Data encryption method, model training method, and house rental customer identification method

Abstract

The application relates to a data encryption method, a model training method and a house rental customer identification method, a data encryption device, computer equipment, a storage medium and a computer program product. Power data is acquired, linear classification is performed on the power data, classified data is obtained, a key corresponding to the classified data is determined, and the classified data is homomorphically encrypted according to the key, so that the safety of the power data in the data interaction process can be ensured. The model is trained by using the homomorphically encrypted power data, the accuracy of the neural network model can be improved, and the accuracy of house rental customer identification is improved.

Description

Technical Field

[0001] This application relates to the field of data interaction and sharing technology, and in particular to a data encryption method, a model training method and a housing rental customer identification method, a data encryption device, a computer device, a computer-readable storage medium and a computer program product. Background Technology

[0002] To more effectively utilize power data and discover patterns and value from it, cross-enterprise data sharing is often necessary. However, due to the importance and confidentiality of power data, the integrity of data transmission must be guaranteed. Therefore, the security and privacy protection of cross-power enterprise data sharing is particularly important.

[0003] Currently, cloud services are commonly used for cross-enterprise data interaction and sharing. However, the cloud service-centric system architecture leads to many data silos, which is not conducive to the development of data-driven analysis.

[0004] Therefore, there is an urgent need for a method that can ensure the security of cross-enterprise data interaction and sharing. Summary of the Invention

[0005] Therefore, it is necessary to provide a data encryption method, model training method, and housing rental customer identification method, as well as a data encryption device, computer equipment, computer-readable storage medium, and computer program product that can protect the security of power data interaction, in response to the above-mentioned technical problems.

[0006] Firstly, this application provides a data encryption method, which includes:

[0007] Acquire power data;

[0008] The power data is linearly classified to obtain the classified data;

[0009] Determine the key corresponding to the categorized data;

[0010] The classified data is homomorphically encrypted using the key.

[0011] In one embodiment, after performing linear classification on the power data to obtain the classified data, the process includes:

[0012] The classified data is divided into intrinsic mode function components at each time scale;

[0013] Obtain the data attributes of each component as the corresponding data attributes of the power data.

[0014] In one embodiment, the constraints that the above-mentioned components meet include: the number of extreme points of the component is equal to the number of zero-crossing points of the component; and / or the power data corresponding to the component is symmetrical on the time axis.

[0015] In one embodiment, the above-mentioned homomorphic encryption of the classified data based on the key includes:

[0016] The key is divided into a master key and a subkey;

[0017] The master key is reconstructed by using the subkey to obtain the reconstructed master key;

[0018] The classified data is homomorphically encrypted using the reconstructed master key.

[0019] In one embodiment, the above-mentioned homomorphic encryption of the classified data based on the reconstructed master key includes:

[0020] The calculation formulas for the encryption algorithm and the decryption algorithm are determined based on the reconstructed master key.

[0021] Calculate the ciphertext and plaintext results of the homomorphic encryption algorithm based on the calculation formulas of the encryption and decryption algorithms.

[0022] Homomorphic encryption is applied to the categorized data based on the ciphertext and plaintext results.

[0023] Secondly, this application provides a model training method, which includes:

[0024] Obtain the power data encrypted using any one of the first aspects;

[0025] The encrypted power data is used as input data to obtain the output data of the neural network model.

[0026] The neural network model is trained based on the output data.

[0027] Thirdly, this application provides a method for identifying housing rental customers, the method comprising:

[0028] Obtain the trained neural network model from the second aspect;

[0029] Obtain rental information from home rental customers;

[0030] Rental information is input into a trained neural network model to identify rental customers.

[0031] Fourthly, this application also provides a data encryption device, which includes:

[0032] The data acquisition module is used to acquire power data.

[0033] The data classification module is used to perform linear classification on power data to obtain classified data.

[0034] The key determination module is used to determine the key corresponding to the classified data;

[0035] The data encryption module is used to perform homomorphic encryption on the classified data based on the key.

[0036] Fifthly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method steps of any one of the first, second, or third aspects.

[0037] Sixthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the method steps of any one of the first, second, or third aspects.

[0038] Seventhly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the method steps of any one of the first, second, or third aspects.

[0039] The aforementioned data encryption method, model training method, housing rental customer identification method, data encryption device, computer equipment, storage medium, and computer program product acquire electricity data, perform linear classification on the electricity data to obtain classified data, determine the key corresponding to the classified data, and perform homomorphic encryption on the classified data based on the key. This ensures the security of electricity data during data interaction. Furthermore, using homomorphically encrypted electricity data for model training improves the accuracy of the neural network model, thereby improving the accuracy of housing rental customer identification. Attached Figure Description

[0040] Figure 1 This is a diagram illustrating the application environment of a data encryption method in one embodiment;

[0041] Figure 2 This is a flowchart illustrating a data encryption method in one embodiment;

[0042] Figure 3 for Figure 2 A flowchart illustrating step S204 in the illustrated embodiment;

[0043] Figure 4 for Figure 3 A flowchart illustrating step S303 in the illustrated embodiment;

[0044] Figure 5 This is a flowchart illustrating the data encryption method in another embodiment;

[0045] Figure 6 This is a flowchart illustrating a model training method in one embodiment;

[0046] Figure 7 This is a flowchart illustrating a method for identifying housing rental customers in one embodiment;

[0047] Figure 8 This is a flowchart illustrating a data interaction and sharing security and privacy protection method in one embodiment;

[0048] Figure 9 This is a structural block diagram of a data encryption device in one embodiment;

[0049] Figure 10 This is a structural block diagram of a model training device in one embodiment;

[0050] Figure 11 This is a structural block diagram of a housing rental customer identification device in one embodiment;

[0051] Figure 12 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0053] The data encryption method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or placed on the cloud or other network servers. Server 104 stores power data from various power companies. Terminal 102 retrieves the power data from server 104, performs linear classification on the power data, obtains the classified data, determines the key corresponding to the classified data, and then performs homomorphic encryption on the classified data based on the key. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Server 104 can be implemented using a standalone server or a server cluster composed of multiple servers.

[0054] In one embodiment, such as Figure 2 As shown, a data encryption method is provided, which can be applied to... Figure 1 Taking terminal 102 as an example, the explanation includes the following steps:

[0055] S201: Obtain power data.

[0056] The electricity data includes data from multiple power companies, encompassing urban / rural categories, household age, average monthly and yearly electricity consumption, electricity consumption during the Spring Festival, holiday electricity consumption percentage, off-peak electricity consumption percentage, payment channels, number of overdue payments, changes in payment mobile phone numbers, and mobile phone number locations. It also includes unstructured data such as high-volume residential electricity consumption and service requests from the 95598 online electricity bill inquiry system. In practical applications, time can be used as a metric to collect electricity data within a specific range, thereby constructing electricity consumption characteristics tailored to tenants.

[0057] Off-peak electricity consumption refers to the total electricity consumption between 11 PM and 7 AM the following morning, i.e., the electricity consumption during off-peak hours. The 95598 online electricity bill inquiry system, also known as the 95598 Smart Interactive Website, is the unified external service website of the State Grid Corporation of China. It provides online services such as browsing electricity information, online business processing, online payment, and self-service information inquiry, offering information consultation, communication, and interactive services to electricity customers, and promoting new businesses such as energy efficiency services, electric vehicles, green energy, and smart electricity use. Unstructured data refers to data with irregular or incomplete structures, without a predefined data model, and inconvenient to represent using a two-dimensional logical table in a database. This includes all formats of office documents, text, images, various reports, images, and audio / video information, etc.

[0058] S202: Perform linear classification on the power data to obtain the classified data.

[0059] Among them, linear classification of the collected power data can be achieved by supervised discrete hashing algorithm. Supervised discrete hashing algorithm refers to using the category labels of images or other high-dimensional data to estimate the output of binary coding model. During the iteration process, the error between the actual output and the expected output is continuously calculated using the image label information and fed back to the model to continuously adjust the model parameters.

[0060] Specifically: for datasets in, Let R represent the label matrix, where R is a real number space, d represents the data dimension, and c is the total number of categories.

[0061] Then the nonlinear embedding of the i-th sample It is expressed as follows:

[0062]

[0063] Where {a1,...,a p} represents P anchor points randomly selected from the training set, and σ is the kernel width.

[0064] therefore,

[0065] Define the function F(X) = U T (X), where the matrix is... Its function is to Mapped to a lower-dimensional space.

[0066] Supervised discrete hashing enables accurate collection of power data through binary encoding. The data collection function is constructed as follows:

[0067]

[0068] stB∈{-1,1} K×N

[0069] Where W is the projection matrix, λ is the regularization parameter, v is the penalty parameter, B is the hash encoding matrix corresponding to the data, and st is a mathematical symbol that means that ... satisfies ... or is subject to ... constraints.

[0070] S203: Determine the key corresponding to the classified data.

[0071] The key is a parameter in the data encryption process; it is a parameter input into the algorithm that converts plaintext to ciphertext or vice versa.

[0072] Here, plaintext refers to a cleartext password, meaning a password transmitted or stored in plaintext. Specifically, when storing or transmitting a password over a network, plaintext characters are used without hiding anything, rather than encrypted ciphertext. For example, if the password is 123, the ciphertext password is ***, and the plaintext password is 123.

[0073] Ciphertext is a message and signal generated by a cryptographic system. Without the recipient's instruction to transform the code, i.e., decryption, it cannot be understood by others, thus ensuring the confidentiality of the transmitted information. Ciphertext is a transformation of information in digital communication to guarantee its confidentiality and authenticity. Here, the information is represented as a sequence on a certain q-ary alphabet set. The original information is called plaintext, and the sequence obtained after encryption transformation is called ciphertext. If the ciphertext is obtained by rearranging the elements in the plaintext according to a certain rule, this transformation is called a shift cipher. If there is a one-to-one correspondence between the alphabet and itself, and each element in the plaintext is replaced with its corresponding element to obtain the ciphertext, this transformation is called a substitution cipher.

[0074] S204: Homomorphically encrypt the categorized data based on the key.

[0075] Homomorphic encryption refers to an encryption function that performs addition and multiplication operations on plaintext over a ring before encryption, resulting in an equivalent outcome to performing the corresponding operations on the ciphertext after encryption. An encryption function with homomorphic properties is one that satisfies the condition Dec(En(a)⊙En(b))=a⊕b for two plaintexts a and b, where En is the encryption operation, Dec is the decryption operation, and ⊙ and ⊕ correspond to operations on the plaintext and ciphertext fields, respectively. When ⊕ represents addition, the encryption is called additive homomorphic encryption; when ⊕ represents multiplication, it is called multiplicative homomorphic encryption. Using homomorphic encryption functions to encrypt categorized data can effectively ensure the privacy and security of power data.

[0076] In the above data encryption method, by acquiring power data, linearly classifying the power data to obtain classified data, determining the key corresponding to the classified data, and performing homomorphic encryption on the classified data based on the key, the security of power data during data exchange can be guaranteed.

[0077] In one embodiment, after performing linear classification on the power data to obtain the classified data, the process includes: dividing the classified data into intrinsic mode function components at each time scale; and obtaining the data attributes of each component as the corresponding data attributes of the power data.

[0078] The Intrinsic Mode Function (IMF) is derived from the time-scale characteristics of sequential data. Different data have different IMFs, and each IMF can be considered an inherent mode in the signal. Therefore, the instantaneous frequency obtained through Hilbert transform has a clear physical meaning and can express the local characteristics of the signal. In other words, by dividing the power data into IMF components at different time scales using a hash algorithm, the characteristics of these components can represent the local characteristics of the power data. Therefore, the data attributes of each component are extracted as the corresponding data attributes of the power data.

[0079] In the above embodiments, the classified data is divided into intrinsic mode function components at each time scale, and the data attributes of each component are obtained as the corresponding data attributes of the power data, which can provide data support for subsequent data encryption.

[0080] In one embodiment, the constraints that the above components meet include: the number of extreme points of the component is equal to the number of zero-crossing points of the component; and / or the power data corresponding to the component is symmetrical on the time axis.

[0081] To avoid generating many spurious components that don't actually exist, certain constraints are imposed on the IMF components: First, the number of local extrema and zero-crossings must be equal or differ by at most one over the entire time range, similar to the narrowband requirement of traditional stationary Gaussian signals. Second, at any given time point, the average of the envelopes of local maxima (upper envelope) and local minima (lower envelope) must be zero. This essentially modifies the classic global requirement into a local requirement, preventing the instantaneous frequency from being affected by unnecessary fluctuations caused by asymmetrical waveforms. In reality, this condition should be "the local mean of the data is zero," but for non-stationary data, calculating the local mean involves the concept of a "local time scale," which is difficult to define. Therefore, the average of the envelopes of local maxima and local minima is zero, making the signal waveform locally symmetrical. In general, using this substitution, the instantaneous frequency still conforms to the physical meaning of the system under study.

[0082] In the above embodiments, setting relevant constraints on the components can improve the reliability of IMF components and provide data support for subsequent data encryption.

[0083] In one embodiment, determining the key corresponding to the categorized data includes: performing equi-joins on the categorized data according to data attributes to obtain interactive shared data; and determining the key corresponding to the interactive shared data according to a homomorphic encryption algorithm.

[0084] Equi-join is a common join operation in relational join operations, and is a special case of conditional join (or θ join) when the join operator is "=", i.e., θ = 0. For large-scale power data equi-joins, if all data streams involved in the join have the same join attribute, it is represented as:

[0085] S1.attr = S2.attr = ... = S n .attr

[0086] Where s represents different routes of the data flow, and attr represents the attribute.

[0087] When a data stream passes through a compute node within the cluster, data with the same connection attribute are assigned to the same compute node using a hash function to establish a connection. If the data streams involved in the connection have different connection attributes, it is represented as: (S1.b1=S2.a2)∧(S2.b2=S3.a3)∧...∧(S n-1 .b n-1 =S n .a n This requires multi-hop implementation within the cluster to complete the entire connection.

[0088] For the data after equi-joining, the keys corresponding to each layer of data in the algorithm are determined using a homomorphic encryption algorithm. Specifically, the parameters are first selected: if p s ={p m1 ,...,p mL If} is a set of parameters, then the expressions for the parameters of each layer are:

[0089] p j = L ,...,q0,χ j ,d j N

[0090] in, L ,...,q0> represents the modulus sequence; χ j Indicates the distribution.

[0091] L ,...,q0> shows a decreasing trend, χ j The distribution is as follows, and the formula for calculating N is:

[0092] The encryption and decryption keys for each layer are The expression for R is as follows:

[0093]

[0094] Where Z[x] represents the password generation function, similar to random number generation; x(d) represents the one-way hash function;

[0095] Based on the combination of encryption and decryption keys for each layer in the above formula, the encryption and decryption key S for the homomorphic encryption algorithm is obtained. k With P k , respectively represented as:

[0096] Sk = (S0,...,S) L )

[0097] P k =(P0,...,P L ,τ(S′1→S0),...,τ(S′ L →S L-1 ))

[0098] in:

[0099]

[0100] τ(S′ L →S L-1 ) = s k (S′ L ,S L-1 )

[0101] ​​​S1' represents the components of the encryption key; τ(S1'→S0) represents the transformation process.

[0102] In the above embodiments, the classified data are equi-joined according to the data attributes to obtain interactive shared data, and the key corresponding to the interactive shared data is determined according to the homomorphic encryption algorithm, providing the necessary conditions for subsequent data encryption.

[0103] In one embodiment, such as Figure 3 As shown, the above-mentioned homomorphic encryption of the categorized data based on the key includes:

[0104] S301: Divide the key into a master key and a subkey.

[0105] The master key is the key corresponding to the interactive data. The master key P is divided into n subkeys. The master key is then reconstructed using the subkeys to improve its security.

[0106] S302: Reconstruct the master key using the subkey to obtain the reconstructed master key.

[0107] Specifically, P can be reconstructed if n is greater than or equal to the threshold t; otherwise, it cannot be reconstructed. If the number of participants reconstructing P is equal to or greater than t, the participants can obtain the following using the subkey:

[0108] (x1,y1),(x2,y2),...,(x t ,y t )

[0109] Then the reconstruction of f(x) and P can be achieved separately:

[0110]

[0111]

[0112] Where i and j represent the number of subkeys and the number of participants, respectively; mod indicates modulo operation, i.e., taking the modulo.

[0113] If n and t are equal, and P∈GF(q), P0 selects t-1 elements from GF(q) and distributes the result of the following formula as a subkey to the participants:

[0114]

[0115] If the number of participants is equal to or exceeds t, the participants can reconstruct P using the subkey. The process is as follows:

[0116]

[0117] The participants refer to various enterprise clients.

[0118] S303: Homomorphically encrypt the classified data based on the reconstructed master key.

[0119] The reconstructed master key is used as the latest key for the classified data, and the classified data is homomorphically encrypted using this key.

[0120] In the above embodiments, the key is divided into a master key and a subkey, and the master key is reconstructed using the subkey to obtain the reconstructed master key. Homomorphic encryption is then performed on the classified data based on the reconstructed master key, which can improve the security of the master key and thus improve the data security during data interaction.

[0121] In one embodiment, such as Figure 4 As shown, the above-mentioned homomorphic encryption of the categorized data based on the reconstructed master key includes:

[0122] S401: Determine the calculation formulas for the encryption algorithm and the decryption algorithm based on the reconstructed master key.

[0123] The calculation formula for the encryption algorithm is expressed as follows:

[0124]

[0125] The calculation formula for the decryption algorithm is expressed as follows:

[0126] m*=(( <c,S i >mod q)mod 2)

[0127] Among them, m and P L T r and Rq are both formulas used in cryptography. Cryptography is the technical science that studies the creation and decryption of codes. It studies the objective laws governing code changes. When applied to creating codes to protect the secrets of communication, it is called coding science; when applied to decrypting codes to obtain communication intelligence, it is called decryption science. Together, they are called cryptography.

[0128] S402: Calculate the ciphertext and plaintext results of the homomorphic encryption algorithm based on the calculation formulas of the encryption and decryption algorithms.

[0129] The ciphertext C = [c1 c2 c3 ... c] is calculated using the function f. t ], then decrypt the ciphertext result C f This is equivalent to using a function f to process plaintext M = [m1 m2 m3 ... m t ], thereby obtaining the plaintext result M f Its mathematical expression is:

[0130] C f =f pk [c1 c2 c3 ... c t ]

[0131] =f[m1 m2 m3 ... m t ] = M f

[0132] S403: Homomorphically encrypt the classified data based on the ciphertext and plaintext results.

[0133] The process of calculating the ciphertext and plaintext results is the process of homomorphically encrypting the classified data. Homomorphic encryption of the classified data is achieved through the OR conversion between ciphertext and plaintext.

[0134] In the above embodiments, the calculation formulas for the encryption algorithm and the decryption algorithm are determined based on the reconstructed master key. The ciphertext and plaintext results of the homomorphic encryption algorithm are calculated based on the calculation formulas for the encryption algorithm and the decryption algorithm. Finally, the classified data is homomorphically encrypted based on the ciphertext and plaintext results, which can ensure the security of power data during data interaction.

[0135] In one embodiment, such as Figure 5 As shown, a data encryption method is provided, which includes the following steps:

[0136] S1: Obtain power data.

[0137] S2: Perform linear classification on the power data to obtain the classified data.

[0138] S3: Divide the classified data into intrinsic mode function components at each time scale; the number of extreme points of the components is equal to the number of zero-crossing points of the components; and / or the power data corresponding to the components are symmetrical on the time axis.

[0139] S4: Obtain the data attributes of each component as the corresponding data attributes of the power data.

[0140] S5: Perform iso-joins on the categorized data based on data attributes to obtain interactive shared data.

[0141] S6: Determine the key corresponding to the shared data based on the homomorphic encryption algorithm.

[0142] S7: Divide the key into a master key and a subkey.

[0143] S8: Reconstruct the master key using the subkey to obtain the reconstructed master key.

[0144] S9: Determine the calculation formulas for the encryption algorithm and the decryption algorithm based on the reconstructed master key.

[0145] S10: Calculate the ciphertext and plaintext results of the homomorphic encryption algorithm based on the calculation formulas of the encryption and decryption algorithms.

[0146] S11: Homomorphically encrypt the classified data based on the ciphertext and plaintext results.

[0147] In the above embodiments, by acquiring power data, linearly classifying the power data to obtain classified data, determining the key corresponding to the classified data, and homomorphically encrypting the classified data according to the key, the security of power data during data interaction can be guaranteed.

[0148] In one embodiment, such as Figure 6 As shown, a model training method is provided, which includes the following steps:

[0149] S601: Obtain power data encrypted using a data encryption method.

[0150] Among these, encrypted power data is more accurate. Using encrypted power data as training data for neural network models can ensure the accuracy of the models. Neural network models are mathematical methods that simulate actual human neural networks, and they play an important role in fields such as system identification, pattern recognition, and intelligent control.

[0151] S602: Use the encrypted power data as input data for the neural network model to obtain the output data of the neural network model.

[0152] By inputting encrypted power data into a neural network model, corresponding output data can be obtained. Since the input power data is encrypted, the corresponding output data is also more accurate.

[0153] S603: Train the neural network model based on the output data.

[0154] The process involves obtaining a set of output data from a neural network model, calculating the loss function using the simulated output data of the neural network model, and adjusting the neural network model parameters based on the output of the loss function until the neural network model achieves its optimal performance, thus ending the training of the neural network model.

[0155] The loss function is a computational function used to measure the difference between the model's predicted value f(x) and the true value Y. It is a non-negative real-valued function, usually represented by L(Y,f(x)). The smaller the loss function, the better the robustness of the model.

[0156] In the above embodiments, by acquiring the power data encrypted by the data encryption method and using the encrypted power data as the input data of the neural network model, the output data of the neural network model is obtained. Finally, the neural network model is trained based on the output data, which can improve the model accuracy of the neural network model.

[0157] In one embodiment, such as Figure 7 As shown, a method for identifying housing rental customers is provided, which includes the following steps:

[0158] S701: Obtain the trained neural network model in the model training method.

[0159] Among them, the trained neural network model has a higher accuracy rate. Using the neural network model to identify rental housing customers can accurately identify them, which helps the power system with precision marketing.

[0160] S702: Obtain rental information from housing rental customers.

[0161] The rental information includes customer information related to electricity data, such as the duration or location of the rental property, which can help identify the rental customer's habits.

[0162] S703: Input rental information into the trained neural network model to identify rental customers.

[0163] Among these methods, the trained neural network model for identifying rental customers demonstrates high accuracy and recall, enabling targeted electricity marketing. Recall, referring to the original sample, indicates how many positive cases were correctly predicted. Recall serves to assess the accuracy of the neural network model's identification results.

[0164] In the above embodiments, by obtaining the trained neural network model and the rental information of housing rental customers in the model training method, and inputting the rental information into the trained neural network model to identify housing rental customers, it is possible to accurately identify rental housing customers and provide assistance for the precise marketing of the power system.

[0165] In one embodiment, such as Figure 8 As shown, a method for protecting the security and privacy of cross-power enterprise data interaction and sharing is provided. The method includes the following steps:

[0166] (1) Data scope confirmation, including: urban and rural categories, household age, average monthly electricity consumption, electricity consumption during the Spring Festival, proportion of electricity consumption during holidays, proportion of electricity consumption during off-peak hours, payment channels, number of overdue payments, changes in the mobile phone number used for payment, and location of the mobile phone number. At the same time, unstructured data such as large electricity consumption by residents and 95598 complaint work orders are combined to construct electricity consumption characteristics for tenants.

[0167] (2) Use the supervised discrete hashing algorithm to perform linear data classification on the data from step (1);

[0168] (3) Dataset attribute extraction: The dataset is divided into IMF components at different time scales using a hash algorithm;

[0169] (4) Data sharing: Integrating information between different data streams in multiple power companies through a hash join algorithm;

[0170] (5) Protection of security and privacy in cross-power enterprise data exchange and sharing;

[0171] (6) Using privacy-protected power company data as training data for neural network models to achieve accurate prediction of rental housing customers.

[0172] In the above embodiments, a neural network algorithm is used to accurately predict rental housing customers, which can accurately identify rental housing customers and help the power system with precision marketing.

[0173] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0174] Based on the same inventive concept, this application also provides a data encryption device for implementing the data encryption method described above. The solution provided by this device is similar to the implementation described in the above method; therefore, the specific limitations in one or more data encryption device embodiments provided below can be found in the limitations of the data encryption method described above, and will not be repeated here.

[0175] In one embodiment, such as Figure 9As shown, a data encryption device is provided, comprising: a data acquisition module 901, a data classification module 902, a key determination module 903, and a data encryption module 904, wherein:

[0176] Data acquisition module 901 is used to acquire power data;

[0177] The data classification module 902 is used to perform linear classification on power data to obtain classified data.

[0178] The key determination module 903 is used to determine the key corresponding to the classified data;

[0179] The data encryption module 904 is used to perform homomorphic encryption on the classified data according to the key.

[0180] In one embodiment, the data classification module is further configured to divide the classified data into intrinsic mode function components at each time scale; and obtain the data attributes of each component as the data attributes corresponding to the power data.

[0181] In one embodiment, the constraints that the components involved in the above data classification module meet include: the number of extreme points of the component is equal to the number of zero-crossing points of the component; and / or the power data corresponding to the component is symmetrical on the time axis.

[0182] In one embodiment, the key determination module includes: a data connection unit and a key determination unit, wherein:

[0183] The data connection unit is used to perform equal-value connections on the categorized data according to data attributes to obtain interactive and shared data.

[0184] The key determination unit is used to determine the key corresponding to the shared data based on the homomorphic encryption algorithm.

[0185] In one embodiment, the data encryption module includes: a key partitioning unit, a key reconstruction unit, and a homomorphic encryption unit, wherein:

[0186] A key partitioning unit is used to divide the key into a master key and a subkey;

[0187] The key reconstruction unit is used to reconstruct the master key using the subkey to obtain the reconstructed master key;

[0188] The homomorphic encryption unit is used to homomorphically encrypt the classified data based on the reconstructed master key.

[0189] In one embodiment, the homomorphic encryption unit comprises: a formula calculation subunit, a result calculation subunit, and a homomorphic encryption subunit, wherein:

[0190] The formula calculation subunit is used to determine the calculation formulas for the encryption algorithm and the decryption algorithm based on the reconstructed master key.

[0191] The result calculation subunit is used to calculate the ciphertext and plaintext results of the homomorphic encryption algorithm according to the calculation formula of the encryption algorithm and the calculation formula of the decryption algorithm.

[0192] The homomorphic encryption subunit is used to perform homomorphic encryption on the classified data based on the ciphertext and plaintext results.

[0193] In one embodiment, such as Figure 10 As shown, a model training device is provided, including: a data acquisition module 1001, a data output module 1002, and a model training module 1003, wherein:

[0194] Data acquisition module 1001 is used to acquire power data encrypted by the data encryption device;

[0195] The data output module 1002 is used to take the encrypted power data as input data of the neural network model and obtain the output data of the neural network model.

[0196] The model training module 1003 is used to train the neural network model based on the output data.

[0197] In one embodiment, such as Figure 11 As shown, a housing rental customer identification device is provided, including: a model acquisition module 1101, an information acquisition module 1102, and a customer identification module 1103, wherein:

[0198] The model acquisition module 1101 is used to acquire the trained neural network model in the model training device;

[0199] Information acquisition module 1102 is used to acquire rental information of housing rental customers;

[0200] The customer identification module 1103 is used to input rental information into the trained neural network model to identify rental customers.

[0201] Each module in the aforementioned data encryption device, model training device, and housing rental customer identification device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0202] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 12As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When executed by the processor, the computer program implements a data encryption method, a model training method, or a method for identifying rental customers. The display unit is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0203] Those skilled in the art will understand that Figure 12 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0204] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to perform the following steps: acquiring power data; performing linear classification on the power data to obtain classified data; determining a key corresponding to the classified data; and performing homomorphic encryption on the classified data based on the key.

[0205] In one embodiment, when the processor executes a computer program involving linear classification of power data to obtain classified data, the process includes: dividing the classified data into intrinsic mode function components at each time scale; and obtaining the data attributes of each component as the data attributes corresponding to the power data.

[0206] In one embodiment, the constraints that the components involved in the processor executing the computer program meet include: the number of extreme points of the component is equal to the number of zero-crossing points of the component; and / or the power data corresponding to the component is symmetrical on the time axis.

[0207] In one embodiment, determining the key corresponding to the categorized data when the processor executes a computer program includes: performing equi-joins on the categorized data according to data attributes to obtain interactive shared data; and determining the key corresponding to the interactive shared data according to a homomorphic encryption algorithm.

[0208] In one embodiment, the homomorphic encryption of classified data based on a key when the processor executes a computer program includes: dividing the key into a master key and a subkey; reconstructing the master key using the subkey to obtain a reconstructed master key; and homomorphically encrypting the classified data based on the reconstructed master key.

[0209] In one embodiment, the process of homomorphically encrypting classified data based on a reconstructed master key when the processor executes a computer program includes: determining the calculation formulas for an encryption algorithm and a decryption algorithm based on the reconstructed master key; calculating the ciphertext and plaintext results of the homomorphic encryption algorithm based on the calculation formulas for the encryption and decryption algorithms; and homomorphically encrypting the classified data based on the ciphertext and plaintext results.

[0210] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to perform the following steps: acquiring power data encrypted by the above-described data encryption method; using the encrypted power data as input data for a neural network model to obtain output data for the neural network model; and training the neural network model based on the output data.

[0211] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to perform the following steps: acquiring a trained neural network model in a model training method; acquiring rental information of housing rental customers; and inputting the rental information into the trained neural network model to identify housing rental customers.

[0212] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, performs the following steps: acquiring power data; performing linear classification on the power data to obtain classified data; determining a key corresponding to the classified data; and performing homomorphic encryption on the classified data based on the key.

[0213] In one embodiment, when a computer program is executed by a processor, it involves linearly classifying power data to obtain classified data, and then includes: dividing the classified data into intrinsic mode function components at each time scale; and obtaining the data attributes of each component as the data attributes corresponding to the power data.

[0214] In one embodiment, the constraints that the components involved in the execution of the computer program by the processor meet include: the number of extreme points of the component is equal to the number of zero-crossing points of the component; and / or the power data corresponding to the component is symmetrical on the time axis.

[0215] In one embodiment, the process of determining the key corresponding to the categorized data when the computer program is executed by the processor includes: performing equi-joins on the categorized data according to data attributes to obtain interactive shared data; and determining the key corresponding to the interactive shared data according to a homomorphic encryption algorithm.

[0216] In one embodiment, when a computer program is executed by a processor, the homomorphic encryption of classified data based on a key includes: dividing the key into a master key and a subkey; reconstructing the master key using the subkey to obtain a reconstructed master key; and homomorphically encrypting the classified data based on the reconstructed master key.

[0217] In one embodiment, when a computer program is executed by a processor, the process of homomorphically encrypting classified data based on a reconstructed master key includes: determining the calculation formulas for an encryption algorithm and a decryption algorithm based on the reconstructed master key; calculating the ciphertext and plaintext results of the homomorphic encryption algorithm based on the calculation formulas for the encryption and decryption algorithms; and homomorphically encrypting the classified data based on the ciphertext and plaintext results.

[0218] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, it performs the following steps: acquiring power data encrypted by the above-described data encryption method; using the encrypted power data as input data for a neural network model to obtain output data for the neural network model; and training the neural network model based on the output data.

[0219] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, performs the following steps: obtaining a trained neural network model in a model training method; obtaining rental information of housing rental customers; and inputting the rental information into the trained neural network model to identify housing rental customers.

[0220] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps: acquiring power data; performing linear classification on the power data to obtain classified data; determining a key corresponding to the classified data; and performing homomorphic encryption on the classified data based on the key.

[0221] In one embodiment, when a computer program is executed by a processor, it involves linearly classifying power data to obtain classified data, and then includes: dividing the classified data into intrinsic mode function components at each time scale; and obtaining the data attributes of each component as the data attributes corresponding to the power data.

[0222] In one embodiment, the constraints that the components involved in the execution of the computer program by the processor meet include: the number of extreme points of the component is equal to the number of zero-crossing points of the component; and / or the power data corresponding to the component is symmetrical on the time axis.

[0223] In one embodiment, the process of determining the key corresponding to the categorized data when the computer program is executed by the processor includes: performing equi-joins on the categorized data according to data attributes to obtain interactive shared data; and determining the key corresponding to the interactive shared data according to a homomorphic encryption algorithm.

[0224] In one embodiment, when a computer program is executed by a processor, the homomorphic encryption of classified data based on a key includes: dividing the key into a master key and a subkey; reconstructing the master key using the subkey to obtain a reconstructed master key; and homomorphically encrypting the classified data based on the reconstructed master key.

[0225] In one embodiment, when a computer program is executed by a processor, the process of homomorphically encrypting classified data based on a reconstructed master key includes: determining the calculation formulas for an encryption algorithm and a decryption algorithm based on the reconstructed master key; calculating the ciphertext and plaintext results of the homomorphic encryption algorithm based on the calculation formulas for the encryption and decryption algorithms; and homomorphically encrypting the classified data based on the ciphertext and plaintext results.

[0226] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps: acquiring power data encrypted using the above-described data encryption method; using the encrypted power data as input data to a neural network model to obtain output data of the neural network model; and training the neural network model based on the output data.

[0227] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps: obtaining a trained neural network model in a model training method; obtaining rental information of housing rental customers; and inputting the rental information into the trained neural network model to identify housing rental customers.

[0228] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0229] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0230] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A data encryption method, characterized in that, The method includes: Acquire power data; The power data is linearly classified using a supervised discrete hash algorithm to obtain the classified data. The classified data is divided into intrinsic mode function components at each time scale; the constraints that the components meet include: the number of extreme points of the component is equal to the number of zero-crossing points of the component; and / or the power data corresponding to the component is symmetrical on the time axis; The data attributes of each component are obtained as the data attributes corresponding to the power data; Based on the data attributes, the categorized data is joined by equal-value pairs to obtain interactive shared data; The key corresponding to the shared interactive data is determined based on a homomorphic encryption algorithm; The categorized data is homomorphically encrypted using the key.

2. The method according to claim 1, characterized in that, The step of homomorphically encrypting the categorized data according to the key includes: The key is divided into a master key and a subkey; The master key is reconstructed using the subkey to obtain the reconstructed master key; The classified data is homomorphically encrypted using the reconstructed master key.

3. The method according to claim 2, characterized in that, The step of homomorphically encrypting the categorized data based on the reconstructed master key includes: The calculation formulas for the encryption algorithm and the decryption algorithm are determined based on the reconstructed master key. The ciphertext and plaintext results of the homomorphic encryption algorithm are calculated based on the calculation formulas of the encryption and decryption algorithms. The classified data is homomorphically encrypted based on the ciphertext and plaintext results.

4. A model training method, characterized in that, The method includes: To obtain power data encrypted according to any one of claims 1 to 3; The encrypted power data is used as input data for a neural network model to obtain the output data of the neural network model. The neural network model is trained based on the output data.

5. A method for identifying housing rental customers, characterized in that, The method includes: Obtain the trained neural network model as described in claim 4; Obtain rental information from housing rental customers, including customer information related to electricity data; The rental information is input into the trained neural network model to identify rental customers.

6. A data encryption device, characterized in that, The device includes: The data acquisition module is used to acquire power data; The data classification module is used to linearly classify the power data using a supervised discrete hash algorithm to obtain classified data. The data classification module is also used to divide the classified data into intrinsic mode function components at each time scale. The constraints met by the components include: the number of extreme points of the component is equal to the number of zero-crossing points of the component; and / or the power data corresponding to the component is symmetric on the time axis. The data attributes of each component are obtained as the data attributes corresponding to the power data. The key determination module includes: a data connection unit, used to perform equal-value connection on the classified data according to the data attributes to obtain interactive shared data; and a key determination unit, used to determine the key corresponding to the interactive shared data according to the homomorphic encryption algorithm. The data encryption module is used to perform homomorphic encryption on the classified data according to the key.

7. The apparatus according to claim 6, characterized in that, The data encryption module includes: A key partitioning unit is used to divide the key into a master key and a subkey; The key reconstruction unit is used to reconstruct the master key using the subkey to obtain the reconstructed master key; The homomorphic encryption unit is used to homomorphically encrypt the classified data based on the reconstructed master key.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 3, 4, or 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 3, 4, or 5.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the method according to any one of claims 1 to 3, 4, or 5.

Images