Data encryption method for new energy power plant

Abstract

The invention discloses a data encryption method for a new energy power plant, and relates to the technical field of data encryption, and the data encryption method comprises the following steps: obtaining a real-time operation data set of the new energy power plant; identifying sensitive data; obtaining sensitive data of different grades; and respectively storing the encrypted sensitive data and non-sensitive data of different levels in a preset security database, and establishing a data access authority management mechanism. According to the method, through sensitive classification and hierarchical protection of the real-time operation data of the new energy power plant, encryption algorithms of different intensities can be adopted for data of different levels, the problems of resource waste and insufficient sensitive data protection under the unified encryption intensity are effectively avoided, the identified sensitive data are encrypted, and the security of the new energy power plant is improved. Risks caused by hostile attacks or accidental leakage are effectively prevented, and confidentiality of core operation data of a power plant is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of data encryption, and in particular to a data encryption method for a new energy power plant. Background Art

[0002] With the rapid development of information technology and the continuous progress of the power industry, data management and protection of power plants have become increasingly important. During the operation of renewable energy power plants, a large amount of data is generated, which contains important operational information and sensitive data. However, these data may face security threats from both internal and external sources during transmission. Therefore, improving the encryption security of data in renewable energy power plants is crucial to ensure the safe operation of the power system.

[0003] However, the existing technology is not convenient for sensitivity classification and hierarchical protection of new energy power plant operating data, is not convenient for using encryption algorithms of different strengths according to different levels of sensitive data, is not convenient for implementing hierarchical access to data of different levels, and cannot ensure that only authorized users can access sensitive data, thereby reducing data security.

[0004] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0005] In view of the problems in the related art, the present invention proposes a data encryption method for a new energy power plant to overcome the above technical problems existing in the existing related art.

[0006] To this end, the specific technical solution adopted by the present invention is as follows:

[0007] A data encryption method for a new energy power plant, the data encryption method comprising the following steps:

[0008] S1. Acquire the real-time operation data of the new energy power plant, and pre-process the real-time operation data to obtain the real-time operation data set of the new energy power plant;

[0009] S2. Based on the data classification algorithm, the real-time operation data set of the new energy power plant is classified to identify sensitive data;

[0010] S3. Establish a sensitivity level model, and classify the identified sensitive data by using the sensitivity level model to obtain sensitive data of different levels;

[0011] S4. Based on the encryption algorithm, the obtained sensitive data of different levels are encrypted, and the encrypted sensitive data and non-sensitive data of different levels are stored in a preset security database respectively, and a data access permission management mechanism is established.

[0012] Furthermore, obtaining the real-time operation data of the new energy power plant and preprocessing the real-time operation data to obtain the real-time operation data set of the new energy power plant includes the following steps:

[0013] S11, performing wavelet decomposition on the acquired original operation data of the new energy power plant to separate the high-frequency noise signal, reconstructing the decomposed signal according to the energy size of the frequency band, and obtaining the real-time operation data after denoising;

[0014] S12, using the empirical mode decomposition technology to decompose the denoised real-time operation data, extract each mode component, and perform smoothing processing;

[0015] S13, calculating the correlation between each mode component and the original real-time operation data, and eliminating components with low correlation;

[0016] S14, reconstructing the optimized mode components into a complete data set to form a real-time operation data set of the new energy power plant.

[0017] Furthermore, based on the data classification algorithm, the real-time operation data set of the new energy power plant is classified, and the identification of sensitive data includes the following steps:

[0018] S21. Acquire a real-time operation data set of a new energy power plant, and perform normalization processing on the real-time operation data set;

[0019] S22. Assign corresponding weights to different data categories according to the importance of the real-time running data set in the classification process;

[0020] S23. Defining classification standards for sensitive data and classification standards for non-sensitive data through weighted data values, and forming ideal classification targets for sensitive data and non-sensitive data respectively;

[0021] S24. Calculate the grey correlation coefficient between each real-time operation data and the sensitive data classification standard and the non-sensitive data classification standard, and quantify the correlation degree between the real-time operation data and the ideal classification target of sensitive data and the ideal classification target of non-sensitive data respectively;

[0022] S25. Based on the grey correlation coefficient, the grey correlation degree between each real-time operation data and the sensitive data classification standard and the non-sensitive data classification standard is calculated respectively;

[0023] S26. Use grey correlation to construct the classification proximity of sensitive data, identify the sensitive data in the real-time running data set, and identify and label the sensitive data.

[0024] Furthermore, the formulas for calculating the grey correlation between each real-time operation data and the sensitive data classification standard and the non-sensitive data classification standard are as follows:

[0025]

[0026] In the formula, H i ′ represents the grey correlation between the i-th real-time operation data and sensitive data;

[0027] H i ″ represents the grey correlation between the ith real-time operation data and non-sensitive data;

[0028] H represents the grey relational degree;

[0029] F represents the total number of real-time running data;

[0030] j represents real-time running data.

[0031] Furthermore, using grey correlation, constructing the classification closeness of sensitive data, identifying the sensitive data in the real-time running data set, and identifying and labeling the sensitive data include the following steps:

[0032] S261. Compare each real-time operation data in the real-time operation data set with the sensitive data classification standard and the non-sensitive data classification standard, and calculate the grey correlation degree between the two;

[0033] S262. Based on the grey correlation of sensitive data and non-sensitive data, a classification proximity model is constructed as the basis for classification;

[0034] S263. According to the classification proximity model, a classification threshold is set, and the proximity of the real-time operation data is compared with the classification threshold. If the proximity of the real-time operation data is greater than the classification threshold, it is determined to be sensitive data; otherwise, it is determined to be non-sensitive data;

[0035] S264. Label the real-time operation data determined to be sensitive data to identify the sensitive data.

[0036] Furthermore, the formula for constructing the classification proximity model is:

[0037]

[0038] Where M i ′ represents the relative closeness between the i-th real-time operation data and the sensitive data;

[0039] M represents relative closeness;

[0040] H i ′ represents the grey correlation between the i-th real-time operation data and sensitive data;

[0041] H i ″ represents the grey correlation between the ith real-time operation data and non-sensitive data;

[0042] H represents the grey relational degree;

[0043] F represents the total number of real-time running data.

[0044] Furthermore, a sensitivity level model is established, and the identified sensitive data is classified into different levels through the sensitivity level model. The steps of obtaining sensitive data of different levels include:

[0045] S31. Compare the identified sensitive data with the preset sensitivity performance indicators, screen out the sensitivity characteristics of the sensitive data, and establish a sensitivity level model;

[0046] S32. Analyze the key factors that affect the sensitivity of sensitive data and set a corresponding sensitivity evaluation level for each key factor;

[0047] S33. Using the sensitivity level model, the screened sensitive data is scored for sensitivity, and each sensitive data is preliminarily classified into different sensitivity levels according to the scoring results;

[0048] S34. Combine the scoring results and the set sensitivity evaluation level to calculate the comprehensive sensitivity value of each data and generate a standardized sensitivity scoring matrix;

[0049] S35. Based on the results of the sensitivity scoring matrix, determine whether the sensitivity level of each sensitive data exceeds the preset threshold. If it exceeds, it is marked as key sensitive data. If it does not exceed, it is classified as medium or low sensitive data.

[0050] Furthermore, the screened sensitive data is scored for sensitivity through the sensitivity level model, and each sensitive data is preliminarily divided into different sensitivity levels according to the scoring results, including the following steps:

[0051] S331, dividing the screened sensitive data into a training set and a test set, and based on the training set, training an integrated sensitivity level model for sensitivity level scoring by selecting features related to the sensitivity of the sensitive data;

[0052] S332. Construct an integrated sensitivity rating model, analyze the sensitivity characteristics of sensitive data through the constructed integrated sensitivity rating model, and generate preliminary sensitivity scoring rules;

[0053] S333. Use the constructed sensitivity level model to calculate the comprehensive sensitivity score of each sensitive data in the test set, and preliminarily divide it into different sensitivity levels according to the score value;

[0054] S334. Based on the preset sensitivity level threshold, the sensitivity score is used to determine whether each sensitive data is highly sensitive data. If the score does not reach the highly sensitive threshold, it is classified as medium or low sensitive data.

[0055] Further, based on the encryption algorithm, the obtained sensitive data of different levels are encrypted, and the encrypted sensitive data of different levels and non-sensitive data are respectively stored in a preset security database, and a data access permission management mechanism is established, including the following steps:

[0056] S41. Generate a random key fragment required for encryption from a preset random number generator initial value according to the level of sensitive data, and initialize a key pool to adapt to different encryption requirements;

[0057] S42, based on sensitive data of different sensitivity levels, generating a random key from a key pool according to the key length required by the encryption algorithm, and assigning a corresponding encryption strength according to the level of the sensitive data;

[0058] S43, based on the encryption algorithm, encrypting the sensitive data of different levels and the generated keys, and storing the encrypted sensitive data separately from its encryption information;

[0059] S44. The encrypted sensitive data is classified and stored in a preset security database according to the sensitivity level, and the non-sensitive data is stored in an unencrypted form in a low-security area, and a hierarchical data access permission management mechanism is established.

[0060] Further, based on sensitive data of different sensitivity levels, generating a random key from a key pool according to the key length required by the encryption algorithm, and assigning a corresponding encryption strength according to the level of the sensitive data includes the following steps:

[0061] S421. Determine the required encryption strength and key length according to the level of sensitive data;

[0062] S422, initializing a random number generator, and randomly extracting a key length matching the target sensitive data level from a preset key pool;

[0063] S423, configuring encryption modes for sensitive data of different levels, storing the generated random keys in a key management system, and assigning a sensitive data level identifier to each key;

[0064] S424. Regularly check the matching between the sensitive data level and the encryption strength. For keys that have not been used for a long time, trigger the key rotation process, generate new random keys and update the encryption strategy.

[0065] Furthermore, the real-time operation data set of the new energy power plant is obtained, and then the data quality evaluation value of the real-time operation data set is obtained to determine whether to optimize the data quality; the specific method for obtaining the data quality evaluation value is as follows: the high-frequency noise energy ratio is obtained according to the relative relationship analysis between the total energy of the high-frequency noise signal after wavelet decomposition and the total signal energy of the original operation data of the new energy power plant; the low-correlation component elimination ratio is obtained according to the relative relationship analysis between the number of eliminated low-correlation mode components and the total number of mode components; the signal-to-noise ratio optimization coefficient is obtained according to the relative relationship analysis between the signal signal-to-noise ratio of the real-time operation data after denoising and the signal signal-to-noise ratio of the original operation data of the new energy power plant; the mode component energy retention coefficient is obtained according to the relative relationship analysis between the energy of the mode component after smoothing and the energy of the mode component of the original operation data of the new energy power plant; the high-frequency noise energy ratio, the low-correlation component elimination ratio, the signal-to-noise ratio optimization coefficient and the mode component energy retention coefficient are processed to obtain the data quality evaluation value.

[0066] Furthermore, it also includes obtaining a classification accuracy evaluation value to determine whether to perform parameter tuning; the specific method for obtaining the classification accuracy evaluation value is as follows: the classification accuracy rate is obtained according to the relative relationship between the number of samples correctly classified as sensitive data, the number of samples correctly classified as non-sensitive data, the number of non-sensitive data samples incorrectly classified as sensitive data, and the number of sensitive data samples incorrectly classified as non-sensitive data; the classification accuracy rate is obtained according to the relative relationship between the number of samples correctly classified as sensitive data and the number of non-sensitive data samples incorrectly classified as sensitive data; the classification precision rate is obtained according to the relative relationship between the number of samples correctly classified as sensitive data and the number of non-sensitive data samples incorrectly classified as non-sensitive data The relative relationship between the number of sense data samples is used to obtain the classification recall rate; the optimized data quality assessment value, classification accuracy, classification precision and classification recall rate are processed to obtain the classification accuracy assessment value; it is determined whether the classification accuracy assessment value is less than the preset classification accuracy threshold in the database: if the classification accuracy assessment value is less than the preset classification accuracy threshold in the database, parameter tuning is performed, otherwise parameter tuning is not performed; it is determined whether the classification accuracy assessment value obtained after parameter tuning is less than the preset classification accuracy threshold in the database: if the classification accuracy assessment value obtained after parameter tuning is less than the preset classification accuracy threshold in the database, feedback is performed, otherwise no feedback is performed.

[0067] Furthermore, it also includes obtaining an encryption effective evaluation value to determine whether to perform data encryption optimization; the specific method for obtaining the encryption effective evaluation value is as follows: obtaining a key length deviation based on the relative relationship between the key length and the preset key length in the database; obtaining a key usage time deviation based on the relative relationship between the key usage time and the preset key usage time in the database; obtaining an encrypted data volume deviation based on the relative relationship between the encrypted data volume and the preset encrypted data volume in the database; processing the key length deviation, the key usage time deviation and the encrypted data volume deviation to obtain a first effective coefficient; when the encryption time is within the preset encryption time range, obtaining an encryption effective evaluation value from the optimized classification accurate evaluation value and the first effective coefficient; when the encryption time is not within the preset encryption time range, obtaining an encryption time deviation from the relative relationship between the encryption time and the maximum encryption time value and the minimum encryption time value in the database; processing the optimized classification accurate evaluation value, the first effective coefficient and the encryption time deviation to obtain an encryption effective evaluation value.

[0068] The beneficial effects of the present invention are:

[0069] 1. The present invention classifies the sensitivity of the real-time operating data of the new energy power plant and protects it in a graded manner. Different levels of data can use encryption algorithms of different strengths, effectively avoiding the problem of resource waste and insufficient protection of sensitive data under uniform encryption strength. The identified sensitive data is encrypted to effectively prevent the risks caused by malicious attacks or accidental leaks, and ensure the confidentiality of the core operating data of the power plant.

[0070] 2. The present invention classifies and identifies data through a data classification algorithm with a high degree of automation, which can reduce manual intervention and management complexity, making data classification more efficient and accurate. Through the data access permission management mechanism, hierarchical access to data of different levels is achieved to ensure that only authorized users can access sensitive data, further improving management efficiency. Different levels of data are treated with corresponding encryption strength and storage strategies, which not only meets the security requirements of highly sensitive data, but also avoids the waste of computing resources caused by using too high encryption strength for low-sensitivity data.

[0071] 3. The present invention obtains the data quality assessment value by processing the high-frequency noise energy ratio obtained by the total energy of the decomposed high-frequency noise signal and the total energy of the original signal, the low-correlation component elimination ratio obtained by the number of eliminated low-correlation mode components and the total number of mode components, the signal-to-noise ratio optimization coefficient obtained by the signal-to-noise ratio of the denoised signal and the signal-to-noise ratio of the original signal, and the mode component energy retention coefficient obtained by the energy of the mode component after smoothing and the energy of the original mode component, thereby quantitatively evaluating the quality of the data after processing, and further realizing the improvement of data quality in new energy power plants.

[0072] 4. The present invention obtains classification accuracy, classification precision and classification recall through the number of samples correctly classified as sensitive data, the number of samples correctly classified as non-sensitive data, the number of non-sensitive data samples incorrectly classified as sensitive data and the number of sensitive data samples incorrectly classified as non-sensitive data; finally, the classification accuracy assessment value is obtained according to the optimized data quality assessment value, classification accuracy, classification precision and classification recall, thereby quantitatively evaluating the accuracy of data classification, and further realizing the improvement of data classification performance in new energy power plants.

[0073] 5. The present invention obtains a first effective coefficient by processing the key length deviation, the key usage time deviation and the encrypted data amount deviation, and then when the encryption time is not within the preset encryption time range, the optimized classification accuracy evaluation value, the first effective coefficient and the encryption time deviation are processed to obtain an encryption effective evaluation value, otherwise the optimized classification accuracy evaluation value and the first effective coefficient are used to obtain an encryption effective evaluation value, thereby quantitatively evaluating the effectiveness of data encryption and further improving data security. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0075] Figure 1 The present invention is a flowchart of a data encryption method for a new energy power plant according to an embodiment of the present invention. DETAILED DESCRIPTION

[0076] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in the field should be able to understand other possible implementations and advantages of the present invention. The components in the figures are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0077] According to an embodiment of the present invention, a data encryption method for a new energy power plant is provided.

[0078] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to the data encryption method for a new energy power plant according to an embodiment of the present invention, the data encryption method comprises the following steps:

[0079] S1. Acquire the real-time operation data of the new energy power plant, and pre-process the real-time operation data to obtain the real-time operation data set of the new energy power plant;

[0080] Specifically, real-time operation data includes: power production related data, environmental monitoring data, equipment operation status data, grid connection data, energy management data, data communication and control information, etc.

[0081] S2. Based on the data classification algorithm, the real-time operation data set of the new energy power plant is classified to identify sensitive data;

[0082] Specifically, sensitive data refers to data with high confidentiality, directly affecting the safety and operational efficiency of power plants, or data that may pose security threats and commercial competition risks. It includes: data related to power production and energy output, equipment operation and maintenance data, grid connection and grid interaction data, safety and protection information, business and operation data, data communication and control information, environment and energy storage data, etc.

[0083] S3. Establish a sensitivity level model, and classify the identified sensitive data by using the sensitivity level model to obtain sensitive data of different levels;

[0084] S4. Based on the encryption algorithm, the obtained sensitive data of different levels are encrypted, and the encrypted sensitive data and non-sensitive data of different levels are stored in a preset security database respectively, and a data access permission management mechanism is established.

[0085] Specifically, the data access permission management mechanism refers to a set of security policies that limit the scope, method, and time of user access to data based on the sensitivity level of the data and the user's permission level. Its core goal is to ensure "minimum authorized access" to data, that is, users can only access the minimum amount of data required for their work or role. The technical support for implementing the data access permission management mechanism includes:

[0086] Access control model:

[0087] 1) Role-based access control (RBAC): Set permissions based on user roles.

[0088] Attribute-based access control (ABAC): Dynamically assign permissions based on data attributes, user attributes, and environment attributes.

[0089] 2) Secure database architecture:

[0090] Data is stored in classified form, and sensitive data is stored in a highly secure encrypted database.

[0091] 3) Rights management system:

[0092] Use permission management platforms (such as IAM tools) to centrally manage user roles and data permissions.

[0093] Preferably, obtaining the real-time operation data of the new energy power plant and preprocessing the real-time operation data to obtain the real-time operation data set of the new energy power plant includes the following steps:

[0094] S11, performing wavelet decomposition on the acquired original operation data of the new energy power plant to separate the high-frequency noise signal, reconstructing the decomposed signal according to the energy size of the frequency band, and obtaining the real-time operation data after denoising;

[0095] Specifically, wavelet decomposition is a signal processing method based on wavelet transform, which is mainly used to decompose complex signals into sub-signals of multiple frequency bands, so as to analyze the different frequency components of the signal. It is a time-frequency analysis method that can provide information of the signal in both the time domain and the frequency domain.

[0096] S12, using the empirical mode decomposition technology to decompose the denoised real-time operation data, extract each mode component, and perform smoothing processing;

[0097] Specifically, Empirical Mode Decomposition (EMD) is a signal processing method mainly used to analyze nonlinear and non-stationary signals. It is an adaptive decomposition algorithm that does not rely on predefined basis functions and can decompose complex signals into several physically meaningful intrinsic mode functions (IMFs). Each IMF reflects the characteristics of the signal at different time scales.

[0098] S13, calculating the correlation between each mode component and the original real-time operation data, and eliminating components with low correlation;

[0099] S14, reconstructing the optimized mode components into a complete data set to form a real-time operation data set of the new energy power plant.

[0100] Preferably, based on a data classification algorithm, data classification is performed on the obtained real-time operation data set of the new energy power plant, and identifying sensitive data includes the following steps:

[0101] S21. Acquire a real-time operation data set of a new energy power plant, and perform normalization processing on the real-time operation data set;

[0102] S22. Assign corresponding weights to different data categories according to the importance of the real-time running data set in the classification process;

[0103] S23. Defining classification standards for sensitive data and classification standards for non-sensitive data through weighted data values, and forming ideal classification targets for sensitive data and non-sensitive data respectively;

[0104] S24. Calculate the grey correlation coefficient between each real-time operation data and the sensitive data classification standard and the non-sensitive data classification standard, and quantify the correlation degree between the real-time operation data and the ideal classification target of sensitive data and the ideal classification target of non-sensitive data respectively;

[0105] S25. Based on the grey correlation coefficient, the grey correlation degree between each real-time operation data and the sensitive data classification standard and the non-sensitive data classification standard is calculated respectively;

[0106] S26. Use grey correlation to construct the classification proximity of sensitive data, identify the sensitive data in the real-time running data set, and identify and label the sensitive data.

[0107] Specifically, the data classification algorithm is the TOP-SIS algorithm based on grey correlation, which is a decision analysis algorithm that combines grey correlation analysis and TOPSIS method. Grey correlation analysis is an analysis method used to process incomplete information and is used for system analysis of small samples and incomplete information. It determines the similarity between factors by calculating the grey correlation between each factor. The TOPSIS method is a multi-attribute decision analysis method.

[0108] Preferably, the formula for calculating the grey correlation between each real-time operation data and the sensitive data classification standard and the non-sensitive data classification standard is:

[0109]

[0110] In the formula, H i ′ represents the grey correlation between the i-th real-time operation data and sensitive data;

[0111] H i ″ represents the grey correlation between the ith real-time operation data and non-sensitive data;

[0112] H represents the grey relational degree;

[0113] F represents the total number of real-time running data;

[0114] j represents real-time running data.

[0115] Preferably, using grey correlation to construct classification proximity of sensitive data, identifying the part of the real-time running data set that belongs to sensitive data, and identifying and labeling the sensitive data include the following steps:

[0116] S261. Compare each real-time operation data in the real-time operation data set with the sensitive data classification standard and the non-sensitive data classification standard, and calculate the grey correlation degree between the two;

[0117] S262. Based on the grey correlation of sensitive data and non-sensitive data, a classification proximity model is constructed as the basis for classification;

[0118] S263. According to the classification proximity model, a classification threshold is set, and the proximity of the real-time operation data is compared with the classification threshold. If the proximity of the real-time operation data is greater than the classification threshold, it is determined to be sensitive data; otherwise, it is determined to be non-sensitive data;

[0119] S264. Label the real-time operation data determined to be sensitive data to identify the sensitive data.

[0120] Preferably, the formula for constructing the classification proximity model is:

[0121]

[0122] Where M i ′ represents the relative closeness between the i-th real-time operation data and the sensitive data;

[0123] M represents relative closeness;

[0124] H i ′ represents the grey correlation between the i-th real-time operation data and sensitive data;

[0125] H i ″ represents the grey correlation between the ith real-time operation data and non-sensitive data;

[0126] H represents the grey relational degree;

[0127] F represents the total number of real-time running data.

[0128] Preferably, a sensitivity level model is established, and the identified sensitive data is classified into different levels by using the sensitivity level model. The steps of obtaining sensitive data of different levels include:

[0129] S31. Compare the identified sensitive data with the preset sensitivity performance indicators, screen out the sensitivity characteristics of the sensitive data, and establish a sensitivity level model;

[0130] S32. Analyze the key factors that affect the sensitivity of sensitive data and set a corresponding sensitivity evaluation level for each key factor;

[0131] S33. Using the sensitivity level model, the screened sensitive data is scored for sensitivity, and each sensitive data is preliminarily classified into different sensitivity levels according to the scoring results;

[0132] S34. Combine the scoring results and the set sensitivity evaluation level to calculate the comprehensive sensitivity value of each data and generate a standardized sensitivity scoring matrix;

[0133] S35. Based on the results of the sensitivity scoring matrix, determine whether the sensitivity level of each sensitive data exceeds the preset threshold. If it exceeds, it is marked as key sensitive data. If it does not exceed, it is classified as medium or low sensitive data.

[0134] Preferably, the screened sensitive data is scored for sensitivity by using a sensitivity rating model, and each sensitive data is preliminarily divided into different sensitivity levels according to the scoring results, including the following steps:

[0135] S331, dividing the screened sensitive data into a training set and a test set, and based on the training set, training an integrated sensitivity level model for sensitivity level scoring by selecting features related to the sensitivity of the sensitive data;

[0136] S332. Construct an integrated sensitivity rating model, analyze the sensitivity characteristics of sensitive data through the constructed integrated sensitivity rating model, and generate preliminary sensitivity scoring rules;

[0137] S333. Use the constructed sensitivity level model to calculate the comprehensive sensitivity score of each sensitive data in the test set, and preliminarily divide it into different sensitivity levels according to the score value;

[0138] S334. Based on the preset sensitivity level threshold, the sensitivity score is used to determine whether each sensitive data is highly sensitive data. If the score does not reach the highly sensitive threshold, it is classified as medium or low sensitive data.

[0139] Preferably, based on the encryption algorithm, the obtained sensitive data of different levels are encrypted, and the encrypted sensitive data of different levels and non-sensitive data are respectively stored in a preset security database, and the establishment of a data access permission management mechanism includes the following steps:

[0140] S41. Generate a random key fragment required for encryption from a preset random number generator initial value according to the level of sensitive data, and initialize a key pool to adapt to different encryption requirements;

[0141] S42, based on sensitive data of different sensitivity levels, generating a random key from a key pool according to the key length required by the encryption algorithm, and assigning a corresponding encryption strength according to the level of the sensitive data;

[0142] S43, based on the encryption algorithm, encrypting the sensitive data of different levels and the generated keys, and storing the encrypted sensitive data separately from its encryption information;

[0143] S44. The encrypted sensitive data is classified and stored in a preset security database according to the sensitivity level, and the non-sensitive data is stored in an unencrypted form in a low-security area, and a hierarchical data access permission management mechanism is established.

[0144] Specifically, the encryption algorithm is the AES algorithm, which is a symmetric encryption algorithm. The specific process is as follows:

[0145] Key expansion: Based on the input key, a set of round keys is generated for subsequent encryption operations.

[0146] Initial round: The plaintext is initially processed, including adding round keys and performing operations such as byte replacement, row shift, and column confusion.

[0147] Multi-round encryption: The result of the initial round is used as input, and multiple rounds of encryption operations are performed in sequence. Each round includes adding round keys and performing operations such as byte replacement, row shift, and column confusion.

[0148] Final Round: In the final round of encryption, the MixColumns operation is omitted.

[0149] Output ciphertext: XOR the result of the final round with the round key to obtain the final ciphertext.

[0150] It should be noted that the number of round keys and encryption rounds in the AES algorithm depends on the key length used. AES has three key lengths to choose from: 128 bits, 192 bits, and 256 bits, and the corresponding number of round keys is 10, 12, and 14 rounds, respectively.

[0151] Preferably, based on sensitive data of different sensitivity levels, generating a random key from a key pool according to the key length required by the encryption algorithm, and assigning a corresponding encryption strength according to the level of the sensitive data includes the following steps:

[0152] S421. Determine the required encryption strength and key length according to the level of sensitive data;

[0153] S422, initializing a random number generator, and randomly extracting a key length matching the target sensitive data level from a preset key pool;

[0154] S423, configuring encryption modes for sensitive data of different levels, storing the generated random keys in a key management system, and assigning a sensitive data level identifier to each key;

[0155] S424. Regularly check the matching between the sensitive data level and the encryption strength. For keys that have not been used for a long time, trigger the key rotation process, generate new random keys and update the encryption strategy.

[0156] Preferably, a real-time operation data set of a new energy power plant is obtained, and then the data quality evaluation value of the real-time operation data set is obtained to determine whether to optimize the data quality; the specific method for obtaining the data quality evaluation value is as follows: the high-frequency noise energy ratio (i.e., ), the high-frequency noise energy ratio is obtained by calculating the ratio of the total energy of the high-frequency noise signal after wavelet decomposition to the total signal energy of the original operation data of the new energy power plant; the low-correlation component elimination ratio (i.e. ), the proportion of low-correlation component elimination is obtained by calculating the ratio of the number of eliminated low-correlation mode components to the total number of mode components; the signal-to-noise ratio optimization coefficient (i.e. ), the signal-to-noise ratio optimization coefficient is obtained by calculating the ratio of the signal-to-noise ratio of the real-time operation data after denoising and the signal-to-noise ratio of the original operation data of the new energy power plant; the mode component energy retention coefficient (i.e. ), the mode component energy retention coefficient is obtained by processing the ratio of the energy of the mode component after smoothing to the energy of the mode component of the original operating data of the new energy power plant; the high-frequency noise energy ratio, the low-correlation component elimination ratio, the signal-to-noise ratio optimization coefficient and the mode component energy retention coefficient are processed to obtain the data quality assessment value.

[0157] The specific method for obtaining the data quality assessment value is as follows:

[0158]

[0159] Where n represents the number of the mode component, n = 1, 2, ..., N, N represents the total number of mode components, ZL represents the data quality assessment value, GPZ′ represents the total energy of the high-frequency noise signal after wavelet decomposition, CSN represents the total signal energy of the original operation data of the new energy power plant, TSM represents the number of low-correlation mode components removed, ZMS represents the total number of mode components, SNR′ represents the signal-to-noise ratio of the real-time operation data after denoising, SNR represents the signal-to-noise ratio of the original operation data of the new energy power plant, MNL n ′ represents the energy of the mode component after smoothing, MNL n Represents the energy of the mode component of the original operating data of the renewable energy power plant.

[0160] Determine whether to optimize data quality based on the data quality assessment value:

[0161] If the data quality assessment value is less than the preset data quality threshold in the database, data quality optimization is performed; if the data quality assessment value is not less than the preset data quality threshold in the database, data quality optimization is not performed; data quality optimization includes data re-cleaning and algorithm parameter adjustment; data re-cleaning includes processing missing values ​​and removing outliers, processing missing values ​​means using linear interpolation method for processing, outliers mean data greater than the sum of the mean value and three times the variance of the real-time operation data set of the new energy power plant and data less than the difference between the mean value and three times the variance of the real-time operation data set of the new energy power plant; algorithm parameter adjustment means optimizing and adjusting the decomposition layers of wavelet decomposition and empirical mode decomposition; selecting the best decomposition layer of wavelet decomposition by cross-validation method; selecting the best decomposition layer of empirical mode decomposition according to the energy attenuation law of the intrinsic mode function component. Among them, the preset data quality threshold is represented by the maximum value of the data quality assessment value in the historical time period.

[0162] Specifically, the data quality assessment value in this algorithm involves processing multiple independent variables (high-frequency noise energy ratio, low-correlation component elimination ratio, signal-to-noise ratio optimization coefficient, and mode component energy retention coefficient), and there is a mutual influence relationship between these independent variables; the higher the high-frequency noise energy ratio, the more high-frequency noise components in the signal, which usually leads to a decrease in the signal-to-noise ratio; the higher the low-correlation component elimination ratio, the more noise or pseudo components are eliminated, which helps to improve the purity of the signal and thus improve the signal-to-noise ratio; reducing the high-frequency noise energy ratio may lead to a decrease in the low-correlation component elimination ratio and the mode component energy retention coefficient; increasing the low-correlation component elimination ratio may also lead to a decrease in the high-frequency noise energy ratio and the mode component energy retention coefficient; an increase in the signal-to-noise ratio optimization coefficient usually means that high-frequency noise is better suppressed, so an increase in the signal-to-noise ratio optimization coefficient may lead to a decrease in the high-frequency noise energy ratio and an increase in the low-correlation component elimination ratio; in summary, the data quality assessment value is negatively correlated with the high-frequency noise energy ratio and the low-correlation component elimination ratio, and the data quality assessment value is positively correlated with the signal-to-noise ratio optimization coefficient. The closer the mode component energy retention coefficient is to 1, the larger the data quality assessment value is.

[0163] By quantitatively analyzing the quality of real-time operating data of renewable energy power plants, it is determined whether the data quality needs to be further optimized, thereby quantitatively evaluating the quality of the processed data, thereby achieving improved data quality in renewable energy power plants and providing a reliable data foundation for subsequent data encryption, analysis and decision-making.

[0164] Preferably, the closeness of the real-time running data is compared with the classification threshold. If the closeness of the real-time running data is greater than the classification threshold, it is determined to be sensitive data, otherwise, it is determined to be non-sensitive data. The classification accuracy evaluation value is then obtained to determine whether to perform parameter tuning. The specific method for obtaining the classification accuracy evaluation value is as follows: the classification accuracy is obtained according to the relative relationship between the number of samples correctly classified as sensitive data, the number of samples correctly classified as non-sensitive data, the number of non-sensitive data samples incorrectly classified as sensitive data, and the number of sensitive data samples incorrectly classified as non-sensitive data; the classification accuracy is obtained according to the relative relationship between the number of samples correctly classified as sensitive data and the number of non-sensitive data samples incorrectly classified as sensitive data; the classification precision is obtained according to the relative relationship between the number of samples correctly classified as sensitive data and the number of samples incorrectly classified as non-sensitive data. The classification recall rate is obtained by the relative relationship between the number of sensitive data samples of the sensitive data; the optimized data quality assessment value, classification accuracy, classification precision and classification recall rate are processed to obtain the classification accuracy assessment value; it is judged whether the classification accuracy assessment value is less than the preset classification accuracy threshold in the database: if the classification accuracy assessment value is less than the preset classification accuracy threshold in the database, parameter tuning is performed, otherwise parameter tuning is not performed; parameter tuning is performed by tuning the parameters in the classification proximity model (such as adjusting the threshold, modifying the training data set, etc.) through the k-fold cross validation method; it is judged whether the classification accuracy assessment value obtained after parameter tuning is less than the preset classification accuracy threshold in the database: if the classification accuracy assessment value obtained after parameter tuning is less than the preset classification accuracy threshold in the database, feedback is performed, otherwise no feedback is performed.

[0165] Among them, the specific method for obtaining the accurate classification evaluation value is as follows:

[0166]

[0167] FL=ZL′*arcosh[exp(ZQL+JQL+ZHL)];

[0168] Wherein, FL represents the classification accuracy evaluation value, ZL′ represents the optimized data quality evaluation value, ZQL represents the classification accuracy, JQL represents the classification precision, ZHL represents the classification recall, TP represents the number of samples correctly classified as sensitive data, TN represents the number of samples correctly classified as non-sensitive data, FP represents the number of non-sensitive data samples incorrectly classified as sensitive data, and FN represents the number of sensitive data samples incorrectly classified as non-sensitive data.

[0169] Specifically, the preset classification accuracy threshold is represented by the maximum value of the classification accuracy evaluation value within the historical time period; the sample set including sensitive data and non-sensitive data is classified and predicted by using the trained classification proximity model, and the classification results are counted; the number of samples correctly classified as sensitive data refers to the number of samples predicted by the model to be sensitive data and are actually sensitive data; the number of samples correctly classified as non-sensitive data refers to the number of samples predicted by the model to be non-sensitive data and are actually non-sensitive data; the number of non-sensitive data samples incorrectly classified as sensitive data refers to the number of samples predicted by the model to be sensitive data but are actually non-sensitive data; the number of sensitive data samples incorrectly classified as non-sensitive data refers to the number of samples predicted by the model to be non-sensitive data but are actually sensitive data.

[0170] The classification accuracy evaluation value in this algorithm involves processing multiple independent variables (classification accuracy, classification precision, and classification recall), and there is a mutual influence relationship between these independent variables; the smaller the data quality evaluation value, the lower the data quality, which may lead to a smaller classification accuracy; the classification accuracy is an important indicator for comprehensively measuring the overall performance of the classification model, but in some cases, such as when the ratio of positive and negative samples is unbalanced, the classification accuracy may not be accurate enough. At this time, the classification accuracy and recall rate can provide more specific information; the classification accuracy reflects the model's prediction accuracy for a certain category of samples, while the classification recall rate reflects the model's prediction completeness for the category of samples. The two are contradictory to a certain extent, that is, improving the classification accuracy may reduce the classification recall rate; in summary, the classification accuracy evaluation value is positively correlated with the classification accuracy, classification precision, and classification recall rate.

[0171] Through the above steps, the accuracy of data classification is quantitatively evaluated, thereby achieving improvement in data classification performance in renewable energy power plants.

[0172] Preferably, based on sensitive data of different sensitivity levels, a random key is generated from a key pool according to the key length required by the encryption algorithm, and the corresponding encryption strength is allocated according to the level of the sensitive data, and then the encryption effectiveness evaluation value is obtained to determine whether to perform data encryption optimization; the specific method for obtaining the encryption effectiveness evaluation value is as follows: the encrypted data includes the key length, the encryption time, the key usage time and the amount of encrypted data; the key length deviation (i.e., the value of the key length difference) is obtained according to the relative relationship between the key length and the preset key length in the database. ), the key length deviation is obtained by performing a ratio operation on the key length and the preset key length in the database; the key usage time deviation (i.e. ), the key usage time deviation is obtained by performing a ratio operation on the key usage time and the preset key usage time in the database; the encrypted data amount deviation (i.e. ), the encrypted data volume deviation is obtained by performing a ratio operation on the encrypted data volume and the preset encrypted data volume in the database; the key length deviation, the key usage time deviation and the encrypted data volume deviation are processed to obtain a first effective coefficient; when the encryption time is within the preset encryption time range, the optimized classification accuracy evaluation value and the first effective coefficient are used to obtain an encryption effective evaluation value; when the encryption time is not within the preset encryption time range, the encryption time deviation (i.e., the relative relationship between the encryption time and the maximum encryption time and the minimum encryption time in the database) is obtained. ), the encryption duration deviation is obtained by performing a difference operation on the encryption duration and the average value of the maximum encryption duration and the minimum encryption duration; the optimized classification accurate evaluation value, the first effective coefficient and the encryption duration deviation are processed to obtain the encryption effective evaluation value.

[0173] The specific method for obtaining the encrypted effective evaluation value is as follows:

[0174]

[0175] In the formula, JM represents the encryption effective evaluation value, JS represents the first effective coefficient, FL' represents the optimized classification accurate evaluation value, JMS represents the encryption duration, and JMS min Indicates the minimum encryption duration, JMS max Indicates the maximum encryption duration, MYC indicates the key length, YXS indicates the key usage duration, MSJ indicates the amount of encrypted data, MYC0 indicates the preset key length, YXS0 indicates the preset key usage duration, and MSJ0 indicates the preset amount of encrypted data.

[0176] Specifically, whether to perform data encryption optimization is determined based on the encryption effectiveness evaluation value: if the encryption effectiveness evaluation value is less than the preset encryption effectiveness threshold in the database, data encryption optimization is performed, otherwise data encryption optimization is not performed; data encryption optimization includes data parallel processing and key segmented storage; data parallel processing means dividing the real-time running data set into sensitive data of different sensitivity levels, and then assigning the sensitive data of different sensitivity levels to different processing units for parallel processing, and finally aggregating the results of all processed units to obtain the final processing result; key segmented storage means dividing the key into multiple sub-keys, and then storing the divided sub-keys in different areas of the security database.

[0177] Among them, the preset encryption validity threshold is represented by the maximum value of the encryption validity evaluation value in the historical time period; the minimum encryption duration is represented by the minimum value of the encryption duration in the historical time period; the maximum encryption duration is represented by the maximum value of the encryption duration in the historical time period; the preset key length is set according to data requirements, and the higher the sensitive data level, the longer the corresponding key length; the preset key usage time is set according to application requirements. For example, if data needs to be stored for a long time, the preset key usage time is set to 1 month; the preset encrypted data volume is represented by the average value of the encrypted data volume in the historical time period.

[0178] Taking the optimized classification accuracy evaluation value of 1.485, encryption duration of 1.573 seconds, minimum encryption duration of 0.5 seconds, maximum encryption duration of 5s, preset key length of 128 bits, preset key usage duration of 30 days, and preset encrypted data volume of 10MB as an example, the change statistics of encryption effective evaluation value are shown in Table 1:

[0179] Table 1 Statistics of changes in encryption effective evaluation values

[0180]

[0181] It can be seen from the first and second groups of data in the table that when the key length increases, the encryption effectiveness evaluation value increases; it can be seen from the second and third groups of data that when the key usage time is closer to the preset key usage time, the encryption effectiveness evaluation value increases; it can be seen from the third and fourth groups of data that when the amount of encrypted data increases, the encryption effectiveness evaluation value decreases.

[0182] The encryption effective evaluation value in this algorithm involves processing multiple independent variables (key length, encryption time, key usage time and encrypted data volume), and there is a mutual influence relationship between these independent variables; the longer the key length, the longer the encryption time may be, because a longer key means a more complex encryption operation, which increases the amount of calculation in the encryption process; the encryption time and the amount of encrypted data are usually positively correlated. When the amount of encrypted data increases, the encryption time will also increase accordingly; in key management, in order to ensure the security of data, keys are usually replaced regularly to avoid the risk of leakage caused by long-term use of keys. When the amount of encrypted data increases, the key usage time may be shortened to reduce the risk of leakage; in summary, the encryption effective evaluation value is positively correlated with the key length, and the encryption effective evaluation value is negatively correlated with the amount of encrypted data. The closer the encryption time is to the preset encryption time range and the closer the key usage time is to the preset key usage market, the greater the encryption effective evaluation value.

[0183] The larger the encryption effective evaluation value is, the more effectively the data is encrypted. Even if the data is stolen or leaked, the attacker cannot easily interpret the content, thus ensuring the privacy and security of the data, thereby improving the accuracy of the classification algorithm. Because the classification algorithm relies on accurate and complete data sets for training and prediction, if the data set is tampered with or damaged, the data quality evaluation value may decrease, which may in turn cause the accuracy of the classification algorithm to be affected and the classification accuracy evaluation value to decrease.

[0184] Through the above steps, the effectiveness of data encryption is quantitatively evaluated, thereby achieving improved data security.

[0185] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A data encryption method for a new energy power plant, characterized in that: The data encryption method comprises the following steps: S1. Acquire the real-time operation data of the new energy power plant, and pre-process the real-time operation data to obtain the real-time operation data set of the new energy power plant; S2. Based on the data classification algorithm, the real-time operation data set of the new energy power plant is classified to identify sensitive data; S3. Establish a sensitivity level model, and classify the identified sensitive data by using the sensitivity level model to obtain sensitive data of different levels; S4. Based on the encryption algorithm, the obtained sensitive data of different levels are encrypted, and the encrypted sensitive data and non-sensitive data of different levels are stored in a preset security database respectively, and a data access permission management mechanism is established.

2. A data encryption method for a new energy power plant according to claim 1, characterized in that: The method of obtaining the real-time operation data of the new energy power plant and preprocessing the real-time operation data to obtain the real-time operation data set of the new energy power plant includes the following steps: S11, performing wavelet decomposition on the acquired original operation data of the new energy power plant to separate the high-frequency noise signal, reconstructing the decomposed signal according to the energy size of the frequency band, and obtaining the real-time operation data after denoising; S12, using the empirical mode decomposition technology to decompose the denoised real-time operation data, extract each mode component, and perform smoothing processing; S13, calculating the correlation between each mode component and the original real-time operation data, and eliminating components with low correlation; S14, reconstructing the optimized mode components into a complete data set to form a real-time operation data set of the new energy power plant.

3. A data encryption method for a new energy power plant according to claim 1, characterized in that: The method of classifying the real-time operation data set of the new energy power plant based on the data classification algorithm and identifying sensitive data includes the following steps: S21. Acquire a real-time operation data set of a new energy power plant, and perform normalization processing on the real-time operation data set; S22. Assign corresponding weights to different data categories according to the importance of the real-time running data set in the classification process; S23. Defining classification standards for sensitive data and classification standards for non-sensitive data through weighted data values, and forming ideal classification targets for sensitive data and non-sensitive data respectively; S24. Calculate the grey correlation coefficient between each real-time operation data and the sensitive data classification standard and the non-sensitive data classification standard, and quantify the correlation degree between the real-time operation data and the ideal classification target of sensitive data and the ideal classification target of non-sensitive data; S25. Based on the grey correlation coefficient, the grey correlation degree between each real-time operation data and the sensitive data classification standard and the non-sensitive data classification standard is calculated respectively; S26. Use grey correlation to construct the classification proximity of sensitive data, identify the sensitive data in the real-time running data set, and identify and label the sensitive data.

4. A data encryption method for a new energy power plant according to claim 3, characterized in that: The formula for calculating the grey correlation between each real-time operation data and the sensitive data classification standard and the non-sensitive data classification standard is: In the formula, H i ′ represents the grey correlation between the i-th real-time operation data and sensitive data; H i ″ represents the grey correlation between the ith real-time operation data and non-sensitive data; H represents the grey relational degree; F represents the total number of real-time running data; j represents real-time running data.

5. A data encryption method for a new energy power plant according to claim 4, characterized in that: The method of using grey correlation to construct classification proximity of sensitive data, identifying the part of sensitive data in the real-time running data set, and identifying and labeling the sensitive data includes the following steps: S261. Compare each real-time operation data in the real-time operation data set with the sensitive data classification standard and the non-sensitive data classification standard, and calculate the grey correlation degree between the two; S262. Based on the grey correlation of sensitive data and non-sensitive data, a classification proximity model is constructed as the basis for classification; S263. According to the classification proximity model, a classification threshold is set, and the proximity of the real-time operation data is compared with the classification threshold. If the proximity of the real-time operation data is greater than the classification threshold, it is determined to be sensitive data; otherwise, it is determined to be non-sensitive data; S264. Label the real-time operation data determined to be sensitive data to identify the sensitive data.

6. A data encryption method for a new energy power plant according to claim 5, characterized in that: The formula for constructing the classification proximity model is: Where M i ′ represents the relative closeness between the i-th real-time operation data and the sensitive data; M represents relative closeness; H i ′ represents the grey correlation between the i-th real-time operation data and sensitive data; H i ″ represents the grey correlation between the ith real-time operation data and non-sensitive data; H represents the grey relational degree; F represents the total number of real-time running data.

7. The data encryption method for a new energy power plant according to claim 1, characterized in that: The step of establishing a sensitivity level model and classifying the identified sensitive data by the sensitivity level model to obtain sensitive data of different levels includes the following steps: S31. Compare the identified sensitive data with the preset sensitivity performance indicators, screen out the sensitivity characteristics of the sensitive data, and establish a sensitivity level model; S32. Analyze the key factors that affect the sensitivity of sensitive data and set a corresponding sensitivity evaluation level for each key factor; S33. Using the sensitivity level model, the screened sensitive data is scored for sensitivity, and each sensitive data is preliminarily classified into different sensitivity levels according to the scoring results; S34. Combine the scoring results and the set sensitivity evaluation level to calculate the comprehensive sensitivity value of each data and generate a standardized sensitivity scoring matrix; S35. Based on the results of the sensitivity scoring matrix, determine whether the sensitivity level of each sensitive data exceeds the preset threshold. If it exceeds, it is marked as key sensitive data. If it does not exceed, it is classified as medium or low sensitive data.

8. A data encryption method for a new energy power plant according to claim 7, characterized in that: The method of using the sensitivity level model to perform sensitivity scoring on the screened sensitive data and preliminarily classifying each sensitive data into different sensitivity levels according to the scoring results includes the following steps: S331, dividing the screened sensitive data into a training set and a test set, and based on the training set, training an integrated sensitivity level model for sensitivity level scoring by selecting features related to the sensitivity of the sensitive data; S332. Construct an integrated sensitivity rating model, analyze the sensitivity characteristics of sensitive data through the constructed integrated sensitivity rating model, and generate preliminary sensitivity scoring rules; S333. Use the constructed sensitivity level model to calculate the comprehensive sensitivity score of each sensitive data in the test set, and preliminarily divide it into different sensitivity levels according to the score value; S334. Based on the preset sensitivity level threshold, the sensitivity score is used to determine whether each sensitive data is highly sensitive data. If the score does not reach the highly sensitive threshold, it is classified as medium or low sensitive data.

9. The data encryption method for a new energy power plant according to claim 1, characterized in that: The method of encrypting the obtained sensitive data of different levels based on the encryption algorithm, and storing the encrypted sensitive data of different levels and non-sensitive data in a preset security database respectively, and establishing a data access permission management mechanism includes the following steps: S41. Generate a random key fragment required for encryption from a preset random number generator initial value according to the level of sensitive data, and initialize a key pool to adapt to different encryption requirements; S42, based on sensitive data of different sensitivity levels, generating a random key from a key pool according to the key length required by the encryption algorithm, and assigning a corresponding encryption strength according to the level of the sensitive data; S43, based on the encryption algorithm, encrypting the sensitive data of different levels and the generated keys, and storing the encrypted sensitive data separately from its encryption information; S44. The encrypted sensitive data is classified and stored in a preset security database according to the sensitivity level, and the non-sensitive data is stored in an unencrypted form in a low-security area, and a hierarchical data access permission management mechanism is established.

10. A data encryption method for a new energy power plant according to claim 9, characterized in that: The method of generating a random key from a key pool based on sensitive data of different sensitivity levels according to the key length required by the encryption algorithm and allocating a corresponding encryption strength according to the level of the sensitive data includes the following steps: S421. Determine the required encryption strength and key length according to the level of sensitive data; S422, initializing a random number generator, and randomly extracting a key length that matches the target sensitive data level from a preset key pool; S423, configuring encryption modes for sensitive data of different levels, storing the generated random keys in a key management system, and assigning a sensitive data level identifier to each key; S424. Regularly check the matching between sensitive data level and encryption strength. For keys that have not been used for a long time, trigger the key rotation process, generate new random keys and update the encryption strategy.

11. A data encryption method for a new energy power plant according to claim 2, characterized in that: The step of obtaining a real-time operating data set of a new energy power plant further includes obtaining a data quality evaluation value of the real-time operating data set to determine whether to perform data quality optimization; The specific method for obtaining the data quality assessment value is as follows: The high-frequency noise energy ratio is obtained by analyzing the relative relationship between the total energy of the high-frequency noise signal after wavelet decomposition and the total energy of the signal of the original operation data of the new energy power plant; The proportion of low-correlation component removal is obtained by analyzing the relative relationship between the number of removed low-correlation pattern components and the number of total pattern components; The signal-to-noise ratio optimization coefficient is obtained by analyzing the relative relationship between the signal-to-noise ratio of the real-time operation data after denoising and the signal-to-noise ratio of the original operation data of the new energy power plant; The energy retention coefficient of the mode component is obtained by analyzing the relative relationship between the energy of the mode component after smoothing and the energy of the mode component of the original operation data of the new energy power plant; The high-frequency noise energy ratio, low-correlation component elimination ratio, signal-to-noise ratio optimization coefficient and mode component energy retention coefficient are processed to obtain the data quality assessment value.

12. A data encryption method for a new energy power plant according to claim 5, characterized in that: It also includes obtaining accurate evaluation values ​​of classification to determine whether to perform parameter tuning; The specific method for obtaining the classification accuracy evaluation value is as follows: The classification accuracy is obtained according to the relative relationship between the number of samples correctly classified as sensitive data, the number of samples correctly classified as non-sensitive data, the number of non-sensitive data samples incorrectly classified as sensitive data, and the number of sensitive data samples incorrectly classified as non-sensitive data; The classification accuracy is obtained based on the relative relationship between the number of samples correctly classified as sensitive data and the number of non-sensitive data samples incorrectly classified as sensitive data; The classification recall rate is obtained based on the relative relationship between the number of samples correctly classified as sensitive data and the number of sensitive data samples incorrectly classified as non-sensitive data; The optimized data quality assessment value, classification accuracy, classification precision and classification recall are processed to obtain a classification accuracy assessment value; Determine whether the classification accuracy evaluation value is less than the preset classification accuracy threshold in the database: If the classification accuracy evaluation value is less than the preset classification accuracy threshold in the database, parameter tuning is performed, otherwise parameter tuning is not performed; Determine whether the classification accuracy evaluation value obtained after parameter tuning is less than the preset classification accuracy threshold in the database: If the classification accuracy evaluation value obtained after parameter tuning is less than the preset classification accuracy threshold in the database, feedback is given, otherwise no feedback is given.

13. The data encryption method for a new energy power plant according to claim 10, characterized in that: It also includes obtaining an encryption effectiveness evaluation value to determine whether to perform data encryption optimization; The specific method for obtaining the encrypted effective evaluation value is as follows: The key length deviation is obtained according to the relative relationship between the key length and the preset key length in the database; The key usage time deviation is obtained according to the relative relationship between the key usage time and the preset key usage time in the database; Obtaining an encrypted data volume deviation according to a relative relationship between the encrypted data volume and a preset encrypted data volume in a database; The key length deviation, the key usage time deviation and the encrypted data amount deviation are processed to obtain a first effective coefficient; When the encryption duration is within the preset encryption duration range, the optimized classification accuracy evaluation value and the first effective coefficient are used to obtain an encryption effective evaluation value; When the encryption duration is not within the preset encryption duration range, the encryption duration deviation is obtained by comparing the relative relationship between the encryption duration and the maximum encryption duration and the minimum encryption duration in the database; The optimized classification accurate evaluation value, the first effective coefficient and the encryption time deviation are processed to obtain the encrypted effective evaluation value.

Images