Power grid data security sharing and summarizing method compatible with privacy protection and multi-center data structure model of power grid data security sharing and summarizing method

Abstract

The invention discloses a power grid data security sharing and summarizing method compatible with privacy protection and a multi-center data structure model thereof, and relates to the technical field of power grid data sharing, and the method comprises the steps: integrating the comprehensive power grid data of a power worker into a multi-stage discrete data unit through keyword extraction and data association; multi-dimensional processing is carried out on the basis; a central data storage unit formed through iteration is combined with a parameter space module and access authority management to ensure safe sharing of data; and privacy protection is enhanced through a privacy protection technology. Through efficient extraction and multi-dimensional integration of the target data, safe sharing and privacy protection are realized, and the requirements of a modern power system for efficient and safe data processing are met.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid data sharing, and in particular to a privacy protection-compatible power grid data secure sharing and aggregation method and a multi-center data structure model thereof. Background Art

[0002] In today's digital age, the working mode of power grid enterprises has undergone tremendous changes. With the expansion of business and the improvement of information technology, a huge amount of data and information has been generated within power grid enterprises, including various documents, emails, chat records, project materials, etc. This information is scattered across different platforms and applications, with more data sources and more data types, which makes it difficult for employees to find and obtain the required information, consuming a lot of time and energy.

[0003] In the internal communication of the State Grid, for example, using dedicated software such as Netcom, employees need to extract specific content from a large amount of communication information, such as project progress or upcoming business trip plans. However, traditional search methods have obvious limitations when dealing with complex needs in work scenarios. The current simple keyword search often cannot effectively screen out information that is highly relevant to actual needs, resulting in retrieval results that are both complex and lack of targeting.

[0004] Based on the development of various artificial intelligence and large language models, especially the recent development and application of AIPC concepts, in order to further improve the accuracy and relevance of information extraction from internal power grid operations, it is possible to link the target information to the pre-trained intelligent LLM for intelligent processing. By using the powerful data analysis capabilities of artificial intelligence models, the context and details of the communication content can be analyzed and understood more accurately. However, since Netcom is designed for internal employees of the State Grid, it places special emphasis on privacy protection and secure sharing.

[0005] To this end, it is necessary to develop and integrate advanced parameter space modules and permission access control modules. These modules not only help protect sensitive information and prevent unauthorized access, but also ensure that data is shared on demand to meet the security and privacy requirements of internal communications. In this way, tools such as Netcom can significantly improve the efficiency and accuracy of information retrieval while ensuring information security and leveraging AI's efficient data processing capabilities.

[0006] A Chinese patent application with application publication number CN 117495608 A discloses a data sharing system for power grid dispatching, which acquires power grid data through a data acquisition module, divides the power grid data into first range data and second range data through a data processing module, and encrypts the first range data and the second range data respectively to ensure the security of the first range data and the second range data. The first identity authentication module can verify the user's first identity, and the second identity authentication module can verify the user's second identity. When the user's first identity authentication passes, the first data sharing module sends the first range data to the user end, and when the user's second identity authentication passes, the second data sharing module sends the second range data to the user end, which can improve the security of the power grid data. The user end is used to display the first range data and the second range data, which can facilitate users to search and query the power grid data.

[0007] However, the data sharing system for power grid dispatch developed by the above technology mainly focuses on data encryption and identity authentication, and rarely involves efficient extraction of target data in multiple dimensions, ensuring secure sharing of different access ports, and effectively protecting privacy. Therefore, it is necessary to further develop a method that can quickly and accurately extract the required information from the comprehensive power grid data, so as to achieve secure sharing and aggregation of data, and effectively protect privacy, as well as a multi-center data structure model, in order to meet the needs of modern power systems for efficient and secure data processing.

[0008] In summary, it is particularly important to construct a privacy-compatible power grid data security sharing and aggregation method and its multi-center data structure model, which can not only achieve efficient extraction of target data, multi-dimensional integration, secure sharing and privacy protection, but also meet the needs of modern power systems for efficient and secure data processing. Summary of the invention

[0009] The technical problem to be solved by the present invention is to provide a privacy-protection compatible power grid data secure sharing and aggregation method and its multi-center data structure model, in order to achieve efficient extraction of target data, multi-dimensional integration, secure sharing and privacy protection, and meet the needs of modern power systems for efficient and secure data processing.

[0010] In order to solve the above technical problems, the technical solutions adopted by the present invention are as follows: A privacy-protection compatible power grid data secure sharing and aggregation method, characterized in that it includes integrating the comprehensive power grid data of power workers into multi-level discrete data units through keyword extraction and data association, and implementing multi-dimensional processing on this basis; a central data storage unit formed through iteration, combined with a parameter space module and an access permission control module, to ensure the secure sharing of data.

[0011] Further preferably, the method also strengthens privacy protection by introducing privacy protection technology.

[0012] Further preferably, in the data processing framework, the method includes at least the following data paradigms: information extraction and association, data integration and processing, data iteration and storage, secure sharing and access control, and privacy protection measures.

[0013] Further preferably, in the specific data processing process, the method comprises the following steps: A. Constructing a primary multi-center discrete data unit: extracting information based on keywords from the comprehensive power grid data of a single power operator, recorded as a primary multi-center discrete data unit; further processing the primary multi-center discrete data unit according to the three dimensions of semantics, time and space to form a primary multi-center discrete data unit; the semantic dimension is formed by setting a two-layer structure, the first layer is the search keyword itself, and the second layer is the semantics associated with the keyword currently searched; the time dimension includes at least two time dimension parameters: a sufficiently small time parameter t1 and a sufficiently large time parameter t2; for t1, as long as the time interval of the dialogue / data line in the comprehensive power grid data is less than t1, it is summarized as a paragraph and directly accommodated in the primary multi-center discrete data unit; the space dimension sets a dialogue / data line lower limit, and when the number of dialogue / data lines is lower than this lower limit, the number of context lines of the dialogue / data is automatically expanded to the set lower limit; a two-layer space parameter structure is set corresponding to the semantic dimension; B. Constructing a secondary multi-center discrete data unit: iterating the data records associated with n individuals to form n primary multi-center discrete data units, which are recorded as secondary multi-center discrete data units; C. Constructing a central data storage unit: Iterating the data process of the secondary multi-center discrete data unit among k related power employees to form k secondary multi-center discrete data units as the central data storage unit; D. Security sharing unit: Build a parameter space module to annotate the data by row or paragraph in the central data storage unit; build an access control module to build different access ports based on the parameter set; E. Privacy protection unit: Privacy protection is performed by removing sensitive words or data desensitization or differential privacy or homomorphic encryption or access control or anonymization or pseudo-anonymization technology.

[0014] Further preferably, in step D, parameter annotation is performed on the data by paragraph in the central data storage unit.

[0015] Further preferably, in step D, the parameter space construction module includes constructing a side / additional / independent parameter space: based on the extraction process of the original data, parameter annotation is performed on any row / segment in the central data storage unit; the access permission control module is constructed including construction based on parameter sets, and different parameter sets are assigned to different access ports, so that different ports can have different access permissions to achieve secure sharing.

[0016] It is further preferred that the data process is further iterated between different operating departments of the same power grid unit and across units to form a higher-level data storage unit.

[0017] A multi-center data structure model for secure sharing and aggregation of power grid data, characterized in that: the data model is constructed according to the method described in any one of claims 1-6, and can achieve information extraction integrity based on power data target-related keyword retrieval, while having a good data hierarchy and data logical structure; it at least includes a primary multi-center discrete data unit, a first-level multi-center discrete data unit, a second-level multi-center discrete data unit, and a central data storage unit; wherein the first level includes at least three dimensions: semantics, time, and space.

[0018] The beneficial effects of adopting the above technical solution are: In summary, the present invention, under the premise of considering the secure sharing of power grid information, constructs a new data model that can perform secure external data links and effective access permission control, which can be used as a basic means and practical attempt to introduce AIPC data processing methods within the power grid system.

[0019] In the specific processing of power grid data, for example, a single power grid project based on Netcom involving multiple people, a long time span, and multiple departments / units, it is necessary to perform global data retrieval and processing on these completely scattered data, so as to quickly and efficiently obtain data sets related to the current data target (such as the progress of the current power grid project in a certain aspect, such as development progress, costs, personnel and other data items of concern), and further connect to a large language model for intelligent data analysis. The present invention lays the foundation for realizing such global and intelligent data processing.

[0020] In terms of the technical details of the data model, the present invention uses multi-dimensional (semantic, temporal, spatial) information extraction to more accurately locate and extract information that is highly relevant to the query requirements, thereby significantly improving the efficiency and accuracy of information retrieval; by integrating data into multi-level discrete data units and processing them in multiple dimensions, dynamic data updates and iterative processing are supported, making the data storage structure more flexible and scalable; this structural design allows data to be iteratively updated and shared between different power grid units and departments, improving data utilization and operational efficiency. Through the parameter space module and access control module, data access rights can be effectively controlled, and only authorized users can access relevant data, thereby enhancing data security; the use of advanced privacy protection technology ensures that even in the data sharing process, personal and sensitive information is fully protected, reducing the risk of data leakage. DETAILED DESCRIPTION

[0021] The following examples describe the present invention in detail. In the description of the following examples, specific details such as specific system structures, technologies, etc. are proposed for illustration rather than for limitation, so as to thoroughly understand the present application embodiments. However, it should be clear to those skilled in the art that the present application can also be implemented in other embodiments without these specific details. In other cases, the detailed description of well-known systems, devices, circuits and methods is omitted to prevent unnecessary details from hindering the description of the present application. It should be understood that when used in the present application specification and the attached claims, the term "comprising" indicates the presence of the described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their collections. It should also be understood that the term "and / or" used in the present application specification and the attached claims refers to any combination of one or more of the items listed in association and all possible combinations, and includes these combinations.

[0022] In addition, in the description of the present application specification and the attached claims, the terms "first", "second", "third", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance. The reference to "one embodiment" or "some embodiments" described in the present application specification means that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in the different parts of this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways. Example 1

[0023] In the data processing method of the present invention, the so-called "comprehensive power grid data" can have multiple specific forms, such as extracting information from operation data, maintenance records, fault reports, network communication chat records, etc., and the above operation data, maintenance records, fault reports, network communication chat records and other data sets constitute a set of "comprehensive power grid data". In addition, specific data sets can also be constructed as comprehensive power grid data according to current data processing needs.

[0024] Here we can take the Netcom communication data set within the power grid as an example; similar to conventional instant messaging tools, the information communication among power grid employees on Netcom is also individual-centric; however, the processing of power grid data is often anchored by "a focused work project or a power event", such as the progress of a current power grid project in a certain aspect, such as development progress, costs, personnel and other data items of concern, or power events such as power outages and failures.

[0025] Therefore, in order to extract and analyze data from this multi-dimensional discrete data set (multi-dimensional discreteness includes different employees, different times, different departments, and different network communication subjects, etc.), for one or several power operation data targets of current concern, especially if considering introducing intelligent language models to assist in data analysis and processing, effective data extraction and integration must be carried out first.

[0026] In the face of these problems and data processing requirements, the data system constructed by the present invention can realize the safe sharing of power grid data compatible with privacy protection as a whole, which mainly includes five parts: information extraction and association, data integration and processing, data iteration and storage, safe sharing and access control, and privacy protection measures. Through keyword extraction and data association, the comprehensive power grid data of power workers are integrated into multi-level discrete data units, and multi-dimensional processing is implemented on this basis; the central data storage unit formed by iteration is combined with the parameter space module and the access permission control module to ensure the safe sharing of data; privacy protection is strengthened through privacy protection technology.

[0027] More specifically, in the data system newly constructed by the present invention, operations can be performed according to the following basic steps: a. Extract information based on keywords from the comprehensive power grid data of a single power operator. Here, as mentioned above, the so-called "comprehensive power grid data" can have multiple specific forms, such as extracting information from operation data, maintenance records, fault reports, Netcom data sets, etc., or constructing specific data sets as comprehensive power grid data based on current data processing needs; here, the Netcom chat data set within the power grid is taken as an example; Specifically, the keywords are directly determined by the data target of current concern. In general, the focus of the current power operation project can be split into several independent data targets in advance, such as line loss, failure rate, business expansion progress, cost budget, power outage accidents, etc. Such an independent data focus target can correspond to a keyword; After having a single keyword, we can use the comprehensive power grid data, namely the Netcom chat data set (Netcom data between any two people under multiple people and their combinations; group Netcom data; global Netcom data set under multiple departments, multiple units, and long time spans) as the initial data set, and then we can perform keyword-based information retrieval and extraction.

[0028] Based on keyword retrieval and extraction, we can get discretely distributed data information one by one. Considering the natural influence of time, different users of Netcom, different communication objects and groups of the same person, the above keyword-based information retrieval and extraction will obtain a discrete and multi-center data set. Thus, the corresponding generation and construction can obtain the "primary multi-center discrete data unit". Here, the key is to ensure that the information extraction based on keyword retrieval is complete, and at the same time ensure that the data of the "primary multi-center discrete data unit" has a good data hierarchy and data logical structure. To this end, the data structure of the "primary multi-center discrete data unit" is constructed as follows: The primary multi-center discrete data units are further processed according to the three dimensions of semantics, time and space, and are further constructed to form the first-level multi-center discrete data units.

[0029] In the semantic dimension, a two-layer structure is set up. The first layer is to search for the keyword itself, and the second layer is to search for the semantics associated with the current keyword. For example, if there are 10 second-layer semantic words related to the keyword, these 10 associated semantic words are constructed into an associated semantic feature vector, that is, a 10-dimensional feature vector is obtained; for example: analyzing other keywords related to "power outage", such as "equipment failure, weather warning, power outage, power outage maintenance, power outage plan, power outage notice, power outage recovery, power outage accident, power outage notice, power outage area", etc.; richer semantic associations can be formed, thereby expanding the space for data search and information extraction.

[0030] The time dimension includes at least two time dimension parameters: a sufficiently small time parameter t1 and a sufficiently large time parameter t2. For t1, as long as the time interval of the conversation / data line in Netcom is less than t1, it is summarized as a paragraph and directly accommodated in the primary multi-center discrete data unit. For t2, for example, if t2 is set to 24h, that is, one day, the larger δt-span can be used to distinguish and integrate the primary multi-center discrete data units. The simplest example is: the retrieval data spans k t2s, that is, k days, which correspond to k primary multi-center discrete data units, which are merged as a first-level multi-center discrete data unit.

[0031] Therefore, we can see that the time dimension has two data values: one is to directly generate a paragraphed data layer structure as a primary multi-center discrete data unit; the other is to merge the primary multi-center discrete data units based on a larger time parameter, and the merged data set is used as a first-level multi-center discrete data unit.

[0032] The spatial dimension refers to the number of context lines. A lower limit is set, which means that when the number of lines is lower than the lower limit, the context / line expansion is performed. A two-layer spatial parameter structure is set corresponding to the semantic dimension: the first layer is core keywords, for example, a spatial parameter of 500 lines is set, and the second layer is associated semantic feature vectors. The spatial parameter is set to 100 lines, and there is no upper limit.

[0033] b. Iterate the data records associated with n individuals to form n first-level multi-center discrete data units, which are recorded as second-level multi-center discrete data units; c. Iterate the data process of the secondary multi-center discrete data unit among k related power employees to form k secondary multi-center discrete data units as central data storage units; the construction of the central data storage unit can be connected to the LLM model to perform AI intelligent data processing and AI search; d. Security sharing unit: Build parameter space modules, such as data sensitivity levels, "public", "internal", "confidential", and parameterize data by row or paragraph in the central data storage unit; build access control modules, and build different access ports based on parameter sets. Build a side / extra / independent parameter space: First, based on the extraction process of the original data, parameterize any row in the central data storage unit (it may be multiple parameters, for example, if the content of this row comes from the chat between A and B, then this row is labeled with 2 parameters; group chats correspond to multiple parameters;) Secondly, based on the construction of parameter sets, assign different parameter sets to different access ports, so that different ports can have different access rights; perform secure sharing; h. Privacy protection unit: Privacy protection is performed by removing sensitive words or data desensitization or differential privacy or homomorphic encryption or access control or anonymization or pseudo-anonymization technology.

[0034] The data process is further iterated within different operating departments of the same power grid unit and across units to form higher-level data storage units. Example 2

[0035] A privacy-protection-compatible power grid data security sharing and aggregation method does not conduct a full network search, but focuses on extracting information based on keywords from comprehensive power grid data by searching users, such as extracting information from operation data, maintenance records, fault reports, Netcom chat records, etc. Not only can it search for comprehensive information, but it can also use AI to integrate the extracted information, such as sorting out comprehensive information about who and in which scenario (such as which group, which document, which meeting, or which business process, etc.).

[0036] The current situation is that the internal documents and project management of the power grid company are in chaos, including various work documents and project group chats. Even different groups of people for the same project set up multiple group chats, so that the information to be sent is often sent to the wrong group. Therefore, it is very troublesome every time we want to track the progress of some projects or find some information and data. So we are looking forward to a method that can process and integrate all the searched information based on the search function so that we can make the next decision.

[0037] The following is a specific use process of the present invention: Assume that X Power Grid Company will have meetings with different themes in the next five days, and different quantities of coffee need to be reserved according to different participants. First, search for coffee in the Netcom and WeChat chat records between employee A and employee B, and set a lower limit of conversation lines, such as 500 lines; then search for semantics related to coffee, such as latte, Americano, iced, heated, sugared, etc., and set a lower limit of conversation lines, such as 200 lines; set a continuous conversation line as a continuous conversation line within 5 seconds, and integrate the searched content into a first-level multi-center discrete data unit. Search for coffee in the Netcom and WeChat chat records between employee A and 15 employees, and form 15 first-level multi-center discrete data units, which are recorded as second-level multi-center discrete data units; iterate the data process of the second-level multi-center discrete data unit among 30 related power employees, and form 30 second-level multi-center discrete data units as central data storage units. The central data storage unit stores the coffee required by different participants of a meeting on a certain day, including the quantity and content of coffee. Build a parameter space module to annotate parameters for each row of data in the central data storage unit; build an access control module to build different access ports based on parameter sets, and assign different parameter sets to different access ports, so that different ports can have different access rights to ensure secure information sharing; then use differential privacy and other technologies to restrict only the administrative staff who finally ordered the coffee to see all the content, and individuals can only see the content they uploaded. Finally, link the central data storage unit to the LLM model for AI intelligent data processing, which can analyze the various needs of the participants in each meeting for coffee, so that administrative staff can place orders and make reservations.

[0038] The present invention uses multi-dimensional (semantic, temporal, spatial) information extraction to more accurately locate and extract information that is highly relevant to the query requirements, thereby significantly improving the efficiency and accuracy of information retrieval; By integrating data into multi-level discrete data units and processing them in multiple dimensions, dynamic data updates and iterative processing are supported, making the data storage structure more flexible and scalable; this structural design allows data to be iteratively updated and shared between different power grid units and departments, improving data utilization and operational efficiency.

[0039] Through the parameter space module and access permission control module, data access permissions can be effectively controlled, and only authorized users can access relevant data, thereby enhancing data security; the use of advanced privacy protection technology ensures that personal and sensitive information is fully protected even during data sharing, reducing the risk of data leakage.

[0040] Finally, it should be noted that the above are only preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the scope of protection of the claims of the present invention.

Claims

1. A privacy protection compatible power grid data secure sharing and aggregation method, characterized in that: This includes integrating the comprehensive power grid data of power operators into multi-level discrete data units through keyword extraction and data association, and implementing multi-dimensional processing on this basis; ensuring the secure sharing of data through the iteratively formed central data storage unit, combined with the parameter space module and the access permission control module.

2. A privacy protection compatible power grid data secure sharing and aggregation method according to claim 1, characterized in that: This method also strengthens privacy protection by introducing privacy protection technology.

3. A privacy protection compatible power grid data secure sharing and aggregation method according to claim 1 or 2, characterized in that: In the data processing framework, the method includes at least the following data paradigms: information extraction and association, data integration and processing, data iteration and storage, secure sharing and access control, and privacy protection measures.

4. A privacy protection compatible power grid data secure sharing and aggregation method according to claim 1 or 2, characterized in that: In the specific data processing process, the method includes the following steps: A. Constructing a primary multi-center discrete data unit: extracting information based on keywords from the comprehensive power grid data of a single power operator, recorded as a primary multi-center discrete data unit; further processing the primary multi-center discrete data unit according to the three dimensions of semantics, time and space to form a primary multi-center discrete data unit; the semantic dimension is formed by setting a two-layer structure, the first layer is the search keyword itself, and the second layer is the semantics associated with the keyword currently searched; the time dimension includes at least two time dimension parameters: a sufficiently small time parameter t1 and a sufficiently large time parameter t2; for t1, as long as the time interval of the dialogue / data line in the comprehensive power grid data is less than t1, it is summarized as a paragraph and directly accommodated in the primary multi-center discrete data unit; the space dimension sets a dialogue / data line lower limit, and when the number of dialogue / data lines is lower than this lower limit, the number of context lines of the dialogue / data is automatically expanded to the set lower limit; a two-layer space parameter structure is set corresponding to the semantic dimension; B. Constructing a secondary multi-center discrete data unit: iterating the data records associated with n individuals to form n primary multi-center discrete data units, which are recorded as secondary multi-center discrete data units; C. Constructing a central data storage unit: Iterating the data process of the secondary multi-center discrete data unit among k related power employees to form k secondary multi-center discrete data units as the central data storage unit; D. Security sharing unit: Build a parameter space module to annotate the data by row or paragraph in the central data storage unit; build an access control module to build different access ports based on the parameter set; E. Privacy protection unit: Privacy protection is performed by removing sensitive words or data desensitization or differential privacy or homomorphic encryption or access control or anonymization or pseudo-anonymization technology.

5. A privacy protection compatible power grid data secure sharing and aggregation method according to claim 4, characterized in that: In step D, the data is parameter-labeled by paragraph in the central data storage unit.

6. A privacy protection compatible power grid data secure sharing and aggregation method according to claim 4, characterized in that: In step D, the parameter space construction module includes constructing a side / additional / independent parameter space: based on the extraction process of the original data, parameter annotation is performed on any row / segment in the central data storage unit; the access permission control module is constructed based on parameter sets, and different parameter sets are assigned to different access ports, so that different ports can have different access permissions to achieve secure sharing.

7. A privacy protection compatible power grid data secure sharing and aggregation method according to claim 4, characterized in that: The data process is further iterated within different operating departments of the same power grid unit and across units to form higher-level data storage units.

8. A multi-center data structure model for secure sharing and aggregation of power grid data, characterized by: The data model is constructed according to the method described in any one of claims 1-7, and can achieve information extraction integrity based on power data target-related keyword retrieval, and has a good data hierarchy and data logical structure; it at least includes a primary multi-center discrete data unit, a first-level multi-center discrete data unit, a second-level multi-center discrete data unit, and a central data storage unit; wherein the first level includes at least three dimensions: semantics, time, and space.