital security chip processing algorithm. In a possible implementation, the digital security chip processing system 300 according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the digital security chip processing system 300 can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the digital security chip processing system 300 can also be one of the many hardware modules of the wireless terminal.

[0076] Alternatively, in another example, the digital security chip processing system 300 and the wireless terminal may be separate devices, and the digital security chip processing system 300 may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in an agreed data format.

[0077] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A digital security chip processing method, characterized in that: include: Obtain security protection requirement information of the digital power grid system and extract a set of type labels of security chips from the database; Performing one-hot encoding on the type label of each security chip in the set of type labels of the security chip to obtain a set of one-hot encoding vectors of the security chip type label; Semantically encoding the security protection requirement information to obtain a semantically encoded feature vector of the security protection requirement information; Performing cross-domain query matching analysis on a set of the semantic encoding feature vector of the security protection requirement information and the one-hot encoding vector of the security chip type label to obtain a requirement-chip type cross-domain query matching feature vector; Generate a recommendation result of a security chip based on the requirement-chip type cross-domain query matching feature vector; Integrating the selected security chip into key nodes of the digital grid system; Utilizing the encryption algorithm of the security chip to encrypt the target data; Among them, performing cross-domain query matching analysis on the set of the semantic encoding feature vector of the security protection requirement information and the one-hot encoding vector of the security chip type label to obtain the requirement-chip type cross-domain query matching feature vector includes: The semantically encoded feature vector of the security protection requirement information is used as a query feature vector, and a feature matrix obtained by arranging a set of the one-hot encoding vectors of the security chip type label is used as an initial key matrix. Based on the feature distribution difference between the query feature vector and each row vector in the initial key matrix, the initial key matrix is ​​locally associated with the enhancement to obtain an updated key matrix; Performing a cross-domain attention query based on a self-attention mechanism on the query feature vector and the update key matrix to obtain the requirement-chip type cross-domain query matching feature vector; The semantically encoded feature vector of the security protection requirement information is used as a query feature vector, and a feature matrix obtained by arranging a set of the unique-hot encoding vectors of the security chip type label is used as an initial key matrix. Based on the feature distribution difference between the query feature vector and each row vector in the initial key matrix, the initial key matrix is ​​locally associated with the enhancement to obtain an updated key matrix, including: Calculating feature distribution difference energy coefficients between the query feature vector and each row vector in the initial key matrix to obtain a set of feature distribution difference energy coefficients; Based on the set of characteristic distribution difference energy coefficients, performing matrix update on the initial bond matrix to obtain the updated bond matrix; Calculating the feature distribution difference energy coefficients between the query feature vector and each row vector in the initial key matrix to obtain a set of feature distribution difference energy coefficients, including: Extracting semantic difference features between the query feature vector and each row vector in the initial key matrix respectively to obtain a set of semantic association difference feature vectors; Calculating the energy coefficient of each semantic association difference feature vector in the set of semantic association difference feature vectors to obtain the set of feature distribution difference energy coefficients; Extracting semantic difference features between the query feature vector and each row vector in the initial key matrix respectively to obtain a set of semantic association difference feature vectors, including: Calculate the position point addition, position point subtraction and position point multiplication between the query feature vector and the row vector to obtain a first matching fusion result, a second matching fusion result and a third matching fusion result; After concatenating the first matching fusion result, the second matching fusion result, and the third matching fusion result into a multi-scale semantic association matching vector, one-dimensional convolution processing and maximum pooling processing are performed on the vector to obtain the semantic association difference feature vector; Calculating the energy coefficient of each semantic association difference feature vector in the set of semantic association difference feature vectors to obtain the set of feature distribution difference energy coefficients includes: Performing weighted summation on the square of the difference between the maximum eigenvalue of the semantic association difference feature vector and its feature mean, the feature variance of the semantic association difference feature vector, and a bias term to obtain a weighted sum value, and dividing the sum of the feature variance of the semantic association difference feature vector and the bias term by the weighted sum value to obtain the feature distribution difference energy coefficient; Based on the set of characteristic distribution difference energy coefficients, performing a matrix update on the initial bond matrix to obtain an updated bond matrix includes: Comparing each feature distribution difference energy coefficient in the set of feature distribution difference energy coefficients with a predetermined threshold, and in response to the feature distribution difference energy coefficient being less than the predetermined threshold, taking a mean vector between a row vector corresponding to the feature distribution difference energy coefficient and the query feature vector as an updated row vector to obtain the updated key matrix; Performing a cross-domain attention query based on a self-attention mechanism on the query feature vector and the update key matrix to obtain the requirement-chip type cross-domain query matching feature vector, including: Processing the query feature vector using a value embedding matrix to obtain a value vector; Using the query feature vector as a query vector and each row vector in the update key matrix as a key vector, the query vector, the key vector and the value vector are input into a cross-domain attention query module based on a Transformer structure to obtain a sequence of demand-chip type cross-domain query attention vectors; The position-wise mean vector of the sequence of the demand-chip type cross-domain query focus vector is calculated to obtain the demand-chip type cross-domain query matching feature vector.

2. The digital security chip processing method according to claim 1, characterized in that: Semantically encoding the security protection requirement information to obtain a semantically encoded feature vector of the security protection requirement information includes: The security protection requirement information is input into a semantic encoder based on the BERT model for contextual semantic encoding to obtain a semantic encoding feature vector of the security protection requirement information.

3. The digital security chip processing method according to claim 2, characterized in that: Based on the requirement-chip type cross-domain query matching feature vector, a security chip recommendation result is generated, including: The requirement-chip type cross-domain query matching feature vector is input into a classifier-based security chip recommender to obtain the recommendation result, where the recommendation result is used to represent a type label of the recommended security chip.

4. A digital security chip processing system, used to execute the digital security chip processing method according to claim 1, characterized in that: include: An information acquisition module, used to obtain the security protection requirements of the digital power grid system and extract a set of type labels of security chips from the database; A security chip type label one-hot encoding module, used for performing one-hot encoding on the type label of each security chip in the set of security chip type labels to obtain a set of security chip type label one-hot encoding vectors; A security protection requirement information semantic encoding module, used for semantically encoding the security protection requirement information to obtain a security protection requirement information semantic encoding feature vector; A cross-domain query matching analysis module, used to perform a cross-domain query matching analysis on a set of the semantic encoding feature vector of the security protection requirement information and the one-hot encoding vector of the security chip type label to obtain a requirement-chip type cross-domain query matching feature vector; A recommendation result generation module, used to generate a recommendation result of a security chip based on the requirement-chip type cross-domain query matching feature vector; A chip integration module, used to integrate the selected security chip into a key node of a digital power grid system; The encryption module is used to encrypt the target data using the encryption algorithm of the security chip.

Images