Risk perception method and system for cross-domain interaction of energy data
By sampling and analyzing energy data and calculating risk characteristics, combined with the BP neural network model, the problem of low risk perception accuracy in cross-domain interaction of energy data is solved, achieving more accurate risk assessment and data security.
Patent Information
- Application Number
- CN202410709179.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-03
- Publication Date
- 2025-09-30
AI Technical Summary
There are problems of data leakage, tampering and semantic mismatch in cross-domain interaction of energy data, which leads to low risk perception accuracy and affects the security and accuracy of data.
By obtaining the original data sequence of energy interaction and the risk comparison data sequence, sampling analysis is performed, a decomposition characteristic risk sequence is constructed, the cross-domain risk index and malicious modification risk characteristic value are calculated, and the risk perception results are obtained using the BP neural network model.
It improves the accuracy of risk perception during cross-domain interaction of energy data, reduces the risk of data leakage and tampering, and ensures data quality and security.
Smart Images
Figure CN120725418A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of data interaction risk assessment, and in particular to a risk perception method and system for cross-domain interaction of energy data. Background Art
[0002] Cross-domain energy data interaction involves the interaction and sharing of massive, heterogeneous data across multiple business systems, departments, and stakeholders. This interaction and sharing is achieved through the normalization and standardization of energy data, enabling different stakeholders to access interactive and shareable energy data using diverse hardware and software to fulfill their data needs. The significance of cross-domain energy data interaction and sharing lies in reducing the cost of energy data collection, improving the efficiency of energy data use, optimizing energy data resource allocation, and meeting the energy data needs of a wider range of stakeholders.
[0003] However, cross-domain interaction of energy data may lead to data security issues such as data leakage, tampering, and loss, thereby affecting the stable operation of the energy system and the accuracy of the data. Furthermore, energy data may contain personal privacy information. If this information is accessed or used by unauthorized third parties, it will infringe on user privacy. Because energy data from different regions, systems, or fields may use different data formats, standards, and semantics, semantic mismatches may occur during data exchange, resulting in the inability to correctly interpret the data. Using traditional logistic binary regression models to perceive the risks of cross-domain interaction of energy data may have defects such as singular matrices, only two output categories, and quality anomalies, resulting in low accuracy in risk perception of cross-domain interaction of energy data. Summary of the Invention
[0004] This application provides a risk perception method and system for cross-domain interaction of energy data to solve the problem of low accuracy of risk perception of cross-domain interaction of source data. The technical solutions adopted are as follows:
[0005] In a first aspect, an embodiment of the present application provides a risk perception method for cross-domain interaction of energy data, the method comprising the following steps:
[0006] Obtaining the energy interaction original data sequence and the risk comparison data sequence, and using both the energy interaction original data sequence and the risk comparison data sequence as the energy data sampling analysis sequence;
[0007] Sampling the energy data sampling and analysis sequence, constructing a decomposition characteristic risk sequence based on the trend characteristics and periodic characteristics of the decomposition of the sampling results of the energy data sampling and analysis sequence; calculating the energy data cross-domain risk index based on the data difference of the decomposition characteristic risk sequence corresponding to the sampling results of the energy data sampling and analysis sequence;
[0008] Obtain a malicious modification risk feature sequence based on the data deviation characteristics between the original energy interaction data sequence and the corresponding risk comparison data sequence; calculate the data malicious modification risk feature value based on the corresponding malicious modification risk feature sequence during the cross-domain interaction of the original energy interaction data; and obtain the data interaction risk increment index between the risk comparison data sequences based on the data malicious modification risk feature value and the energy data cross-domain risk index;
[0009] The energy data interaction risk index is calculated based on the data interaction risk increasing index between the risk comparison data sequences corresponding to the energy interaction original data; and the risk perception result of cross-domain interaction of energy data is obtained based on the energy data interaction risk index corresponding to the energy interaction original data.
[0010] Preferably, the method for obtaining the energy interaction original data sequence and the risk comparison data sequence is:
[0011] Obtain a set of original energy data before interaction, including power supply data, grid data, load data and energy storage data, and take the sequence composed of all data in each type of data in ascending chronological order as an energy interaction original data sequence. Obtain the data of each energy interaction original data sequence after a preset number of interactions in the energy data cross-domain interaction platform, and take the sequence composed of the data obtained after each interaction as a risk comparison data sequence.
[0012] Preferably, the method of sampling the energy data sampling analysis sequence and constructing a decomposition feature risk sequence based on the trend characteristics and period characteristics decomposed from the sampling results of the energy data sampling analysis sequence is:
[0013] For each energy data sampling and analysis sequence, a sampling algorithm is used to obtain a sampling result of each energy data sampling and analysis sequence, and a sequence composed of data sampled each time in the sampling result is used as a sampling sequence;
[0014] The input is each sampling sequence corresponding to the energy data sampling analysis sequence. The STL algorithm is used to obtain the trend component, periodic component and residual component of each sampling sequence. The trend component, periodic component and residual component of each sampling sequence are used as the decomposition characteristic risk sequence of each sampling sequence.
[0015] Preferably, the method for calculating the energy data cross-domain risk index based on the data difference of the decomposed characteristic risk sequence corresponding to the sampling results of the energy data sampling analysis sequence is:
[0016]
[0017] Where H h1 represents the energy data cross-domain risk index of the energy interaction raw data sequence h1 in all energy data sampling analysis sequences; fx,z and f y,z They represent the zth decomposition characteristic risk sequence of the xth and yth sampling sequences of the energy interaction original data sequence h1 in all energy data sampling analysis sequences, respectively, md(f x,z ,f y,z ) represents f x,z and f y,z The Manhattan distance between x and α y They respectively represent the range of elements in the x-th and y-th sampling sequences of the energy interaction original data sequence h1 in all energy data sampling analysis sequences; n represents the number of sampling sequences of the energy interaction original data sequence h1 in all energy data sampling analysis sequences; w represents the number of decomposition characteristic risk sequences of each sampling sequence of the energy interaction original data sequence h1 in all energy data sampling analysis sequences.
[0018] Preferably, the method for obtaining the malicious modification risk feature sequence based on the data deviation feature between the energy interaction original data sequence and the corresponding risk comparison data sequence is:
[0019] For each risk comparison data sequence corresponding to the energy interaction original data sequence, calculate the absolute value of the difference between the energy interaction original data sequence and the elements at the same position in each risk comparison data sequence, and use the ratio of the absolute value to the elements at the same position in the energy interaction original data sequence as the malicious modification risk index of the elements at the same position in each risk comparison data sequence. Use the elements in each risk comparison data sequence whose malicious modification risk index is greater than a preset threshold as malicious modification risk elements. Retain the value of the malicious modification risk element in each risk comparison data sequence, set the value of the non-malicious modification risk element in each risk comparison data sequence to 0, and use the updated result of each risk comparison data sequence as the malicious modification risk feature sequence of the data sequence after each interaction.
[0020] Preferably, the method for calculating the data malicious modification risk characteristic value according to the malicious modification risk characteristic sequence corresponding to the cross-domain interaction process of the energy interaction original data is:
[0021]
[0022] Where, T b,b-1 The risk characteristic value of the data malicious modification of the risk comparison data sequence after the b-th and b-1-th interactions of the original energy interaction data sequence h1; L b,v Represents the vth element in the risk feature sequence of malicious modification of the risk comparison data sequence after the bth interaction of the original energy interaction data sequence h1, L b-1,uRepresents the u-th element in the malicious modification risk feature sequence of the risk comparison data sequence after the b-1th interaction of the energy interaction original data sequence h1; ω b and ω b-1 They respectively represent the number of malicious modification risk elements in the risk comparison data sequence after the b-th and b-1-th interactions of the energy interaction original data sequence h1; r represents the number of elements in the malicious modification risk feature sequence of the risk comparison data sequence after each interaction of the energy interaction original data sequence h1.
[0023] Preferably, the method for obtaining the data interaction risk increasing index between risk comparison data sequences based on the data malicious modification risk characteristic value and the energy data cross-domain risk index is:
[0024] For two risk comparison data sequences with adjacent interaction times corresponding to the original energy interaction data sequences, the sum of the energy data cross-domain risk indices corresponding to the two risk comparison data sequences is used as the numerator, the product of the energy data cross-domain risk index corresponding to the original energy interaction data sequence and 2 is used as the denominator, and the ratio of the numerator to the denominator and the product of the data malicious modification risk characteristic value between the two risk comparison data sequences is used as the data interaction risk increasing index of the two risk comparison data sequences.
[0025] Preferably, the specific method for calculating the energy data interaction risk index based on the data interaction risk increasing index between the risk comparison data sequences corresponding to the energy interaction original data is:
[0026]
[0027] Where, represents the energy data interaction risk index of the energy interaction raw data sequence h1; R b,b-1 The data interaction risk increasing index of the energy interaction original data sequence h1 after the b-th and b-1-th interactions; l 1,b and l 1,b-1 They represent the risk comparison data sequences after the b-th and b-1-th interactions of the original energy interaction data sequence h1, edr(l 1,b ,l 1,b-1 ) means l 1,b and l 1,b-1 The EDR edit distance between them; a represents the number of risk comparison data sequences after interaction corresponding to the original energy interaction data sequence h1.
[0028] Preferably, the method for obtaining the risk perception result of cross-domain interaction of energy data according to the energy data interaction risk index corresponding to the original energy interaction data is:
[0029] The set consisting of all energy interaction original data sequences corresponding to each set of original energy data before interaction is regarded as an energy interaction original data set, and the sequence consisting of energy data interaction risk indexes corresponding to all sequences in each energy interaction original data set is regarded as the energy data interaction risk index sequence corresponding to each energy interaction original data set;
[0030] The energy data interaction risk index sequence corresponding to the first preset number of energy interaction original data sets obtained on the energy data cross-domain interaction platform is used as the training set of the BP neural network model, and the energy data interaction risk index sequence corresponding to the second preset number of energy interaction original data sets obtained on the energy data cross-domain interaction platform is used as the test set of the BP neural network model. The BP neural network model is used to obtain the risk perception results of energy data cross-domain interaction.
[0031] In a second aspect, an embodiment of the present application also provides a risk perception system for cross-domain interaction of energy data, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above methods when executing the computer program.
[0032] The beneficial effects of the present application are: by analyzing the differences in the characteristics of trend components, periodic components and residual components of local sampling data of energy data, the energy data cross-domain risk index of the original energy interaction data sequence and the data sequence after each interaction in the process of cross-domain interaction of energy data is obtained, and the data sensitive change characteristics and distinguishing characteristics are reflected through the energy data cross-domain risk index, thereby expressing the possibility of risks in the energy data during cross-domain interaction; further, by analyzing the differences in the energy data cross-domain risk index between the data after each interaction and the original energy interaction data and the trend characteristics of the risks of energy data during the interaction, the data interaction risk increasing index is obtained, and the energy data interaction risk index is calculated according to the data interaction risk increasing index. The degree of risk of cross-domain interaction of energy data is reflected based on the energy data interaction risk index. Its beneficial effect is that it takes into account the differences in the sensitivity and distinguishing characteristics of the original data before the energy data interaction and the characteristics of the energy data after the interaction, as well as the risk change trend of the data quality deviation after the interaction, to more accurately reflect the perception of risks in the process of cross-domain interaction of energy data, thereby effectively reducing the risks of cross-domain interaction of energy data. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.
[0034] Figure 1 A flowchart of a risk perception method for cross-domain interaction of energy data provided by one embodiment of the present application;
[0035] Figure 2 A schematic diagram of an implementation process for obtaining risk perception results of cross-domain interaction of energy data provided by an embodiment of the present application. DETAILED DESCRIPTION
[0036] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.
[0037] See also Figure 1 , which shows a flow chart of a risk perception method for cross-domain interaction of energy data provided by an embodiment of the present application, the method comprising the following steps:
[0038] Step S001: Obtain energy interaction original data sequence and risk comparison data sequence.
[0039] A set of original energy data before interaction is obtained on the energy data cross-domain interaction platform, including power supply data, grid data, load data and energy storage data, where the number of each type of data obtained is 5000, and the sequence of all data in each type of data in ascending time order is used as the energy interaction original data sequence, which is recorded as h1, h2, h3 and h4 respectively; further, the data sequence after a (the size is an empirical value of 500) interactions of power supply data, grid data, load data and energy storage data in the energy data cross-domain interaction platform is obtained respectively, and the data sequence after each interaction is recorded as l 1,b 、l 2,b 、l 3,b and l 4,b , where l 1,b 、l 2,b 、l 3,b and l 4,b They represent the data sequences after the b-th interaction of power supply data, grid data, load data and energy storage data in the energy data cross-domain interaction platform, and the data sequence obtained after each interaction is used as a risk comparison data sequence.
[0040] Since the acquired data may contain missing values and outliers, the input is the original data before the interaction and the data of a interactions. The data cleaning algorithm is used to remove duplication, process missing values and process outliers on the input data. The specific processing process of the data cleaning algorithm is a well-known technology and will not be repeated here.
[0041] At this point, the preprocessed energy interaction raw data sequence and a risk comparison data sequence are obtained.
[0042] Step S002: Sample the energy interaction original data sequence and the risk comparison data sequence respectively to obtain sampling sequences, and obtain the energy data cross-domain risk index based on the differences in trend components, periodic components, and residual components between different sampling sequences.
[0043] Nowadays, cross-domain interaction and sharing of energy data has become a development trend. The establishment and application of cross-domain interaction platforms for energy data have made great progress. However, cross-domain interaction and sharing of energy data require multiple subjects to upload energy data to a third-party platform to meet the needs of different subjects for energy data. In the process of uploading energy data and interacting with energy data between different subjects, there may be risks such as data leakage and malicious modification of data, which reduces the security and accuracy of cross-domain interaction of energy data. Therefore, it is necessary to conduct a risk assessment on the process of cross-domain interaction of energy data, and to perform security protection on the cross-domain interaction of energy data based on the risk assessment results, thereby reducing the risk of privacy data leakage in energy data and improving the quality of data for cross-domain interaction of energy data.
[0044] Furthermore, since the risks of cross-domain interaction of energy data can be considered from two aspects, namely the data characteristics of energy data for privacy protection and the characteristics of data deviation during the interaction of energy data, if the privacy protection characteristics of energy data are obvious and the characteristics of data deviation during the cross-domain interaction of energy data are significant, it means that there is a high possibility of risks in the cross-domain interaction of energy data, and there is a risk of data leakage and data quality degradation.
[0045] Furthermore, the data characteristics of the privacy protection of the original data before the cross-domain interaction of energy data are analyzed based on the data sensitive change characteristics in the energy interaction original data sequence, that is, the more significant the data sensitive change characteristics are, the more obvious the distinguishing characteristics of the energy interaction original data are, and the greater the possibility of data leakage. Specifically, the energy interaction original data sequence and the risk comparison data sequence are both used as energy data sampling analysis sequences, and each energy data sampling analysis sequence is input. The reservoir sampling algorithm is used to obtain the sampling results of each energy data sampling analysis sequence, where the number of samples is 500 and the number of sampling times is n=50. The specific implementation process of the reservoir sampling algorithm is a well-known technology and will not be repeated here. Take the energy interaction original data sequence h1 in all energy data sampling analysis sequences as an example for analysis, and take the sequence composed of all data in each sampling result corresponding to h1 in ascending time order as a sampling sequence. Therefore, all sampling sequences corresponding to h1 are [f1,f2,…,f n ], where f n Indicates the sampling sequence of h1 corresponding to every n sampling results.
[0046] Furthermore, the input is each sampling sequence corresponding to h1, and the STL (Seasonal and Trend decomposition using Loess) algorithm is used to obtain the trend component, period component and residual component of each sampling sequence corresponding to h1. The trend component, period component and residual component of each sampling sequence are used as the decomposition characteristic risk sequence of each sampling sequence, where the decomposition characteristic risk sequence of the nth sampling sequence is denoted as f n,1 、f n,2 and f n,3 , and f n,1 、f n,2 and f n,3 The specific calculation process of the STL algorithm corresponds to the trend component, periodic component and residual component respectively. It is a well-known technology and will not be described in detail here. The energy data cross-domain risk index of h1 is obtained according to the trend component, periodic component and residual component of each sampling sequence corresponding to h1. The specific calculation formula is as follows:
[0047]
[0048] Where H h1 represents the energy data cross-domain risk index of the energy interaction raw data sequence h1 in all energy data sampling analysis sequences; f x,z and f y,z They represent the zth decomposition characteristic risk sequence of the xth and yth sampling sequences of the energy interaction original data sequence h1 in all energy data sampling analysis sequences, respectively, md(f x,z ,f y,z ) represents fx,z and f y,z The specific calculation process of Manhattan distance is a well-known technology and will not be described in detail; α x and α y They respectively represent the range of elements in the x-th and y-th sampling sequences of the energy interaction original data sequence h1 in all energy data sampling analysis sequences; n represents the number of sampling sequences of the energy interaction original data sequence h1 in all energy data sampling analysis sequences; w represents the number of decomposition characteristic risk sequences of each sampling sequence of the energy interaction original data sequence h1 in all energy data sampling analysis sequences.
[0049] If the numerical variation range of different sampled data in the energy interaction original data sequence h1 in all energy data sampling analysis sequences is small, then the calculated |α x -α y The smaller the value of |, and the greater the differences in trend components, period components, and residual components between different sampling data in h1, the more likely the calculated md(f x,z ,f y,z ) is larger, that is, the calculated The larger the value is, the greater the energy data cross-domain risk index of the energy interaction raw data sequence h1 The larger the value of , the greater the differences in the trend change characteristics, periodic change characteristics, and residual data change characteristics of the sampled data within the same numerical change range in h1. The data-sensitive change characteristics and distinguishing characteristics of the energy interaction original data sequence h1 are significant, and the risks in cross-domain interaction are greater.
[0050] At this point, the cross-domain risk index of energy data has been obtained.
[0051] Step S003 , calculating the energy data interaction risk index based on the data deviation feature analysis results between the original data and the interaction data and the energy data cross-domain risk index.
[0052] During cross-domain energy data interaction, the likelihood of malicious tampering increases with increasing frequency of interaction. Because the cross-domain energy data interaction platform is jointly managed by multiple entities, the quality of energy data gradually decreases with increasing interaction frequency. This means that the scope for malicious tampering with energy data increases through multiple interactions, leading to a gradual decline in energy data quality and an increased risk of data leakage. Therefore, the energy data interaction risk index is calculated by analyzing the difference in the energy data cross-domain risk index and data deviation characteristics between the data after each interaction and the original energy interaction data.
[0053] Specifically, taking the energy interaction original data sequence h1 and the corresponding a-th interaction data as an example, for the risk comparison data sequence l after the b-th interaction of the energy interaction original data sequence h1 1,b , calculate h1 and l 1,b The absolute value of the difference between the elements at the same position in h1 and the ratio of the elements at the same position in h2 is taken as l 1,b The malicious modification risk index of the same position element in l 1,b Elements with a malicious modification risk index greater than 0.1 are considered malicious modification risk elements, and l 1,b The method for obtaining malicious modification risk elements is the same. The malicious modification risk elements in the risk comparison data sequence after each interaction of the energy interaction original data sequence h1 can be obtained respectively, and the values of the malicious modification risk elements in the risk comparison data sequence after each interaction of the energy interaction original data sequence h1 are retained, and the values of the non-malicious modification risk elements in the risk comparison data sequence after each interaction of the energy interaction original data sequence h1 are set to 0. The updated results of each risk comparison data sequence are used as the malicious modification risk feature sequence of the risk comparison data sequence after each interaction. It should be noted that when there is no deviation in the energy interaction original data sequence during the cross-domain interaction process, the elements and the number of elements of the energy interaction original data sequence and each risk comparison data sequence are the same.
[0054] The data interaction risk increasing index is obtained based on the difference between the original energy interaction data sequence h1 and the risk comparison data sequence after each interaction, and the malicious modification risk feature sequence of the risk comparison data sequence after each interaction. The specific calculation formula is as follows:
[0055]
[0056]
[0057] Where, T b,b-1 The risk characteristic value of the data malicious modification of the risk comparison data sequence after the b-th and b-1-th interactions of the original energy interaction data sequence h1; L b,v Represents the vth element in the risk feature sequence of malicious modification of the risk comparison data sequence after the bth interaction of the original energy interaction data sequence h1, L b-1,u Represents the u-th element in the malicious modification risk feature sequence of the risk comparison data sequence after the b-1th interaction of the energy interaction original data sequence h1; ω b and ω b-1 They represent the number of malicious modification risk elements in the risk comparison data sequence after the b-th and b-1-th interactions of the energy interaction original data sequence h1 respectively; r represents the number of elements in the malicious modification risk feature sequence of the risk comparison data sequence after each interaction of the energy interaction original data sequence h1;
[0058] R b,b-1 The data interaction risk increasing index represents the risk comparison data sequence after the b-th and b-1-th interactions of the energy interaction original data sequence h1; The energy data cross-domain risk index corresponding to the energy interaction raw data sequence h1, and They respectively represent the energy data cross-domain risk index corresponding to the risk comparison data sequences after the b-th and b-1-th interactions of the energy interaction original data sequence h1.
[0059] If the position difference and value difference of the risk comparison data sequence after the adjacent b-th and b-1-th interactions in the energy interaction original data sequence h1 are large, then the calculated At the same time, the number of malicious modification risk elements in the risk comparison data sequence after the bth and b-1th interactions is quite different, then |ω b -ω b-1 The larger the value of |, the risk characteristic value T of the malicious modification of the data of the energy interaction original data sequence h1 after the b-th and b-1 interactions is calculated. b,b-1 The bigger.
[0060] Furthermore, the energy data cross-domain risk index of the risk comparison data sequence after the b-th and b-1-th interactions of the energy interaction original data sequence h1 shows an obvious growth trend compared with the energy data cross-domain risk index of the energy interaction original data sequence h1. The larger the value is, the risk characteristic value T of the malicious modification of the data sequence after the b-th and b-1-th interactions of the energy interaction original data sequence h1 is. b,b-1 The larger the value is, the risk of the energy interaction original data sequence h1 after the b-th and b-1-th interactions is compared with the data interaction risk increasing index R of the data sequence. b,b-1 The larger the value is, the greater the possibility of risk in the data after the interaction of the energy interaction original data sequence h1 increases with the increase in the number of interactions.
[0061] Furthermore, the energy data interaction risk index is calculated based on the data interaction risk increasing index of the risk comparison data sequence after the b-th interaction and the b-1-th interaction of the energy interaction original data sequence h1. The specific calculation formula is as follows:
[0062]
[0063] Where, represents the energy data interaction risk index of the energy interaction raw data sequence h1; R b,b-1The data interaction risk increasing index of the energy interaction original data sequence h1 after the b-th and b-1-th interactions; l 1,b and l 1,b-1 They represent the risk comparison data sequences after the b-th and b-1-th interactions of the original energy interaction data sequence h1, edr(l 1,b ,l 1,b-1 ) means l 1,b and l 1,b-1 The EDR edit distance between them, the specific calculation process of EDR (Edit Distance on Real Sequence) edit distance is a well-known technology and will not be repeated here; a represents the number of risk comparison data sequences after the interaction of the energy interaction original data sequence h1.
[0064] If the risk after the adjacent b-th and b-1-th interactions in the energy interaction original data sequence h1 is compared with the data interaction risk increasing index R of the data sequence b,b-1 The value of is large, and the numerical deviation characteristics of the risk comparison data sequence after the b-th and b-1-th interactions are large, then The larger the value is, the greater the energy data interaction risk index of the calculated energy interaction original data sequence h1 is. It is larger, indicating that the risk of malicious data tampering and data leakage in cross-domain interaction of the energy interaction original data sequence h1 is higher.
[0065] At this point, the energy data interaction risk index is obtained.
[0066] Step S004: obtaining a risk perception result of cross-domain interaction of energy data based on an energy data interaction risk index corresponding to cross-domain interaction of the original energy interaction data.
[0067] According to step S003, the energy data interaction risk index of the energy interaction original data sequence h1 can be obtained. Furthermore, the energy data interaction risk index corresponding to the energy interaction original data sequences h2, h3 and h4 can be obtained: and Take h1, h2, h3 and h4 as an energy interaction raw data set, and each energy interaction raw data set corresponds to an energy data interaction risk index sequence. The energy data interaction risk index sequence corresponding to the p (the size is taken as the empirical value 500) group of energy interaction original data sets obtained on the energy data cross-domain interaction platform is used as the training set of the BP (Back Propagation) neural network model, and the energy data interaction risk index sequence corresponding to the s (the size is taken as the empirical value 100) group of energy interaction original data sets obtained on the energy data cross-domain interaction platform is used as the test set of the BP neural network model. The specific training process of the BP neural network model is a well-known technology and will not be repeated here.
[0068] Furthermore, the input is an energy data interaction risk index sequence corresponding to an obtained energy interaction original data set, and the trained BP neural network model is used to obtain the risk perception results of energy data cross-domain interaction. The risk perception results include high energy data cross-domain interaction risk, medium energy data cross-domain interaction risk, and low energy data cross-domain interaction risk. The specific implementation process of obtaining the risk perception results of energy data cross-domain interaction is as follows: Figure 2 shown.
[0069] Based on the same inventive concept as the above method, an embodiment of the present application also provides a risk perception system for cross-domain interaction of energy data, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of any one of the above-mentioned risk perception methods for cross-domain interaction of energy data are implemented.
[0070] The various embodiments in this specification are described in a progressive manner. Similar parts between the various embodiments can be referenced to each other. Each embodiment focuses on the differences from other embodiments. The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent replacements, improvements, etc. made within the principles of this application shall be included in the scope of protection of this application.
Claims
1. A risk perception method for cross-domain interaction of energy data, characterized in that: The method comprises the following steps: Obtaining the energy interaction original data sequence and the risk comparison data sequence, and using both the energy interaction original data sequence and the risk comparison data sequence as the energy data sampling analysis sequence; Sampling the energy data sampling and analysis sequence, constructing a decomposition characteristic risk sequence based on the trend characteristics and periodic characteristics of the decomposition of the sampling results of the energy data sampling and analysis sequence; calculating the energy data cross-domain risk index based on the data difference of the decomposition characteristic risk sequence corresponding to the sampling results of the energy data sampling and analysis sequence; Obtain a malicious modification risk feature sequence based on the data deviation characteristics between the original energy interaction data sequence and the corresponding risk comparison data sequence; calculate the data malicious modification risk feature value based on the corresponding malicious modification risk feature sequence during the cross-domain interaction of the original energy interaction data; and obtain the data interaction risk increment index between the risk comparison data sequences based on the data malicious modification risk feature value and the energy data cross-domain risk index; The energy data interaction risk index is calculated based on the data interaction risk increasing index between the risk comparison data sequences corresponding to the energy interaction original data; and the risk perception result of cross-domain interaction of energy data is obtained based on the energy data interaction risk index corresponding to the energy interaction original data.
2. The risk perception method for cross-domain interaction of energy data according to claim 1 is characterized in that: The method for obtaining the energy interaction original data sequence and the risk comparison data sequence is: Obtain a set of original energy data before interaction, including power supply data, grid data, load data and energy storage data, and take the sequence composed of all data in each type of data in ascending chronological order as an energy interaction original data sequence. Obtain the data of each energy interaction original data sequence after a preset number of interactions in the energy data cross-domain interaction platform, and take the sequence composed of the data obtained after each interaction as a risk comparison data sequence.
3. The risk perception method for cross-domain interaction of energy data according to claim 1 is characterized in that: The method for sampling the energy data sampling analysis sequence and constructing a decomposition characteristic risk sequence based on the trend characteristics and period characteristics decomposed from the sampling results of the energy data sampling analysis sequence is: For each energy data sampling and analysis sequence, a sampling algorithm is used to obtain a sampling result of each energy data sampling and analysis sequence, and a sequence composed of data sampled each time in the sampling result is used as a sampling sequence; The input is each sampling sequence corresponding to the energy data sampling analysis sequence. The STL algorithm is used to obtain the trend component, periodic component and residual component of each sampling sequence. The trend component, periodic component and residual component of each sampling sequence are used as the decomposition characteristic risk sequence of each sampling sequence.
4. The risk perception method for cross-domain interaction of energy data according to claim 1 is characterized in that: The method for calculating the energy data cross-domain risk index based on the data difference of the decomposed characteristic risk sequence corresponding to the sampling results of the energy data sampling analysis sequence is: Where H h1 represents the energy data cross-domain risk index of the energy interaction raw data sequence h1 in all energy data sampling analysis sequences; f x,z and f y,z They represent the zth decomposition characteristic risk sequence of the xth and yth sampling sequences of the energy interaction original data sequence h1 in all energy data sampling analysis sequences, respectively, md(f x,z ,f y,z ) represents f x,z and f y,z The Manhattan distance between x and α y They respectively represent the range of elements in the x-th and y-th sampling sequences of the energy interaction original data sequence h1 in all energy data sampling analysis sequences; n represents the number of sampling sequences of the energy interaction original data sequence h1 in all energy data sampling analysis sequences; w represents the number of decomposition characteristic risk sequences of each sampling sequence of the energy interaction original data sequence h1 in all energy data sampling analysis sequences.
5. The risk perception method for cross-domain interaction of energy data according to claim 1 is characterized in that: The method for obtaining a malicious modification risk feature sequence based on the data deviation characteristics between the energy interaction original data sequence and the corresponding risk comparison data sequence is: For each risk comparison data sequence corresponding to the energy interaction original data sequence, calculate the absolute value of the difference between the energy interaction original data sequence and the elements at the same position in each risk comparison data sequence, and use the ratio of the absolute value to the elements at the same position in the energy interaction original data sequence as the malicious modification risk index of the elements at the same position in each risk comparison data sequence. Use the elements in each risk comparison data sequence whose malicious modification risk index is greater than a preset threshold as malicious modification risk elements. Retain the value of the malicious modification risk element in each risk comparison data sequence, set the value of the non-malicious modification risk element in each risk comparison data sequence to 0, and use the updated result of each risk comparison data sequence as the malicious modification risk feature sequence of the data sequence after each interaction.
6. The risk perception method for cross-domain interaction of energy data according to claim 1 is characterized in that: The method for calculating the data malicious modification risk characteristic value according to the malicious modification risk characteristic sequence corresponding to the cross-domain interaction process of the energy interaction original data is: Where, T b,b-1 The risk characteristic value of the data malicious modification of the risk comparison data sequence after the b-th and b-1-th interactions of the original energy interaction data sequence h1; L b,v Represents the vth element in the risk feature sequence of malicious modification of the risk comparison data sequence after the bth interaction of the original energy interaction data sequence h1, L b-1,u Represents the u-th element in the malicious modification risk feature sequence of the risk comparison data sequence after the b-1th interaction of the energy interaction original data sequence h1; ω b and ω b-1 They respectively represent the number of malicious modification risk elements in the risk comparison data sequence after the b-th and b-1-th interactions of the energy interaction original data sequence h1; r represents the number of elements in the malicious modification risk feature sequence of the risk comparison data sequence after each interaction of the energy interaction original data sequence h1.
7. The risk perception method for cross-domain interaction of energy data according to claim 1 is characterized in that: The method for obtaining the data interaction risk increasing index between risk comparison data sequences based on the data malicious modification risk characteristic value and the energy data cross-domain risk index is: For two risk comparison data sequences with adjacent interaction times corresponding to the original energy interaction data sequences, the sum of the energy data cross-domain risk indices corresponding to the two risk comparison data sequences is used as the numerator, the product of the energy data cross-domain risk index corresponding to the original energy interaction data sequence and 2 is used as the denominator, and the ratio of the numerator to the denominator and the product of the data malicious modification risk characteristic value between the two risk comparison data sequences is used as the data interaction risk increasing index of the two risk comparison data sequences.
8. The risk perception method for cross-domain interaction of energy data according to claim 1 is characterized in that: The specific method for calculating the energy data interaction risk index based on the data interaction risk increasing index between the risk comparison data sequences corresponding to the energy interaction original data is: Where K h1 represents the energy data interaction risk index of the energy interaction raw data sequence h1; R b,b-1 The data interaction risk increasing index of the energy interaction original data sequence h1 after the b-th and b-1-th interactions; l 1,b and l 1,b-1 They represent the risk comparison data sequences after the b-th and b-1-th interactions of the original energy interaction data sequence h1, edr(l 1,b ,l 1,b-1 ) means l 1,b and l 1,b-1 The EDR edit distance between them; a represents the number of risk comparison data sequences after interaction corresponding to the original energy interaction data sequence h1.
9. The risk perception method for cross-domain interaction of energy data according to claim 1 is characterized in that: The method for obtaining the risk perception result of cross-domain interaction of energy data based on the energy data interaction risk index corresponding to the original energy interaction data is as follows: The set consisting of all energy interaction original data sequences corresponding to each set of original energy data before interaction is regarded as an energy interaction original data set, and the sequence consisting of energy data interaction risk indexes corresponding to all sequences in each energy interaction original data set is regarded as the energy data interaction risk index sequence corresponding to each energy interaction original data set; The energy data interaction risk index sequence corresponding to the first preset number of energy interaction original data sets obtained on the energy data cross-domain interaction platform is used as the training set of the BP neural network model, and the energy data interaction risk index sequence corresponding to the second preset number of energy interaction original data sets obtained on the energy data cross-domain interaction platform is used as the test set of the BP neural network model. The BP neural network model is used to obtain the risk perception results of energy data cross-domain interaction.
10. A risk perception system for cross-domain interaction of energy data, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the risk perception method for cross-domain interaction of energy data as described in any one of claims 1 to 9 are implemented.