Vehicle pile network encryption credible data classification method

Abstract

The invention discloses a vehicle pile network encryption credible data classification method, and belongs to the technical field of vehicle network interaction data security, and the method comprises the steps: obtaining vehicle pile network operation interaction data; based on the vehicle pile network operation interaction data, generating standardized encrypted credible data by adopting an encryption algorithm; classifying the standardized encrypted credible data by using the trained homomorphic classification network to obtain a classification result; wherein in the training process of the homomorphic classification network, the weighting coefficient of the hidden layer of the homomorphic classification network is dynamically determined through an adaptive optimization algorithm. The method can break through the limitation of the prior art, realizes credible classification in an encryption state, guarantees the safe and stable operation of the vehicle network interaction system, and is adaptive to a low signal-to-noise ratio environment.

Description

Technical Field

[0001] This invention relates to a method for classifying encrypted and trusted data in a vehicle-to-grid network, belonging to the field of vehicle-to-grid interactive data security technology. Background Technology

[0002] With the large-scale expansion of the electric vehicle industry and the widespread adoption of vehicle-to-grid (V2G) systems, the secure management and accurate classification of vehicle-charging-grid interaction data have become core requirements for ensuring stable system operation. This data covers key parameters such as electric vehicle charging power, voltage, current, charging pile temperature, and the active and reactive power and power frequency deviation of the power distribution network. Its transmission and processing often face complex low signal-to-noise ratio environments and are susceptible to external noise interference.

[0003] Existing data classification technologies generally suffer from two problems: First, most methods require data decryption before classification, which can easily lead to data leakage risks and fail to meet the high security requirements of vehicle-to-grid (V2G) data. Second, there is a lack of effective data preprocessing and network optimization mechanisms. In low signal-to-noise ratio environments, noise interference can severely affect the accuracy of data classification, leading to a significant decrease in the reliability of classification results and failing to provide reliable support for the decision-making of V2G interactive systems. Summary of the Invention

[0004] The purpose of this invention is to provide a method for classifying encrypted and trusted data in a vehicle-charging network, which can achieve trusted classification in an encrypted state and is suitable for low signal-to-noise ratio environments.

[0005] To achieve the above objectives, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for classifying encrypted and trusted data in a vehicle-charging network, comprising: Obtain interactive data on the operation of the vehicle-charging station network; Based on the interactive data of the vehicle-charging network, standardized encrypted and trusted data is generated using encryption algorithms. The standardized encrypted trusted data is classified using a pre-trained homomorphic classification network to obtain the classification results; In the training process of the homomorphic classification network, the weighting coefficients of the hidden layers of the homomorphic classification network are dynamically determined through an adaptive optimization algorithm.

[0006] In conjunction with the first aspect, further, the vehicle-charging-network operation interaction data includes the charging power, charging voltage, charging current of electric vehicles, the temperature of charging piles, and the power parameters of the distribution network; among which, the power parameters of the distribution network include the active power, reactive power, and power frequency deviation of the distribution network.

[0007] In conjunction with the first aspect, further, based on the vehicle-charging station network operation and interaction data, standardized encrypted and trusted data is generated using encryption algorithms, including: The vehicle-charging station network operation interaction data is represented in a matrix format to obtain matrixed vehicle-charging station network operation interaction data. : ; in, , , These represent the charging power, charging voltage, and charging current of the electric vehicle, respectively. One sampling point, The temperature of the charging pile is indicated by the number of... One sampling point, , , These represent the active power, reactive power, and power frequency deviation of the distribution network, respectively. One sampling point, This indicates the total number of sampling points for the vehicle-charging-network operation interaction data; Normalize the matrix-based vehicle-pile network operation interaction data to obtain normalized vehicle-pile network operation interaction data. ; The CKKS homomorphic encryption algorithm is used to encrypt the normalized vehicle-to-charging network operation interaction data to obtain standardized encrypted and reliable data. : ; in, This indicates CKKS homomorphic encryption computation.

[0008] In conjunction with the first aspect, further training methods for homomorphic classification networks include: Initialize the relevant parameters of the homomorphic classification network; among them, the relevant parameters of the homomorphic classification network include the number of neurons in the input layer of the homomorphic classification network. Number of neurons in the hidden layer Number of neurons in the output layer The weight matrix of the input layer and the bias vector of the hidden layer ; Standardized encrypted trusted data is used as encrypted training samples and input into the hidden layer of the homomorphic classification network. In the ciphertext state, based on the weighting coefficients of the hidden layer of the homomorphic classification network and the encrypted training samples, the output matrix of the encrypted hidden layer of the homomorphic classification network is obtained through homomorphic addition and multiplication operations and approximate calculation via activation function. Based on the output matrix of the encrypted hidden layer of the homomorphic classification network and the corresponding encrypted training labels, the weight matrix of the output layer of the homomorphic classification network is estimated by ciphertext calculation, and the weight matrix of the output layer of the homomorphic classification network is homomorphically encrypted by CKKS to obtain the encrypted weight matrix of the output layer of the homomorphic classification network. The relevant parameters of the homomorphic classification network, the weighting coefficients of the hidden layers, and the encrypted weight matrix of the output layer are saved to obtain the trained homomorphic classification network; wherein, when performing the classification task, the weight matrix of the output layer of the homomorphic classification network is always in an encrypted state.

[0009] Building upon the first aspect, the further calculation formula for the hidden layer of a homomorphic classification network is as follows: ; in, This indicates CKKS homomorphic encryption computation. This represents the data matrix of the hidden layers of a homomorphic classification network. This represents standardized encrypted trusted data. and These are all weighting coefficients of the hidden layers in a homomorphic classification network.

[0010] Building upon the first aspect, further, the objective function of the adaptive optimization algorithm used to dynamically determine the weighting coefficients of the hidden layers in a homomorphic classification network is: ; in, This indicates CKKS homomorphic encryption computation. This indicates the CKKS homomorphic decryption computation. Represents the hyperbolic tangent function. This indicates a summation calculation. This represents the data matrix of the hidden layers of a homomorphic classification network. This represents normalized vehicle-charging station network operation and interaction data. and These are all weighting coefficients of the hidden layers in a homomorphic classification network.

[0011] In conjunction with the first aspect, further methods for solving the objective function of adaptive optimization algorithms include: Randomly generate several groups and The candidate values ​​are selected, and each set of candidate values ​​constitutes a parameter set to be optimized, and the positions of each parameter set are initialized. The positions of each parameter group are iteratively optimized. In each iteration, the fitness value corresponding to the current parameter group is calculated based on its current position, and the parameter group with the lowest fitness value is selected as the current globally optimal parameter group. The positions of each parameter group are updated based on the current globally optimal parameter group's position, the historical optimal positions of each parameter group, and a time-decreasing flight coefficient. Constraints are adjusted for positions exceeding the value boundaries after the update. This process continues until the iteration termination condition is met, at which point the final globally optimal parameter group is selected as the global optimal parameter group. and The optimal solution; The formula for calculating the position of each parameter group is: ; in, , They represent the first , During the nth iteration The position of each parameter group Indicates the first During the nth iteration The flight coefficients that decay over time corresponding to each parameter set, , Indicates exponential calculation. Indicates the attenuation coefficient. Indicates the first The position of the globally optimal parameter set at the next iteration. Represents norm calculation, Indicates the first During the nth iteration The flight coefficients that decay over time corresponding to each parameter set, and All are iteration coefficients. Indicates the first During the nth iteration The historical best position of each parameter group Indicates the first During the nth iteration The fitness values ​​corresponding to each parameter group Indicates the first The fitness value corresponding to the globally optimal parameter set at the nth iteration, i.e., the th... The optimal fitness value at the next iteration. This indicates the maximum number of iterations.

[0012] Building upon the first aspect, the formula for applying CKKS homomorphic encryption to the weight matrix of the output layer of a homomorphic classification network is as follows: ; in, This indicates CKKS homomorphic encryption computation. This represents the weight matrix of the encrypted output layer of a homomorphic classification network. This represents the data matrix of the hidden layers of a homomorphic classification network. A mapping vector representing pointers to the locations where encrypted data is stored. Indicates transpose. This indicates homomorphic computation.

[0013] In conjunction with the first aspect, the classification result is further expressed as follows: ; in, Indicates the classification result. This represents the data matrix of the hidden layers of the homomorphic classification network during the classification stage. This represents the weight matrix of the encrypted output layer of a homomorphic classification network. This indicates CKKS homomorphic encryption computation. This indicates the CKKS homomorphic decryption computation. This indicates homomorphic computation.

[0014] Secondly, the present invention provides a vehicle-charging-pile network encrypted trusted data classification system, comprising: The data acquisition module is used to acquire interactive data on the operation of the vehicle-charging station network; The encryption module is used to generate standardized, encrypted, and trusted data based on the interactive data of the vehicle-charging network. The classification module is used to classify standardized encrypted trusted data using a trained homomorphic classification network to obtain classification results; In the training process of the homomorphic classification network, the weighting coefficients of the hidden layers of the homomorphic classification network are dynamically determined through an adaptive optimization algorithm.

[0015] Thirdly, the present invention provides a computer device, comprising: Storage medium: used to store computer programs; Processor: Used to execute the computer program to implement the vehicle-pile network encrypted trusted data classification method described in the first aspect.

[0016] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the vehicle-pile network encrypted trusted data classification method described in the first aspect.

[0017] Fifthly, the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the vehicle-pile network encrypted trusted data classification method described in the first aspect.

[0018] Compared with the prior art, the beneficial effects of the present invention are: The vehicle-to-pile network encrypted trusted data classification method provided by this invention utilizes CKKS homomorphic encryption technology to achieve end-to-end encrypted processing of vehicle-to-pile network data from acquisition and preprocessing to network training and classification. This avoids the leakage risk of existing technologies that require decryption before processing, ensuring data security and privacy. Based on data matrixization, normalization preprocessing, and adaptive optimization algorithms to adjust network parameters, it effectively suppresses noise interference in low signal-to-noise ratio environments, improves classification accuracy, and solves the problem of insufficient classification accuracy in existing technologies, providing reliable data support for V2G systems. Attached Figure Description

[0019] Figure 1 This is a flowchart of the vehicle-charging-network encrypted trusted data classification method provided in an embodiment of the present invention; Figure 2 This is a flowchart of the training process of the homomorphic classification network provided in an embodiment of the present invention; Figure 3 This is a flowchart of the weighting coefficient determination process provided in an embodiment of the present invention. Detailed Implementation

[0020] The technical solution of the present invention will be further described in detail below with reference to specific embodiments.

[0021] Embodiments of the present invention are described in detail below. Examples of these embodiments are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. Unless otherwise specified, embodiments of the present invention and the technical features thereof can be combined with each other.

[0022] This invention provides a method for classifying encrypted trusted data in a vehicle-charging network, comprising: Obtain interactive data on the operation of the vehicle-charging station network; Based on the interactive data of the vehicle-charging network, standardized encrypted and trusted data is generated using encryption algorithms. The standardized encrypted trusted data is classified using a pre-trained homomorphic classification network to obtain the classification results; In the training process of the homomorphic classification network, the weighting coefficients of the hidden layers of the homomorphic classification network are dynamically determined through an adaptive optimization algorithm.

[0023] The vehicle-to-grid network encrypted trusted data classification method provided in this invention can overcome the limitations of existing technologies, achieve trusted classification in an encrypted state, ensure the safe and stable operation of the vehicle-to-grid interaction system, and adapt to low signal-to-noise ratio environments.

[0024] Figure 1 This is a flowchart of the vehicle-charging-pile network encrypted trusted data classification method provided in this embodiment. This flowchart only shows the logical order of the method in this embodiment. Provided there are no conflicts, different flowcharts may be used. Figure 1 Complete the steps shown or described in the order indicated.

[0025] The vehicle-charging network encrypted trusted data classification method provided in this embodiment can be applied to a terminal and can be executed by a vehicle-charging network encrypted trusted data classification system. This system can be implemented by software and / or hardware and can be integrated into the terminal, such as any tablet computer or computer device with communication function.

[0026] This invention provides a method for classifying encrypted trusted data in a vehicle-charging network, which specifically includes the following steps: Step 1: Obtain vehicle-charging station network operation and interaction data; In this embodiment, the vehicle-charging-network operation interaction data includes the electric vehicle's charging power, charging voltage, charging current, charging pile temperature, and power parameters of the power distribution network.

[0027] Specifically, the power parameters of the distribution network include the active power, reactive power, and power frequency deviation of the distribution network.

[0028] Step 2: Based on the vehicle-charging station network operation and interaction data, use encryption algorithms to generate standardized encrypted and trusted data; In this embodiment, the process of generating standardized encrypted and trusted data based on vehicle-charging network operation interaction data and using encryption algorithms specifically includes the following steps: Step 1: Represent the vehicle-charging-network operation interaction data in a matrix format to obtain matrixed vehicle-charging-network operation interaction data. : ; in, , , These represent the charging power, charging voltage, and charging current of the electric vehicle, respectively. One sampling point, The temperature of the charging pile is indicated by the number of... One sampling point, , , These represent the active power, reactive power, and power frequency deviation of the distribution network, respectively. One sampling point, This indicates the total number of sampling points for the vehicle-charging-network operation interaction data; Step 2: Normalize the matrixed vehicle-pile network operation interaction data to obtain normalized vehicle-pile network operation interaction data. ; Step 3: Use the CKKS homomorphic encryption algorithm to encrypt the normalized vehicle-to-charging network operation interaction data to obtain standardized encrypted trusted data. : ; in, This indicates CKKS homomorphic encryption computation.

[0029] Step 3: Use the trained homomorphic classification network to classify the standardized encrypted trusted data and obtain the classification results.

[0030] In this embodiment, the training method for the homomorphic classification network specifically includes the following steps: Step 1: Initialize the relevant parameters of the homomorphic classification network; Specifically, the relevant parameters of homomorphic classification networks include the number of neurons in the input layer of the homomorphic classification network. Number of neurons in the hidden layer Number of neurons in the output layer The weight matrix of the input layer and the bias vector of the hidden layer .

[0031] Step 2: Input standardized encrypted trusted data as encrypted training samples into the hidden layer of the homomorphic classification network. In the ciphertext state, based on the weighting coefficients of the hidden layer of the homomorphic classification network and the encrypted training samples, the output matrix of the encrypted hidden layer of the homomorphic classification network is obtained through homomorphic addition and multiplication operations and approximate calculation via activation function. Specifically, the formula for calculating the hidden layer of a homomorphic classification network is as follows: ; in, This indicates CKKS homomorphic encryption computation. This represents the data matrix of the hidden layers of a homomorphic classification network. This represents standardized encrypted trusted data. and These are all weighting coefficients of the hidden layers in a homomorphic classification network.

[0032] Step 3: Based on the output matrix of the encrypted hidden layer of the homomorphic classification network and the corresponding encrypted training labels, estimate the weight matrix of the output layer of the homomorphic classification network through ciphertext calculation, and perform CKKS homomorphic encryption on the weight matrix of the output layer of the homomorphic classification network to obtain the encrypted weight matrix of the output layer of the homomorphic classification network. Specifically, the formula for applying CKKS homomorphic encryption to the weight matrix of the output layer of a homomorphic classification network is as follows: ; in, This indicates CKKS homomorphic encryption computation. This represents the weight matrix of the encrypted output layer of a homomorphic classification network. This represents the data matrix of the hidden layers of a homomorphic classification network. A mapping vector representing pointers to the locations where encrypted data is stored. Indicates transpose. This indicates homomorphic computation.

[0033] Step 4: Save the relevant parameters of the homomorphic classification network, the weighting coefficients of the hidden layers, and the encrypted weight matrix of the output layer to obtain the trained homomorphic classification network.

[0034] Specifically, when performing classification tasks, the weight matrix of the output layer of the homomorphic classification network is always encrypted.

[0035] In this embodiment, during the training process of the homomorphic classification network, the weighting coefficients of the hidden layers of the homomorphic classification network are dynamically determined through an adaptive optimization algorithm.

[0036] Specifically, the objective function of the adaptive optimization algorithm used to dynamically determine the weighting coefficients of the hidden layers in a homomorphic classification network is: ; in, This indicates CKKS homomorphic encryption computation. This indicates the CKKS homomorphic decryption computation. Represents the hyperbolic tangent function. This indicates a summation calculation. This represents the data matrix of the hidden layers of a homomorphic classification network. This represents normalized vehicle-charging station network operation and interaction data. and These are all weighting coefficients of the hidden layers in a homomorphic classification network.

[0037] In this embodiment, the method for solving the objective function of the adaptive optimization algorithm specifically includes the following steps: Step 1: Randomly generate several groups and The candidate values ​​are selected, and each set of candidate values ​​constitutes a parameter set to be optimized, and the positions of each parameter set are initialized. Step 2: Iteratively optimize the positions of each parameter group. In each iteration, calculate the fitness value of each parameter group based on its current position, and select the parameter group with the lowest fitness value as the current globally optimal parameter group. Update the positions of each parameter group based on the current globally optimal parameter group's position, the historical optimal positions of each parameter group, and a time-decreasing flight coefficient, and adjust the constraints on positions that exceed the value boundaries after the update. Continue this process until the iteration termination condition is met, and then use the final globally optimal parameter group as the final position. and The optimal solution.

[0038] Specifically, the formula for calculating the position of each parameter group is as follows: ; in, , They represent the first , During the nth iteration The position of each parameter group Indicates the first During the nth iteration The flight coefficients that decay over time corresponding to each parameter set, , Indicates exponential calculation. Indicates the attenuation coefficient. Indicates the first The position of the globally optimal parameter set at the next iteration. Represents norm calculation, Indicates the first During the nth iteration The flight coefficients that decay over time corresponding to each parameter set, and All are iteration coefficients. Indicates the first During the nth iteration The historical best position of each parameter group Indicates the first During the nth iteration The fitness values ​​corresponding to each parameter group Indicates the first The fitness value corresponding to the globally optimal parameter set at the nth iteration, i.e., the th... The optimal fitness value at the next iteration. This indicates the maximum number of iterations.

[0039] In this embodiment, a trained homomorphic classification network is used to classify standardized encrypted trusted data to obtain classification results. : ; in, This represents the data matrix of the hidden layers of the homomorphic classification network during the classification stage. This represents the weight matrix of the encrypted output layer of a homomorphic classification network. This indicates CKKS homomorphic encryption computation. This indicates the CKKS homomorphic decryption computation. This indicates homomorphic computation.

[0040] This invention provides a vehicle-charging-pile network encrypted trusted data classification system, comprising: The data acquisition module is used to acquire interactive data on the operation of the vehicle-charging station network; The encryption module is used to generate standardized, encrypted, and trusted data based on the interactive data of the vehicle-charging network. The classification module is used to classify standardized encrypted trusted data using a trained homomorphic classification network to obtain classification results; In the training process of the homomorphic classification network, the weighting coefficients of the hidden layers of the homomorphic classification network are dynamically determined through an adaptive optimization algorithm.

[0041] The vehicle-pile network encrypted trusted data classification system provided in this embodiment of the invention can execute the vehicle-pile network encrypted trusted data classification method provided in this embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0042] This invention provides a computer device, comprising: Storage medium: used to store computer programs; Processor: Used to execute computer programs to implement the vehicle-pile network encrypted trusted data classification method provided in the embodiments of the present invention.

[0043] This invention provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the vehicle-pile network encrypted trusted data classification method provided in this invention.

[0044] This invention provides a computer program product, including a computer program that, when executed by a processor, implements the vehicle-charging-network encrypted trusted data classification method provided in this invention.

[0045] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0046] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart. Figure 1 One or more processes and / or boxes Figure 1 A system that specifies functions in one or more boxes.

[0047] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including an instruction set implemented in a process. Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0048] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0049] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for classifying encrypted and trusted data in a vehicle-charging network, characterized in that, include: Obtain interactive data on the operation of the vehicle-charging station network; Based on the interactive data of the vehicle-charging network, standardized encrypted and trusted data is generated using encryption algorithms. The standardized encrypted trusted data is classified using a pre-trained homomorphic classification network to obtain the classification results; In the training process of the homomorphic classification network, the weighting coefficients of the hidden layers of the homomorphic classification network are dynamically determined through an adaptive optimization algorithm.

2. The vehicle-charging network encrypted trusted data classification method according to claim 1, characterized in that, The vehicle-charging-network operation interaction data includes the charging power, charging voltage, charging current of electric vehicles, the temperature of charging piles, and the power parameters of the distribution network; among which, the power parameters of the distribution network include the active power, reactive power, and power frequency deviation of the distribution network.

3. The vehicle-charging network encrypted trusted data classification method according to claim 1, characterized in that, Based on the interactive data of the vehicle-charging station network, standardized encrypted and trusted data is generated using encryption algorithms, including: The vehicle-charging station network operation interaction data is represented in a matrix format to obtain matrixed vehicle-charging station network operation interaction data. : ； in, , , These represent the charging power, charging voltage, and charging current of the electric vehicle, respectively. One sampling point, The temperature of the charging pile is indicated by the number of... One sampling point, , , These represent the active power, reactive power, and power frequency deviation of the distribution network, respectively. One sampling point, This indicates the total number of sampling points for the vehicle-charging-network operation interaction data; Normalize the matrix-based vehicle-pile network operation interaction data to obtain normalized vehicle-pile network operation interaction data. ; The CKKS homomorphic encryption algorithm is used to encrypt the normalized vehicle-to-charging network operation interaction data to obtain standardized encrypted and reliable data. : ； in, This indicates CKKS homomorphic encryption computation.

4. The vehicle-charging network encrypted trusted data classification method according to claim 1, characterized in that, Training methods for homomorphic classification networks include: Initialize the relevant parameters of the homomorphic classification network; among them, the relevant parameters of the homomorphic classification network include the number of neurons in the input layer of the homomorphic classification network. Number of neurons in the hidden layer Number of neurons in the output layer The weight matrix of the input layer and the bias vector of the hidden layer ; Standardized encrypted trusted data is used as encrypted training samples and input into the hidden layer of the homomorphic classification network. In the ciphertext state, based on the weighting coefficients of the hidden layer of the homomorphic classification network and the encrypted training samples, the output matrix of the encrypted hidden layer of the homomorphic classification network is obtained through homomorphic addition and multiplication operations and approximate calculation via activation function. Based on the output matrix of the encrypted hidden layer of the homomorphic classification network and the corresponding encrypted training labels, the weight matrix of the output layer of the homomorphic classification network is estimated by ciphertext calculation, and the weight matrix of the output layer of the homomorphic classification network is homomorphically encrypted by CKKS to obtain the encrypted weight matrix of the output layer of the homomorphic classification network. The relevant parameters of the homomorphic classification network, the weighting coefficients of the hidden layers, and the encrypted weight matrix of the output layer are saved to obtain the trained homomorphic classification network; wherein, when performing the classification task, the weight matrix of the output layer of the homomorphic classification network is always in an encrypted state.

5. The vehicle-charging network encrypted trusted data classification method according to claim 4, characterized in that, The formula for calculating the hidden layer of a homomorphic classification network is: ； in, This indicates CKKS homomorphic encryption computation. This represents the data matrix of the hidden layers of a homomorphic classification network. This represents standardized, encrypted, and trusted data. and These are all weighting coefficients of the hidden layers in a homomorphic classification network.

6. The vehicle-pile network encrypted trusted data classification method according to claim 4, characterized in that, The objective function of the adaptive optimization algorithm used to dynamically determine the weighting coefficients of the hidden layers in a homomorphic classification network is: ； in, This indicates CKKS homomorphic encryption computation. This indicates the CKKS homomorphic decryption computation. Represents the hyperbolic tangent function. This indicates a summation calculation. This represents the data matrix of the hidden layers of a homomorphic classification network. This represents normalized vehicle-charging station network operation and interaction data. and These are all weighting coefficients of the hidden layers in a homomorphic classification network.

7. The vehicle-charging network encrypted trusted data classification method according to claim 6, characterized in that, Methods for solving the objective function of adaptive optimization algorithms include: Randomly generate several groups and The candidate values ​​are selected, and each set of candidate values ​​constitutes a parameter set to be optimized, and the positions of each parameter set are initialized. The positions of each parameter group are iteratively optimized. In each iteration, the fitness value corresponding to the current parameter group is calculated based on its current position, and the parameter group with the lowest fitness value is selected as the current globally optimal parameter group. The positions of each parameter group are updated based on the current globally optimal parameter group's position, the historical optimal positions of each parameter group, and a time-decreasing flight coefficient. Constraints are adjusted for positions exceeding the value boundaries after the update. This process continues until the iteration termination condition is met, at which point the final globally optimal parameter group is selected as the global optimal parameter group. and The optimal solution; The formula for calculating the position of each parameter group is: ； in, , They represent the first , During the nth iteration The position of each parameter group Indicates the first During the nth iteration The flight coefficients that decay over time corresponding to each parameter set, , Indicates exponential calculation. Indicates the attenuation coefficient. Indicates the first The position of the globally optimal parameter set at the next iteration. Represents norm calculation, Indicates the first During the nth iteration The flight coefficients that decay over time corresponding to each parameter set, and All are iteration coefficients. Indicates the first During the nth iteration The historical best position of each parameter group Indicates the first During the nth iteration The fitness values ​​corresponding to each parameter group Indicates the first The fitness value corresponding to the globally optimal parameter set at the nth iteration, i.e., the th... The optimal fitness value at the next iteration. This indicates the maximum number of iterations.

8. The vehicle-pile network encrypted trusted data classification method according to claim 4, characterized in that, The formula for applying CKKS homomorphic encryption to the weight matrix of the output layer of a homomorphic classification network is as follows: ； in, This indicates CKKS homomorphic encryption computation. This represents the weight matrix of the encrypted output layer of a homomorphic classification network. This represents the data matrix of the hidden layers of a homomorphic classification network. A mapping vector representing pointers to the locations where encrypted data is stored. Indicates transpose. This indicates homomorphic computation.

9. The vehicle-charging network encrypted trusted data classification method according to claim 1, characterized in that, The classification results are represented as follows: ； in, Indicates the classification result. This represents the data matrix of the hidden layers of the homomorphic classification network during the classification stage. This represents the weight matrix of the encrypted output layer of a homomorphic classification network. This indicates CKKS homomorphic encryption computation. This indicates the CKKS homomorphic decryption computation. This indicates homomorphic computation.

10. A vehicle-charging-pile network encrypted trusted data classification system, characterized in that, include: The data acquisition module is used to acquire interactive data on the operation of the vehicle-charging station network; The encryption module is used to generate standardized, encrypted, and trusted data based on the interactive data of the vehicle-charging network. The classification module is used to classify standardized encrypted trusted data using a trained homomorphic classification network to obtain classification results; In the training process of the homomorphic classification network, the weighting coefficients of the hidden layers of the homomorphic classification network are dynamically determined through an adaptive optimization algorithm.

Images