CN115834257A - Cloud electric power data safety protection method and protection system - Google Patents

Cloud electric power data safety protection method and protection system Download PDF

Info

Publication number
CN115834257A
CN115834257A CN202310133859.9A CN202310133859A CN115834257A CN 115834257 A CN115834257 A CN 115834257A CN 202310133859 A CN202310133859 A CN 202310133859A CN 115834257 A CN115834257 A CN 115834257A
Authority
CN
China
Prior art keywords
interference
data
matrix
private
private data
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202310133859.9A
Other languages
Chinese (zh)
Other versions
CN115834257B (en
Inventor
张实君
来骥
闫忠平
杨峰
李平舟
赵佳琦
李硕
徐相森
曾婧
杨睿
李贤�
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
State Grid Corp of China SGCC
State Grid Jibei Electric Power Co Ltd
Information and Telecommunication Branch of State Grid Jibei Electric Power Co Ltd
Original Assignee
State Grid Corp of China SGCC
State Grid Jibei Electric Power Co Ltd
Information and Telecommunication Branch of State Grid Jibei Electric Power Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by State Grid Corp of China SGCC, State Grid Jibei Electric Power Co Ltd, Information and Telecommunication Branch of State Grid Jibei Electric Power Co Ltd filed Critical State Grid Corp of China SGCC
Priority to CN202310133859.9A priority Critical patent/CN115834257B/en
Publication of CN115834257A publication Critical patent/CN115834257A/en
Application granted granted Critical
Publication of CN115834257B publication Critical patent/CN115834257B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S40/00—Systems for electrical power generation, transmission, distribution or end-user application management characterised by the use of communication or information technologies, or communication or information technology specific aspects supporting them
    • Y04S40/20—Information technology specific aspects, e.g. CAD, simulation, modelling, system security

Landscapes

  • Storage Device Security (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention provides a cloud electric power data safety protection method and a protection system, and relates to the technical field of cloud electric power data safety.A processor processes and stores data generated by a terminal device in a cloud electric power data system, marks a node address of each data, and divides the data generated by the terminal device into a private data block and a common data block; classifying the private data blocks, and performing key processing of different degrees on the private data blocks of different types; performing matrix compression on the private data matrix after the key processing by using a matrix compression algorithm to obtain an interference matrix; and carrying out data interference mining on the interference matrix to form an interference attribute data set, and replacing the interference attribute value of the interference attribute data set with the interference noise value, so that the safety of cloud electric power data transmission and storage is improved.

Description

一种云端电力数据安全防护方法及防护系统A cloud power data security protection method and protection system

技术领域technical field

本发明涉及电力数据安全技术领域，具体涉及一种云端电力数据安全防护方法及防护系统。The invention relates to the technical field of power data security, in particular to a cloud power data security protection method and protection system.

背景技术Background technique

目前电力系统传统IT基础设施中，网络管理、存储管理、计算管理是割裂的三大系统，导致所有的应用和安全策略部署，不得不面临多系统的协同配合，以及大量的人工操作。同时，目前电力系统传统IT基础设施在发生网络故障或者攻击时，只是通过强制机械关停或重启的方式进行应对，这种应对机制较简单，不能确保电力系统中的各个网络单元之间互不影响，不能将攻击或者故障限制在一定范围之内。因此，如何将传统IT资源向软件定义数据中心上迁移，在保证对当前运行的业务与应用影响降到最小的同时，还能够保证电力系统中的各个网络单元之间互不影响，并将攻击或者故障限制在一定范围之内，是目前有待研究攻克的技术难点。At present, in the traditional IT infrastructure of the power system, network management, storage management, and computing management are three separate systems. As a result, all applications and security policy deployments have to face the coordination of multiple systems and a large number of manual operations. At the same time, when a network failure or attack occurs in the traditional IT infrastructure of the power system, it only responds by forcing the machine to shut down or restart. Influence, attacks or failures cannot be limited to a certain range. Therefore, how to migrate traditional IT resources to the software-defined data center, while ensuring the minimum impact on the currently running business and applications, can also ensure that each network unit in the power system does not affect each other, and will Or the fault is limited within a certain range, which is a technical difficulty to be studied and overcome at present.

电网设备的运行和防护对于保证电网的稳定和可靠性至关重要。随着电网规模的扩大和对客户服务的更高要求，运行和维护电网的任务逐渐变得更加艰巨。传统运行和防护电网模式已经无法满足日益复杂的需求。在传统模式下，不同部门数据之间相互兼容较差，技术人员因为没有办法有效地分析和诊断电网状态，无法实时防护电力数据。因此，大数据、云计算、物联网和移动互联网技术正在推动智能电网发展与建设，开发新的运行和防护技术变得极其关键，也是智能电网发展的方向之一。The operation and protection of grid equipment is crucial to ensure the stability and reliability of the grid. The task of operating and maintaining the grid has gradually become more difficult as the scale of the grid has expanded and the requirements for customer service have increased. The traditional operation and protection grid mode can no longer meet the increasingly complex needs. In the traditional mode, data from different departments are not compatible with each other, and technicians cannot protect power data in real time because they have no way to effectively analyze and diagnose the status of the power grid. Therefore, big data, cloud computing, Internet of Things and mobile Internet technologies are promoting the development and construction of smart grids. The development of new operation and protection technologies has become extremely critical, and it is also one of the directions for the development of smart grids.

但无线通信链路的开放性也导致了其通信过程极易遭到第三者窃听、通信内容易被篡改等数据安全的威胁。此外，传统技术中对网络层数据密钥处理的密钥处理算法安全程度一般取决于数学模型的求解难度，无法保证通信安全。这些问题对计算能力有限、存储空间较小的无线传感网络都非常不利。However, the openness of the wireless communication link also makes the communication process vulnerable to data security threats such as eavesdropping by a third party and tampering in the communication. In addition, the security level of key processing algorithms for network layer data key processing in traditional technologies generally depends on the difficulty of solving mathematical models, which cannot guarantee communication security. These problems are very unfavorable to wireless sensor networks with limited computing power and small storage space.

为了保障云存储系统中的电力数据安全，特别是私密，目前常见的方法仍然是基于传统的数据密钥处理技术，即简单地用某种密钥处理技术将密钥处理后的数据托管到云存储系统中。现行机制存在的问题有：数据文件的内容对数据信息不加区分，作为一个整体进行相同的密钥处理；数据文件的整体存储，用户访问策略不能细分，即不能通过访问策略控制相应用户对私密数据的访问；数据托管方无法改变公有云本身存在的数据存储的不安全性；电力行业自身拥有的云存储设备没有充分利用，以提升存储安全性。In order to ensure the security of power data in the cloud storage system, especially the privacy, the current common method is still based on the traditional data key processing technology, that is, simply use some key processing technology to host the data processed by the key to the cloud in the storage system. The problems existing in the current mechanism are: the content of the data file does not distinguish the data information, and the same key is processed as a whole; the overall storage of the data file, the user access policy cannot be subdivided, that is, the corresponding user cannot be controlled by the access policy. Access to private data; the data custodian cannot change the insecurity of data storage in the public cloud itself; the cloud storage equipment owned by the power industry itself is not fully utilized to improve storage security.

发明内容Contents of the invention

为了解决上述技术问题，本发明提出了一种云端电力数据安全防护方法，包括如下步骤：In order to solve the above technical problems, the present invention proposes a cloud power data security protection method, comprising the following steps:

S1、对终端装置产生的数据进行处理及存储，将终端装置产生的数据分割为私密数据块和普通数据块；S1. Process and store the data generated by the terminal device, and divide the data generated by the terminal device into private data blocks and common data blocks;

S2、将私密数据块进行分类，将不同类别的私密数据块进行不同程度的密钥处理，得到密钥处理后的私密数据矩阵；S2. Classifying the private data blocks, performing key processing on different types of private data blocks to obtain a private data matrix after key processing;

S3、利用矩阵压缩算法对密钥处理后的所述私密数据矩阵进行矩阵压缩，得到干扰矩阵；S3. Using a matrix compression algorithm to perform matrix compression on the secret data matrix after key processing to obtain an interference matrix;

S4、对干扰矩阵进行数据干扰挖掘，形成干扰属性数据集，用干扰噪声值替换干扰属性数据集中的干扰属性值。S4. Perform data interference mining on the interference matrix to form an interference attribute data set, and replace the interference attribute values in the interference attribute data set with interference noise values.

进一步地，步骤S3中：Further, in step S3:

给定一组m列×n行的私密数据矩阵

Figure SMS_1
，私密数据矩阵
Figure SMS_2
通过矩阵压缩算法压缩得到m列×t行的左矩阵
Figure SMS_3
和t列×n行的右矩阵
Figure SMS_4
：Given a set of private data matrices with m columns x n rows
Figure SMS_1
, the secret data matrix
Figure SMS_2
The left matrix of m columns × t rows is obtained by compressing the matrix compression algorithm
Figure SMS_3
and the right matrix of t columns by n rows
Figure SMS_4
:

Figure SMS_5
；
Figure SMS_5
;

其中，私密数据矩阵

Figure SMS_6
最大化压缩为左矩阵
Figure SMS_7
和右矩阵
Figure SMS_8
，左矩阵
Figure SMS_9
和右矩阵
Figure SMS_10
压缩为多个干扰矩阵：m列×a行的干扰矩阵
Figure SMS_11
和a列×n行的干扰矩阵
Figure SMS_12
。Among them, the private data matrix
Figure SMS_6
maximize compression as left matrix
Figure SMS_7
and the right matrix
Figure SMS_8
, the left matrix
Figure SMS_9
and the right matrix
Figure SMS_10
Compressed into multiple interference matrices: interference matrix with m columns × a rows
Figure SMS_11
and an interference matrix of a column by n row
Figure SMS_12
.

进一步地，使用迭代算法求解左矩阵

Figure SMS_13
和右矩阵
Figure SMS_14
的干扰矩阵，定义干扰度为
Figure SMS_15
，通过如下公式计算：Further, use an iterative algorithm to solve the left matrix
Figure SMS_13
and the right matrix
Figure SMS_14
The interference matrix, define the interference degree as
Figure SMS_15
, calculated by the following formula:

Figure SMS_16
；
Figure SMS_17
；
Figure SMS_16
;
Figure SMS_17
;

其中，

Figure SMS_18
为私密数据矩阵
Figure SMS_19
的i列×
Figure SMS_20
行的子矩阵，使干扰度
Figure SMS_21
最大的
Figure SMS_22
、
Figure SMS_23
，即为干扰矩阵。in,
Figure SMS_18
is the secret data matrix
Figure SMS_19
column i ×
Figure SMS_20
Row submatrix such that the disturbance
Figure SMS_21
biggest
Figure SMS_22
,
Figure SMS_23
, which is the interference matrix.

进一步地，步骤S4包括如下步骤：Further, step S4 includes the following steps:

S41、获取干扰矩阵

Figure SMS_24
、
Figure SMS_27
的数据，用干扰矩阵
Figure SMS_30
中的最后一列乘以干扰矩阵
Figure SMS_26
中的第一行，得到原始干扰数据
Figure SMS_28
，从原始干扰数据
Figure SMS_29
中随机选取k个数据作为干扰属性值，构成干扰属性数据集：
Figure SMS_31
；其中
Figure SMS_25
为干扰属性数据集中的第j个干扰属性值。S41. Obtain the interference matrix
Figure SMS_24
,
Figure SMS_27
data, using the interference matrix
Figure SMS_30
The last column in is multiplied by the interference matrix
Figure SMS_26
In the first row, get the original interference data
Figure SMS_28
, from the raw noise data
Figure SMS_29
Randomly select k pieces of data as the interference attribute values to form the interference attribute data set:
Figure SMS_31
;in
Figure SMS_25
is the jth disturbance attribute value in the disturbance attribute dataset.

S42、对于每个干扰属性值

Figure SMS_32
，定义一个干扰区间，表示为：S42. For each disturbance attribute value
Figure SMS_32
, define an interference interval, expressed as:

Figure SMS_33
；
Figure SMS_33
;

其中

Figure SMS_34
分别为干扰区间的下限和上限，分别由下述公式计算：in
Figure SMS_34
are the lower limit and upper limit of the interference interval, respectively, calculated by the following formulas:

Figure SMS_35
；
Figure SMS_35
;

Figure SMS_36
；
Figure SMS_36
;

其中，

Figure SMS_37
为干扰属性数据集
Figure SMS_38
的归一化值。in,
Figure SMS_37
is the disturbance attribute data set
Figure SMS_38
normalized value of .

S43、得到干扰属性值

Figure SMS_39
的干扰区间为
Figure SMS_40
，由以下公式计算干扰噪声值
Figure SMS_41
：S43. Get the disturbance attribute value
Figure SMS_39
The interference interval is
Figure SMS_40
, the interference noise value is calculated by the following formula
Figure SMS_41
:

Figure SMS_42
；
Figure SMS_42
;

其中N为随机数。where N is a random number.

S44、将干扰噪声值

Figure SMS_43
替换干扰属性数据集
Figure SMS_44
中的干扰属性值
Figure SMS_45
。S44, will disturb the noise value
Figure SMS_43
Replace the noise attribute dataset
Figure SMS_44
Interference attribute values in
Figure SMS_45
.

进一步地，步骤S2具体包括如下步骤：Further, step S2 specifically includes the following steps:

S21、将私密数据块传入时间卷积块，私密数据块在时间卷积块中先通过时间卷积层，再使用批量归一化通过ReLU激活函数得到该时间卷积块的输出，传送到下一个时间卷积块，重复２次上述过程；S21. Pass the private data block into the time convolution block. In the time convolution block, the private data block first passes through the time convolution layer, and then uses batch normalization to obtain the output of the time convolution block through the ReLU activation function, and transmits it to For the next time convolution block, repeat the above process twice;

S22、经过３个堆叠的时间卷积块输出的数据进入全局平均池化层和时间递归神经网络；S22. The data output by the three stacked time convolution blocks enter the global average pooling layer and the time recurrent neural network;

S23、将全局平均池化层和时间递归神经网络的输出结果进行串联，发送到分类层进行分类，得到不同类别的私密数据块；S23. Connect the output results of the global average pooling layer and the time recurrent neural network in series, and send them to the classification layer for classification to obtain private data blocks of different categories;

S24、将不同类别的私密数据块进行不同程度的密钥处理，得到密钥处理后的私密数据矩阵。S24. Perform different levels of key processing on different types of private data blocks to obtain a key-processed private data matrix.

本发明还提出了一种云端电力数据安全防护系统，用于实现云端电力数据安全防护方法，包括：电力数据系统、处理器、终端装置、云端处理层；The present invention also proposes a cloud power data security protection system for implementing a cloud power data security protection method, including: a power data system, a processor, a terminal device, and a cloud processing layer;

在所述电力数据系统中，所述处理器对终端装置产生的数据进行处理及存储，将终端装置产生的数据分为私密数据块和普通数据块；In the power data system, the processor processes and stores the data generated by the terminal device, and divides the data generated by the terminal device into private data blocks and common data blocks;

云端处理层包括级别分类单元、矩阵压缩单元、替换单元和云服务器；所述级别分类单元将私密数据块进行分类，将不同类别的私密数据块进行不同程度的密钥处理，得到密钥处理后的私密数据矩阵；所述矩阵压缩单元利用矩阵压缩算法对密钥处理后的私密数据矩阵进行矩阵压缩，得到干扰矩阵；所述替换单元对干扰矩阵进行数据干扰挖掘，形成干扰属性数据集，用干扰噪声值替换干扰属性数据集中的干扰属性值。The cloud processing layer includes a level classification unit, a matrix compression unit, a replacement unit, and a cloud server; the level classification unit classifies the private data blocks, performs different key processing on different types of private data blocks, and obtains key processing The secret data matrix; the matrix compression unit uses a matrix compression algorithm to perform matrix compression on the private data matrix after key processing to obtain an interference matrix; the replacement unit performs data interference mining on the interference matrix to form an interference attribute data set, using Disturbance noise values replace the disturbance attribute values in the disturbance attribute dataset.

进一步地，所述处理器包括标记单元和分割单元，所述标记单元按照数据属性，对每个数据的结点地址进行标记，所述分割单元用于根据标记将终端装置产生的数据分为私密数据块和普通数据块。Further, the processor includes a labeling unit and a segmentation unit, the labeling unit labels the node address of each data according to the data attribute, and the segmentation unit is used to divide the data generated by the terminal device into private Data blocks and normal data blocks.

进一步地，所述级别分类单元还包括密钥因子控制装置、密钥因子生成装置和密钥分发装置；所述密钥因子控制装置，用于生成密钥因子划分参数；所述密钥因子生成装置接收密钥因子划分参数，并将密钥因子划分参数划分为多个划分子参数，作为私钥因子；所述密钥分发装置根据私密数据矩阵每一行的序号将多个私钥因子发送给私密数据矩阵的每一行，对私密数据进行密钥处理。Further, the level classification unit also includes a key factor control device, a key factor generation device and a key distribution device; the key factor control device is used to generate key factor division parameters; the key factor generation The device receives the key factor division parameter, and divides the key factor division parameter into multiple sub-parameters as the private key factor; the key distribution device sends multiple private key factors to Each row of the private data matrix performs key processing on the private data.

相比于现有技术，本发明具有如下有益技术效果：Compared with the prior art, the present invention has the following beneficial technical effects:

对终端装置产生的数据进行处理及存储，将终端装置产生的数据分为私密数据块和普通数据块；将私密数据块进行分类，将不同类别的私密数据块进行不同程度的密钥处理；利用矩阵压缩算法对密钥处理后的私密数据矩阵进行矩阵压缩，得到干扰矩阵；对干扰矩阵进行数据干扰挖掘，形成干扰属性数据集，用干扰噪声值替换干扰属性数据集中的干扰属性值，提升了传输和存储的安全性。Process and store the data generated by the terminal device, divide the data generated by the terminal device into private data blocks and ordinary data blocks; classify the private data blocks, and perform different key processing on different types of private data blocks; use The matrix compression algorithm performs matrix compression on the private data matrix after key processing to obtain the interference matrix; performs data interference mining on the interference matrix to form an interference attribute data set, and replaces the interference attribute values in the interference attribute data set with the interference noise value, improving the Security of transmission and storage.

附图说明Description of drawings

为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings that need to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those skilled in the art, other drawings can also be obtained based on these drawings without any creative effort.

图1为本发明的云端电力数据安全防护方法的流程图。FIG. 1 is a flow chart of the cloud power data security protection method of the present invention.

图2为本发明的数据干扰挖掘及替换的方法流程图。FIG. 2 is a flow chart of the data interference mining and replacement method of the present invention.

图3为本发明的云端电力数据安全防护系统的结构示意图。FIG. 3 is a schematic structural diagram of the cloud power data security protection system of the present invention.

具体实施方式Detailed ways

为使本申请实施例的目的、技术方案和优点更加清楚，下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments It is a part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the scope of protection of this application.

在本发明的具体实施例附图中，为了更好、更清楚的描述系统中的各元件的工作原理，表现所述装置中各部分的连接关系，只是明显区分了各元件之间的相对位置关系，并不能构成对元件或结构内的信号传输方向、连接顺序及各部分结构大小、尺寸、形状的限定。In the accompanying drawings of specific embodiments of the present invention, in order to better and more clearly describe the working principle of each component in the system, the connection relationship of each part in the device is shown, and only the relative positions between the components are clearly distinguished. The relationship does not constitute a limitation on the signal transmission direction, connection sequence, and the size, size, and shape of each part of the component or structure.

如图1所示，为本发明的云端电力数据安全防护方法的流程图，包括如下步骤：As shown in Figure 1, it is a flow chart of the cloud power data security protection method of the present invention, including the following steps:

S1、在云端电力数据系统中，处理器对终端装置产生的数据进行处理及存储，将终端装置产生的数据分为私密数据块和普通数据块。S1. In the cloud power data system, the processor processes and stores the data generated by the terminal device, and divides the data generated by the terminal device into private data blocks and common data blocks.

S11、对终端装置产生的数据进行索引存储，采用关键字识别技术，对每个数据的结点地址进行标记。S11. Index and store the data generated by the terminal device, and mark the node address of each data by using keyword recognition technology.

索引存储为在储存数据结点信息的同时，还建立附加的索引表。索引表由若干索引项组成。若每个结点在索引表中都有一个索引项，则该索引表称之为稠密索引。索引项的一般形式是：关键字+地址。关键字是能唯一标记一个结点的数据项，稠密索引中索引项的地址指示结点所在的存储位置。Index storage is to create an additional index table while storing data node information. An index table consists of several index entries. If each node has an index entry in the index table, the index table is called a dense index. The general form of an index item is: keyword + address. The key is a data item that can uniquely mark a node, and the address of the index item in the dense index indicates the storage location of the node.

S12、读取终端装置产生的数据的结点地址，根据结点地址的标记信息判定该结点地址为私密数据块的开始位置或者普通数据块的开始位置。S12. Read the node address of the data generated by the terminal device, and determine whether the node address is the start position of the private data block or the start position of the normal data block according to the tag information of the node address.

S13、依次读取下一个结点地址，直到下一结点地址的标记信息与开始位置的标记信息不同，对数据进行分割，生成私密数据块和普通数据块。S13. Read the address of the next node in sequence until the tag information of the next node address is different from the tag information of the start position, and divide the data to generate a private data block and a common data block.

当下一结点地址的标记信息与开始位置的标记信息不同时，说明两个位置所对应的数据属于不同属性的数据，所以需要以该结点地址作为分界点进行分割。When the tag information of the next node address is different from the tag information of the start position, it means that the data corresponding to the two positions belong to data with different attributes, so the node address needs to be used as the dividing point for segmentation.

优选地，可以将每个终端装置产生的数据文件压缩成多个数据块，完全具备接收到其中任意至少一半数量的数据块时，才能恢复原文件；当任意一半或少于一半数量的数据块丢失或损坏时仍能恢复原文件，从而提高了可靠性和可用性；同时任意不足一半数量的数据块被窃取时，不能还原成原文件，从而提高了安全性。若所有数据块均存储于云端处理层，则防护系统仍能获取数据私密信息。Preferably, the data files generated by each terminal device can be compressed into multiple data blocks, and the original file can only be restored when any at least half of the data blocks are fully equipped; when any half or less than half of the data blocks The original file can still be restored when it is lost or damaged, thereby improving reliability and availability; at the same time, when any data block less than half is stolen, it cannot be restored to the original file, thereby improving security. If all data blocks are stored in the cloud processing layer, the protection system can still obtain data private information.

S2、将私密数据块进行分类，将不同类别的私密数据块进行不同程度的密钥处理。S2. Classify the private data blocks, and perform different levels of key processing on different types of private data blocks.

处理器从云端处理层接收私密数据块，对私密数据块进行级别分类处理以便于之后的密钥处理。The processor receives the private data blocks from the cloud processing layer, and classifies the private data blocks to facilitate subsequent key processing.

优选地，采用LSTM-FCN对数据进行分类，分类过程如下。Preferably, LSTM-FCN is used to classify the data, and the classification process is as follows.

S21、将私密数据块传入时间卷积块，私密数据块在时间卷积块中先通过时间卷积层，再使用批量归一化，之后通过ReLU激活函数得到该时间卷积块的输出，输出再作为输入传送到下一个时间卷积块，重复２次上述过程。S21. Pass the private data block into the time convolution block. The private data block first passes through the time convolution layer in the time convolution block, then uses batch normalization, and then obtains the output of the time convolution block through the ReLU activation function. The output is then sent to the next time convolution block as the input, and the above process is repeated twice.

S22、经过３个堆叠的时间卷积块的私密数据块进入全局平均池化层和时间递归神经网络。S22. The private data block after three stacked temporal convolution blocks enters the global average pooling layer and the temporal recurrent neural network.

S23、将全局平均池化层和时间递归神经网络的输出进行串联，发送到分类层进行分类，得到不同类别的私密数据块。S23. Connect the outputs of the global average pooling layer and the time recurrent neural network in series, and send them to the classification layer for classification to obtain private data blocks of different categories.

S24、将不同类别的私密数据块进行不同程度的密钥处理，得到密钥处理后的私密数据矩阵并存储。S24. Perform different levels of key processing on different types of private data blocks to obtain and store a key-processed private data matrix.

将私密数据块按照不同类别划分为不同级别的私密数据块，按照私密数据块的级别进行不同程度的密钥处理，得到密钥处理后的私密数据矩阵，私密数据矩阵的每一行代表一个密钥处理级别。Divide the private data blocks into different levels of private data blocks according to different categories, and perform different levels of key processing according to the level of private data blocks to obtain the private data matrix after key processing. Each row of the private data matrix represents a key processing level.

在优选实施例中，通过密钥因子控制装置生成密钥因子划分参数，每一个密钥因子划分参数对应一个密钥处理级别；通过密钥因子生成装置接收密钥因子划分参数，并将密钥因子划分参数划分为多个划分子参数，作为私钥因子，通过密钥分发装置根据与私密数据矩阵的每一行的序号将多个私钥因子发送给私密数据矩阵的每一行，从而对该行的私密数据进行密钥处理。In a preferred embodiment, the key factor division parameter is generated by the key factor control device, and each key factor division parameter corresponds to a key processing level; the key factor generation device receives the key factor division parameter, and the key The factor division parameter is divided into a plurality of division sub-parameters, and as the private key factor, a plurality of private key factors are sent to each row of the private data matrix by the key distribution device according to the sequence number of each row of the private data matrix, so that the row private data for key processing.

S3、利用矩阵压缩算法对密钥处理后的私密数据矩阵进行矩阵压缩，得到干扰矩阵。S3. Use a matrix compression algorithm to perform matrix compression on the private data matrix after the key processing to obtain an interference matrix.

矩阵压缩算法将私密数据矩阵压缩为左右两个矩阵的乘积，私密数据矩阵中的一列向量为左矩阵中所有列向量的加权和，而权重系数为右矩阵中对应列向量中的元素。基于迭代计算的矩阵压缩算法具有收敛速度快、左右矩阵存储空间容量小的特点。The matrix compression algorithm compresses the private data matrix into the product of the left and right matrices. A column vector in the private data matrix is the weighted sum of all column vectors in the left matrix, and the weight coefficient is the element in the corresponding column vector in the right matrix. The matrix compression algorithm based on iterative calculation has the characteristics of fast convergence speed and small storage space capacity of left and right matrices.

具体地，给定一组m列×n行的私密数据矩阵

Figure SMS_46
，私密数据矩阵
Figure SMS_47
通过矩阵压缩算法压缩得到左右两个矩阵：m列×t行的左矩阵
Figure SMS_48
和t列×n行的右矩阵
Figure SMS_49
：Specifically, given a set of secret data matrix with m columns×n rows
Figure SMS_46
, the secret data matrix
Figure SMS_47
The left and right matrices are obtained by compressing the matrix compression algorithm: the left matrix of m columns × t rows
Figure SMS_48
and the right matrix of t columns by n rows
Figure SMS_49
:

Figure SMS_50
；
Figure SMS_50
;

其中，t值的选择满足（m+n）t<mn，则私密数据矩阵

Figure SMS_52
可最大化压缩为左矩阵
Figure SMS_55
和右矩阵
Figure SMS_59
，左矩阵
Figure SMS_51
的每一列包含了一个基向量，这组基向量构成了一个t维的空间，右矩阵
Figure SMS_54
的每一列则为私密数据矩阵
Figure SMS_57
对应列向量在该t维空间中的投影。左矩阵
Figure SMS_60
和右矩阵
Figure SMS_53
又可以压缩为多个干扰矩阵，即m列×a行的干扰矩阵
Figure SMS_56
和a列×n行的干扰矩阵
Figure SMS_58
。Among them, the choice of t value satisfies (m+n)t<mn, then the secret data matrix
Figure SMS_52
Maximally compressible to the left matrix
Figure SMS_55
and the right matrix
Figure SMS_59
, the left matrix
Figure SMS_51
Each column of contains a basis vector, and this set of basis vectors constitutes a t-dimensional space, the right matrix
Figure SMS_54
Each column of is the secret data matrix
Figure SMS_57
The projection of the corresponding column vector in this t-dimensional space. left matrix
Figure SMS_60
and the right matrix
Figure SMS_53
It can also be compressed into multiple interference matrices, that is, the interference matrix of m columns × a rows
Figure SMS_56
and an interference matrix of a column by n row
Figure SMS_58
.

使用迭代算法求解左矩阵

Figure SMS_61
和右矩阵
Figure SMS_62
的干扰矩阵，定义干扰度为
Figure SMS_63
，通过如下公式计算：Solve the left matrix using an iterative algorithm
Figure SMS_61
and the right matrix
Figure SMS_62
The interference matrix, define the interference degree as
Figure SMS_63
, calculated by the following formula:

Figure SMS_64
；
Figure SMS_65
；
Figure SMS_64
;
Figure SMS_65
;

其中，

Figure SMS_66
为私密数据矩阵
Figure SMS_67
的i列×
Figure SMS_68
行的子矩阵，使干扰度
Figure SMS_69
最大的
Figure SMS_70
、
Figure SMS_71
，即为干扰矩阵。in,
Figure SMS_66
is the secret data matrix
Figure SMS_67
column i ×
Figure SMS_68
Row submatrix such that the disturbance
Figure SMS_69
biggest
Figure SMS_70
,
Figure SMS_71
, which is the interference matrix.

S4、对干扰矩阵

Figure SMS_72
、
Figure SMS_73
进行数据干扰挖掘，形成干扰属性数据集，用干扰噪声值替换干扰属性数据集的干扰属性值。S4, pair interference matrix
Figure SMS_72
,
Figure SMS_73
Data interference mining is carried out to form the interference attribute data set, and the interference attribute value of the interference attribute data set is replaced with the interference noise value.

为了提高预测准确度和私密安全水平，本发明基于随机干扰方法进行数据干扰挖掘，随机干扰方法由四个处理步骤组成，并将该随机干扰方法应用于数据干扰挖掘的保护阶段，如图2所示，具体包括如下步骤：In order to improve the prediction accuracy and privacy security level, the present invention carries out data interference mining based on the random interference method. The random interference method is composed of four processing steps, and the random interference method is applied to the protection stage of data interference mining, as shown in FIG. 2 , specifically include the following steps:

S41、获取干扰矩阵

Figure SMS_75
、
Figure SMS_77
的数据，用干扰矩阵
Figure SMS_79
中的最后一列乘以干扰矩阵
Figure SMS_76
中的第一行，得到原始干扰数据
Figure SMS_78
，从原始干扰数据
Figure SMS_80
中随机选取k个数据作为干扰属性值，构成干扰属性数据集为
Figure SMS_81
；其中
Figure SMS_74
为干扰属性数据集中的第j个干扰属性值。S41. Obtain the interference matrix
Figure SMS_75
,
Figure SMS_77
data, using the interference matrix
Figure SMS_79
The last column in is multiplied by the interference matrix
Figure SMS_76
In the first row, get the original interference data
Figure SMS_78
, from the raw noise data
Figure SMS_80
Randomly select k data as the interference attribute value, and form the interference attribute data set as
Figure SMS_81
;in
Figure SMS_74
is the jth disturbance attribute value in the disturbance attribute dataset.

S42、定义干扰属性干扰区间。S42. Define an interference attribute interference interval.

对于每个干扰属性值

Figure SMS_82
，定义一个干扰区间，表示为：For each disturbance attribute value
Figure SMS_82
, define an interference interval, expressed as:

Figure SMS_83
；
Figure SMS_83
;

其中

Figure SMS_84
分别为干扰属性干扰区间的下限和上限，分别由下述公式计算：in
Figure SMS_84
are the lower limit and upper limit of the interference interval of the interference attribute, respectively, and are calculated by the following formulas:

Figure SMS_85
；
Figure SMS_85
;

Figure SMS_86
；
Figure SMS_86
;

其中，

Figure SMS_87
为干扰属性数据集为
Figure SMS_88
的归一化值。in,
Figure SMS_87
For the interference attribute data set is
Figure SMS_88
normalized value of .

S43、生成随机干扰数据。S43. Generate random interference data.

基于上述两步，可以得到干扰属性值

Figure SMS_89
的干扰区间为
Figure SMS_90
，由以下公式计算干扰噪声值
Figure SMS_91
：Based on the above two steps, the interference attribute value can be obtained
Figure SMS_89
The interference interval is
Figure SMS_90
, the interference noise value is calculated by the following formula
Figure SMS_91
:

Figure SMS_92
；
Figure SMS_92
;

其中N为随机数。where N is a random number.

S44、将干扰噪声值

Figure SMS_93
替换干扰属性数据集
Figure SMS_94
中的干扰属性值
Figure SMS_95
。S44, will disturb the noise value
Figure SMS_93
Replace the noise attribute dataset
Figure SMS_94
Interference attribute values in
Figure SMS_95
.

随机扰乱技术通过随机化噪声对原始数据进行伪装，可以实现电力数据系统中对用户私密的保护。The random scrambling technology camouflages the original data through random noise, which can realize the protection of user privacy in the power data system.

如图3所示，为本发明的云端电力数据安全防护系统的结构示意图，该安全防护系统包括：电力数据系统、处理器、终端装置、云端处理层。As shown in FIG. 3 , it is a schematic structural diagram of the cloud power data security protection system of the present invention. The security protection system includes: a power data system, a processor, a terminal device, and a cloud processing layer.

在电力数据系统中，处理器对终端装置产生的数据进行处理及存储，将终端装置产生的数据分为私密数据块和普通数据块；In the power data system, the processor processes and stores the data generated by the terminal device, and divides the data generated by the terminal device into private data blocks and ordinary data blocks;

处理器包括标记单元和分割单元，标记单元按照数据属性，对每个数据的结点地址进行标记，分割单元用于将终端装置产生的数据分为私密数据块和普通数据块。The processor includes a marking unit and a splitting unit. The marking unit marks the node address of each data according to the data attribute. The splitting unit is used to divide the data generated by the terminal device into private data blocks and common data blocks.

云端处理层包括级别分类单元、矩阵压缩单元、替换单元和云服务器。The cloud processing layer includes level classification unit, matrix compression unit, replacement unit and cloud server.

级别分类单元将私密数据块进行分类，将不同类别的私密数据块进行不同程度的密钥处理，得到密钥处理后的私密数据矩阵。The level classification unit classifies the private data blocks, performs different key processing on different types of private data blocks, and obtains a secret data matrix after key processing.

在优选实施例中，级别分类单元还包括密钥因子控制装置、密钥因子生成装置和密钥分发装置。密钥因子控制装置，用于生成密钥因子划分参数，每一个密钥因子划分参数对应一个密钥处理级别；接收密钥因子生成装置发送的对应每个密钥处理级别的私钥因子，私钥因子由密钥因子划分参数生成，密钥分发装置将私钥因子赋予给私密数据矩阵的每一行。In a preferred embodiment, the class classification unit further includes key factor control means, key factor generation means and key distribution means. The key factor control device is used to generate key factor division parameters, and each key factor division parameter corresponds to a key processing level; receiving the private key factor corresponding to each key processing level sent by the key factor generation device, the private key factor The key factor is generated by dividing parameters of the key factor, and the key distribution device assigns the private key factor to each row of the secret data matrix.

密钥因子生成装置，接收密钥因子划分参数，并将密钥因子划分参数划分为多个划分子参数，作为私钥因子，将多个私钥因子发送至密钥分发装置；密钥分发装置根据与私密数据矩阵的每一行的序号将多个私钥因子发送给私密数据矩阵的每一行，从而对该行的私密数据进行密钥处理。The key factor generation device receives the key factor division parameter, and divides the key factor division parameter into a plurality of division sub-parameters, and sends the multiple private key factors to the key distribution device as the private key factor; the key distribution device A plurality of private key factors are sent to each row of the private data matrix according to the serial number of each row of the private data matrix, so as to perform key processing on the private data of the row.

矩阵压缩单元对利用矩阵压缩算法对密钥处理后的所述私密数据矩阵进行矩阵压缩，得到干扰矩阵。The matrix compression unit performs matrix compression on the secret data matrix after key processing by using a matrix compression algorithm to obtain an interference matrix.

替换单元，对干扰矩阵进行数据干扰挖掘，形成干扰属性数据集，用干扰噪声值替换干扰属性数据集中的干扰属性值。The replacement unit performs data interference mining on the interference matrix to form an interference attribute data set, and replaces interference attribute values in the interference attribute data set with interference noise values.

在上述实施例中，可以全部或部分地通过软件、硬件、固件或者其任意组合来实现。当使用软件实现时，可以全部或部分地以计算机程序产品的形式实现。所述计算机程序产品包括一个或多个计算机指令。在计算机上加载和执行所述计算机程序指令时，全部或部分地产生按照本申请实施例所述的流程或功能。所述计算机可以是通用计算机、专用计算机、计算机网络、或者其他可编程装置。所述计算机指令可以存储在计算机可读存储介质中，或者通过所述计算机可读存储介质进行传输。所述计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包含一个或多个可用介质集成的服务器、数据中心等数据存储设备。所述可用介质可以是磁性介质，(例如，软盘、硬盘、磁带)、光介质(例如，DVD)、或者半导体介质(例如，固态硬盘(solid state disk，SSD))等。In the above embodiments, all or part of them may be implemented by software, hardware, firmware or any combination thereof. When implemented using software, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on the computer, the processes or functions according to the embodiments of the present application will be generated in whole or in part. The computer can be a general purpose computer, a special purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in or transmitted via a computer-readable storage medium. The computer-readable storage medium may be any available medium that can be accessed by a computer, or a data storage device such as a server or a data center integrated with one or more available media. The available medium may be a magnetic medium (for example, a floppy disk, a hard disk, or a magnetic tape), an optical medium (for example, DVD), or a semiconductor medium (for example, a solid state disk (solid state disk, SSD)) and the like.

以上所述，仅为本申请的具体实施方式，但本申请的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本申请揭露的技术范围内，可轻易想到各种等效的修改或替换，这些修改或替换都应涵盖在本申请的保护范围之内。因此，本申请的保护范围应以权利要求的保护范围为准。The above is only a specific embodiment of the application, but the scope of protection of the application is not limited thereto. Any person familiar with the technical field can easily think of various equivalents within the scope of the technology disclosed in the application. Modifications or replacements, these modifications or replacements shall be covered within the scope of protection of this application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims (8)

1. A cloud electric power data safety protection method is characterized by comprising the following steps:
s1, processing and storing data generated by a terminal device, and dividing the data generated by the terminal device into a private data block and a common data block;
s2, classifying the private data blocks, and performing key processing of different degrees on the private data blocks of different classes to obtain private data matrixes after the key processing;
s3, performing matrix compression on the private data matrix after key processing by using a matrix compression algorithm to obtain an interference matrix;
and S4, carrying out data interference mining on the interference matrix to form an interference attribute data set, and replacing the interference attribute value in the interference attribute data set with the interference noise value.
2. The cloud electric power data security protection method according to claim 1, wherein in step S3:
private data matrix given a set of m columns by n rows
Figure QLYQS_1
Private data matrix
Figure QLYQS_2
Obtaining a left matrix of m columns multiplied by t rows through matrix compression algorithm
Figure QLYQS_3
And t columns by n rows of the right matrix
Figure QLYQS_4
：
Figure QLYQS_5
；
Wherein the private data matrix
Figure QLYQS_6
Maximum compression into left matrix
Figure QLYQS_7
And right matrix
Figure QLYQS_8
Left matrix
Figure QLYQS_9
And right matrix
Figure QLYQS_10
Compressed into a plurality of interference matrices: interference matrix of m columns x a rows
Figure QLYQS_11
And a column x n row interference matrix
Figure QLYQS_12
。
3. The cloud power data security protection method of claim 2,
solving left matrix using iterative algorithm
Figure QLYQS_13
And right matrix
Figure QLYQS_14
The interference matrix of (2) defines the interference degree as
Figure QLYQS_15
Calculated by the following formula:
Figure QLYQS_16
；
Figure QLYQS_17
；
wherein ,
Figure QLYQS_18
as a private data matrix
Figure QLYQS_19
I column of
Figure QLYQS_20
Sub-matrices of rows to make degree of interference
Figure QLYQS_21
Largest size
Figure QLYQS_22
、
Figure QLYQS_23
I.e. an interference matrix.
4. The cloud electric power data security protection method according to claim 3, wherein the step S4 comprises the following steps:
s41, obtaining an interference matrix
Figure QLYQS_25
、
Figure QLYQS_28
Data of (2), using interference matrices
Figure QLYQS_30
The last column in (3) is multiplied by the interference matrix
Figure QLYQS_26
To obtain original interference data
Figure QLYQS_27
From the original interference data
Figure QLYQS_29
Selecting k data as interference attribute values at random to form an interference attribute data set:
Figure QLYQS_31
； wherein
Figure QLYQS_24
Is the jth interference attribute value in the interference attribute data set;
s42, for the interference attribute value
Figure QLYQS_32
Defining an interference interval
Figure QLYQS_33
Expressed as:
Figure QLYQS_34
；
wherein
Figure QLYQS_35
The lower limit and the upper limit of the interference interval are respectively calculated by the following formulas:
Figure QLYQS_36
；
Figure QLYQS_37
；
wherein ,
Figure QLYQS_38
for interference attribute data sets
Figure QLYQS_39
A normalized value of (d);
s43, obtaining an interference attribute value
Figure QLYQS_40
Has an interference interval of
Figure QLYQS_41
Calculating an interference noise value by the following formula
Figure QLYQS_42
：
Figure QLYQS_43
；
Wherein N is a random number;
s44, using interference noise value
Figure QLYQS_44
Replacement of interference attribute datasets
Figure QLYQS_45
Interference attribute value in
Figure QLYQS_46
。
5. The cloud electric power data security protection method according to claim 1, wherein the step S2 specifically includes the following steps:
s21, the private data block is transmitted into a time convolution block, the private data block passes through a time convolution layer in the time convolution block, then the output of the time convolution block is obtained through an activation function by using batch normalization, and the output is transmitted to the next time convolution block;
s22, data output by 3 stacked time volume blocks enter a global average pooling layer and a time recurrent neural network;
s23, serially connecting the output results of the global average pooling layer and the time recursive neural network, and sending the output results to a classification layer for classification to obtain private data blocks of different classes;
and S24, carrying out key processing of different degrees on the private data blocks of different types to obtain a private data matrix after the key processing.
6. A cloud electric power data security protection system for implementing the cloud electric power data security protection method according to any one of claims 1 to 5, comprising: the system comprises a power data system, a processor, a terminal device and a cloud processing layer;
in the electric power data system, the processor processes and stores data generated by the terminal device, and divides the data generated by the terminal device into a private data block and a common data block;
the cloud processing layer comprises a level classification unit, a matrix compression unit, a replacement unit and a cloud server; the level classification unit classifies the private data blocks, and performs different degrees of key processing on the private data blocks of different types to obtain private data matrixes after the key processing; the matrix compression unit performs matrix compression on the private data matrix after the key processing by using a matrix compression algorithm to obtain an interference matrix; and the replacement unit performs data interference mining on the interference matrix to form an interference attribute data set, and replaces the interference attribute value in the interference attribute data set with the interference noise value.
7. The cloud power data security protection system of claim 6, wherein the processor comprises a marking unit and a partitioning unit, the marking unit marks a node address of each data, and the partitioning unit is configured to divide data generated by the terminal device into a private data block and a normal data block according to the marks.
8. The cloud power data security protection system of claim 6, wherein the level classification unit further comprises a key factor control device, a key factor generation device, and a key distribution device; the key factor control device is used for generating key factor dividing parameters; the key factor generation device receives the key factor division parameter and divides the key factor division parameter into a plurality of division subparameters which serve as private key factors; and the key distribution device sends a plurality of private key factors to each row of the private data matrix according to the serial number of each row of the private data matrix, and performs key processing on the private data of each row.
CN202310133859.9A 2023-02-20 2023-02-20 A cloud power data security protection method and protection system Active CN115834257B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310133859.9A CN115834257B (en) 2023-02-20 2023-02-20 A cloud power data security protection method and protection system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310133859.9A CN115834257B (en) 2023-02-20 2023-02-20 A cloud power data security protection method and protection system

Publications (2)

Publication Number Publication Date
CN115834257A true CN115834257A (en) 2023-03-21
CN115834257B CN115834257B (en) 2023-05-19

Family

ID=85521868

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310133859.9A Active CN115834257B (en) 2023-02-20 2023-02-20 A cloud power data security protection method and protection system

Country Status (1)

Country Link
CN (1) CN115834257B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116186018A (en) * 2023-04-25 2023-05-30 国网冀北电力有限公司 Power data identification and analysis method based on safety control

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101401341A (en) * 2005-11-18 2009-04-01 安全第一公司 Secure data parser method and system
US20090136023A1 (en) * 2007-11-26 2009-05-28 National Kaohsiung University Of Applied Sciences Data Encryption Method Using Discrete Fractional Hadamard Transformation
CN111835742A (en) * 2020-07-03 2020-10-27 南京普建维思信息技术有限公司 Data security management system and method based on distributed copy storage
CN111970106A (en) * 2020-08-19 2020-11-20 北京邮电大学 Short ciphertext attribute-based encryption method and system supporting full homomorphism in lattice
CN114630319A (en) * 2022-03-16 2022-06-14 黄文孝 Power transmission and transformation monitoring data safety management system and method for smart power grid

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101401341A (en) * 2005-11-18 2009-04-01 安全第一公司 Secure data parser method and system
CN103384196A (en) * 2005-11-18 2013-11-06 安全第一公司 Secure data parser method and system
US20090136023A1 (en) * 2007-11-26 2009-05-28 National Kaohsiung University Of Applied Sciences Data Encryption Method Using Discrete Fractional Hadamard Transformation
CN111835742A (en) * 2020-07-03 2020-10-27 南京普建维思信息技术有限公司 Data security management system and method based on distributed copy storage
CN111970106A (en) * 2020-08-19 2020-11-20 北京邮电大学 Short ciphertext attribute-based encryption method and system supporting full homomorphism in lattice
CN114630319A (en) * 2022-03-16 2022-06-14 黄文孝 Power transmission and transformation monitoring data safety management system and method for smart power grid

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116186018A (en) * 2023-04-25 2023-05-30 国网冀北电力有限公司 Power data identification and analysis method based on safety control
CN116186018B (en) * 2023-04-25 2023-07-14 国网冀北电力有限公司 Power data identification and analysis method based on safety control

Also Published As

Publication number Publication date
CN115834257B (en) 2023-05-19

Similar Documents

Publication Publication Date Title
JP7441582B2 (en) Methods, devices, computer-readable storage media and programs for detecting data breaches
CN111565205B (en) Network attack identification method and device, computer equipment and storage medium
US20230289661A1 (en) Root cause discovery engine
KR102433425B1 (en) Multi-signal analysis to identify damaged areas
US20210126936A1 (en) Predicting vulnerabilities affecting assets of an enterprise system
US20180262525A1 (en) Multi-modal, multi-disciplinary feature discovery to detect cyber threats in electric power grid
CN113961759B (en) Abnormality detection method based on attribute map representation learning
WO2022142001A1 (en) Target object evaluation method based on multi-score card fusion, and related device therefor
CN110532804A (en) A kind of secure storage control method based on big data
Jeong et al. Anomaly teletraffic intrusion detection systems on hadoop-based platforms: A survey of some problems and solutions
US12301589B2 (en) Intrusion detection using machine learning
CN113746780B (en) Abnormal host detection method, device, medium and equipment based on host image
US20230370486A1 (en) Systems and methods for dynamic vulnerability scoring
CN115834257B (en) A cloud power data security protection method and protection system
US11625438B2 (en) Monitoring information processing systems utilizing co-clustering of strings in different sets of data records
CN116957112A (en) Training method, device, equipment and storage medium of joint model
CN118747186B (en) User equipment login bitmap data storage method, device, electronic device and medium
CN115543991B (en) Data restoration method, device and related equipment based on feature sampling
US12223056B1 (en) Graph-based detection of abusive computational nodes
US20240202370A1 (en) Telemetry data protection for software applications
CN112087448B (en) Security log extraction method, device and computer equipment
US20230029312A1 (en) Similarity-based search for fraud prevention
CN114493850A (en) Artificial intelligence-based online notarization method, system and storage medium
CN113221023B (en) Information pushing method and device
US20230342488A1 (en) Generating and processing personal information chains using machine learning techniques

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant