A Method and System for Off-Vehicle Management of Privacy Data of Intelligent Connected Vehicles

Abstract

The present application relates to the technical field of data processing, and provides a method and system for off-vehicle management of privacy data of intelligent connected vehicles. The method includes: classifying and processing an off-vehicle data set according to a predetermined policy to obtain multiple data types and multiple standard data sets; determining multiple unique identifiers based on the matching of multiple data types, and generating multiple encryption keys; calling multiple standard encoders based on the matching of multiple data types to perform differential encoding processing on the multiple standard data sets respectively to obtain multiple differential data sets; encrypting the multiple differential data sets respectively according to the multiple encryption keys, and communicating and transmitting them to a cloud server for storage and processing. The present application solves the technical problems of high privacy leakage risk and low data transmission efficiency in the processing of privacy data in intelligent connected vehicles, and realizes the effect of ensuring the security of privacy data, improving the data transmission speed and processing efficiency by combining data privacy protection and real-time data processing.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method and system for managing privacy data off-vehicle of an intelligent connected vehicle. Background Art

[0002] In the context of the rapid development of intelligent connected vehicles, vehicles not only have highly automated functions, but also generate a large amount of data through various sensors and network connections. These data include driving behavior, real-time location, vehicle status, and environmental information, which greatly enriches the user experience. However, this data generation is also accompanied by serious privacy protection challenges. Users' personal information and behavior patterns are vulnerable to attacks or abuse during data transmission and storage, resulting in privacy leakage. At the same time, existing privacy protection measures are often not perfect and difficult to adapt to the rapidly changing technological environment. In addition, traditional data processing methods are usually concentrated in the cloud, resulting in high latency and bandwidth occupation, which cannot meet the needs of real-time data processing. This not only affects the timeliness of data, but may also form a contradiction between user privacy security and data utilization efficiency. In addition, increasingly stringent data protection regulations in various countries (such as GDPR, etc.) have put forward higher compliance requirements for enterprises, making how to improve data processing efficiency and security while ensuring data privacy a key issue that the industry needs to solve urgently. Summary of the invention

[0003] The present application provides a method and system for off-vehicle management of privacy data of an intelligent connected vehicle, aiming to solve the technical problems of high risk of privacy leakage and low data transmission efficiency when processing privacy data in an intelligent connected vehicle.

[0004] In view of the above problems, the present application provides a method and system for off-vehicle privacy data management of an intelligent connected vehicle.

[0005] In the first aspect disclosed in this application, a method for off-vehicle management of privacy data of an intelligent connected vehicle is provided. The method includes: at an edge node of a target vehicle, classifying and de-identifying an off-vehicle data set according to a predetermined policy to obtain multiple data types and multiple standard data sets; determining multiple unique identifiers based on the matching of the multiple data types and generating multiple encryption keys; invoking multiple standard encoders based on the matching of the multiple data types to perform differential encoding processing on the multiple standard data sets respectively to obtain multiple differential data sets; encrypting the multiple differential data sets respectively according to the multiple encryption keys and communicating and transmitting them to a cloud server for storage and processing; wherein, classifying and de-identifying the off-vehicle data set according to a predetermined policy to obtain multiple data types and multiple standard data sets includes: identifying and dividing the off-vehicle data set according to predetermined data types to obtain multiple data types and multiple initial data sets, wherein the predetermined data types at least include driving data, vehicle status, environmental parameters, and location information; invoking multiple privacy feature recognition models based on the matching of the multiple data types to extract privacy features from the multiple initial data sets to obtain multiple privacy feature sets; masking the multiple privacy feature sets according to a dynamic de-identification scheme to obtain multiple de-identified feature sets, wherein the dynamic de-identification scheme is sent by the cloud server to the edge node and can be updated regularly; using the multiple de-identified feature sets to update the multiple initial data sets to obtain multiple standard data sets; wherein, invoking multiple standard encoders based on the matching of the multiple data types includes: at the cloud server, randomly selecting a first data type of the predetermined data types and invoking a first historical data set within a predetermined historical window with the first data type as a constraint; dividing the first historical data set according to predetermined attribute features to obtain multiple attribute data sets; performing feature clustering analysis on the multiple attribute data sets respectively to determine multiple high-frequency attribute data, and constructing a first standard encoder based on the multiple high-frequency attribute data; establishing a first mapping between the first data type and the first standard encoder, constructing an integrated encoder based on the first mapping, and sending the integrated encoder to the edge node; at the edge node, inputting the multiple data types into the integrated encoder to obtain the multiple standard encoders.

[0006] Another aspect disclosed in this application provides a privacy data off-vehicle management system for intelligent connected vehicles. The system includes: a data classification and processing module, which is used to classify and de-identify an off-vehicle data set according to a predetermined policy at the edge node of the target vehicle to obtain multiple data types and multiple standard data sets; an encryption key generation module, which is used to match and determine multiple unique identifiers based on the multiple data types and generate multiple encryption keys; a differential coding processing module, which is used to call multiple standard encoders based on the matching of the multiple data types and perform differential coding processing on the multiple standard data sets respectively to obtain multiple differential data sets; a data encryption module, which is used to encrypt the multiple differential data sets according to the multiple encryption keys respectively and communicate and transmit them to the cloud server for storage and processing; wherein, the data classification and processing module includes: identifying and dividing the off-vehicle data set according to a predetermined data type to obtain multiple data types and multiple initial data sets, where the predetermined data type at least includes driving data, vehicle status, environmental parameters, and location information; calling multiple privacy feature recognition models based on the matching of the multiple data types and extracting privacy features from the multiple initial data sets to obtain multiple privacy feature sets; masking the multiple privacy feature sets according to a dynamic de-identification scheme to obtain multiple de-identified feature sets, where the dynamic de-identification scheme is sent from the cloud server to the edge node and can be updated regularly; using the multiple de-identified feature sets to update the multiple initial data sets to obtain multiple standard data sets; wherein, the differential coding processing module includes: at the cloud server, randomly selecting a first data type of the predetermined data type and calling a first historical data set within a predetermined historical window with the first data type as a constraint; dividing the first historical data set according to a predetermined attribute feature to obtain multiple attribute data sets; performing feature clustering analysis on the multiple attribute data sets respectively to determine multiple high-frequency attribute data, and constructing a first standard encoder based on the multiple high-frequency attribute data; establishing a first mapping between the first data type and the first standard encoder, constructing an integrated encoder based on the first mapping, and sending the integrated encoder to the edge node; at the edge node, inputting the multiple data types into the integrated encoder to obtain the multiple standard encoders.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The above privacy data off-vehicle management method for intelligent connected vehicles classifies and de-identifies the off-vehicle data set through a predetermined policy, thereby generating multiple data types and standard data sets. This process effectively protects user privacy. Subsequently, multiple unique identifiers and encryption keys are generated according to these data type matches, laying a foundation for subsequent data protection. After that, a standard encoder is used to perform differential encoding on the standard data set to generate a differential data set, which not only improves the data processing efficiency but also reduces the leakage risk. Finally, the generated encryption key is used to encrypt the differential data set and securely transmit it to the cloud server for storage and processing. This method ensures the privacy security of users while quickly and securely transmitting data, improving the overall data management efficiency.

[0009] The above description is only an overview of the technical solution of this application. In order to be able to more clearly understand the technical means of this application, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of this application more obvious and understandable, the following specifically gives the specific implementation manners of this application. Brief Description of the Drawings

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0011] Figure 1 It is a schematic flowchart of a privacy data off-vehicle management method for an intelligent connected vehicle in an embodiment.

[0012] Figure 2 It is an architecture diagram of a privacy data off-vehicle management system for an intelligent connected vehicle in an embodiment.

[0013] Description of the reference numerals: data classification processing module 1, encryption key generation module 2, differential encoding processing module 3, data encryption module 4. Detailed Description of the Embodiments

[0014] The embodiments of this application provide a privacy data off-vehicle management method and system for intelligent connected vehicles, solving the technical problems of high privacy leakage risk and low data transmission efficiency during the processing of privacy data in intelligent connected vehicles.

[0015] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts belong to the scope of protection of the present application.

[0016] It should be noted that the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products, or devices.

[0017] Embodiment 1, as Figure 1 shown, the present application provides a method for managing off-vehicle privacy data of an intelligent connected vehicle, and the method includes:

[0018] At the edge node of the target vehicle, classify and de-identify the off-vehicle data set according to a predetermined policy to obtain multiple data types and multiple standard data sets.

[0019] In the embodiments of the present application, at the edge node of the target vehicle, the system terminal classifies the data generated off-vehicle according to a predetermined policy. This process uses multiple privacy feature recognition models to identify and extract different types of data, such as driving behavior, location information, and environmental parameters. Subsequently, based on the dynamic de-identification scheme issued by the cloud server, these data are de-identified, removing or replacing information that may disclose the user's identity. The results of this classification and de-identification form multiple data types and standard data sets, ensuring the effective protection of user privacy in subsequent data management, while facilitating subsequent analysis and utilization.

[0020] Further, the present application provides classifying and de-identifying the off-vehicle data set according to a predetermined policy to obtain multiple data types and multiple standard data sets, including:

[0021] Identify and partition the off-vehicle data set according to a predetermined data type to obtain multiple data types and multiple initial data sets, where the predetermined data type at least includes driving data, vehicle status, environmental parameters, and location information; match and call multiple privacy feature recognition models according to the multiple data types, and extract privacy features from the multiple initial data sets to obtain multiple privacy feature sets; perform data masking on the multiple privacy feature sets according to a dynamic de-identification scheme to obtain multiple de-identified feature sets, where the dynamic de-identification scheme is sent from the cloud server to the edge node and can be updated regularly; use the multiple de-identified feature sets to update the multiple initial data sets to obtain multiple standard data sets.

[0022] Preferably, when processing the off-vehicle data set, the system terminal identifies and partitions the off-vehicle data set according to a predetermined data type. These data types include, but are not limited to, driving data, vehicle status, environmental parameters, and location information. Among them, the driving data records information such as the vehicle's speed, acceleration, and driving direction. The vehicle status includes indicators such as engine status, fuel level, and fault warnings. The environmental parameters involve external environmental information such as temperature, humidity, and road conditions. The location information includes data such as GPS coordinates, driving routes, and parking locations. Through this identification and partitioning, the system terminal divides the off-vehicle data set into multiple initial data sets, each corresponding to a data type, which is convenient for subsequent processing. Subsequently, use the data type of each initial data set to match the corresponding model from multiple privacy feature recognition models, and use the matched privacy feature recognition model to extract privacy features from the corresponding initial data set. These models can identify which information belongs to sensitive data, such as the user's identity information and specific location. Through this process, multiple privacy feature sets are generated to ensure that sensitive information is marked for subsequent processing. After that, perform data masking on these privacy feature sets according to the dynamic de-identification scheme. This scheme is a strategy for protecting sensitive data according to real-time requirements and environmental changes. It is sent from the cloud server to the edge node and can be updated regularly according to the latest privacy protection requirements to ensure that privacy data remains anonymous during storage and processing and reduce the risk of data leakage. The dynamic de-identification scheme can obscure privacy information such as the owner's name, address information, and age. For example, use a serial number to obscure and hide the owner's name, and change the exact 30 years old to 30 - 40 years old. After the data masking is completed, multiple de-identified feature sets are generated, and the sensitive information in these feature sets is effectively masked. Then, use these de-identified feature sets to update the initial data sets, that is, replace the data in the initial data sets to generate multiple standard data sets. These standard data sets not only retain the necessary data for subsequent analysis but also effectively protect the user's privacy information, ensuring that privacy and security requirements can be met when processing and storing data.

[0023] Determine multiple unique identifiers based on the matching of the multiple data types, and generate multiple encryption keys.

[0024] In one embodiment, the system terminal matches multiple data types based on a type-identifier mapping table to identify unique identifiers for each data type. These unique identifiers can be unique codes or labels for each data type, used to distinguish and identify different data sets. Based on these unique identifiers, the system terminal generates multiple encryption keys. Each encryption key corresponds to the corresponding data type and identifier, ensuring that each data can be independently protected during the data encryption process. This method not only enhances the security of the data, but also ensures that each data set can be encrypted with a specific key during data transmission and storage, avoiding the risk of unauthorized access and leakage.

[0025] Furthermore, the present application provides for determining multiple unique identifiers, including:

[0026] Input the multiple data types into a type-identifier mapping table, and match to obtain multiple unique identifiers. The type-identifier mapping table is sent from the cloud server to the edge node and can be dynamically updated.

[0027] Preferably, the system terminal inputs multiple data types into the type-identifier mapping table, and obtains the corresponding unique identifiers through lookup and matching. This mapping table contains information on each data type and its associated identifier, which can help the system terminal quickly identify and classify data. This mapping table is also sent from the cloud server to the edge node, ensuring the efficiency and consistency of data processing. In addition, the mapping table has the ability to be dynamically updated, and can be adjusted according to the latest requirements and technological changes to adapt to new data types or privacy protection requirements. In this way, when the system terminal processes data, it can always rely on the latest identifier mapping to ensure data security and management flexibility.

[0028] Based on the matching of the multiple data types, call multiple standard encoders to perform differential encoding processing on the multiple standard data sets respectively, and obtain multiple differential data sets.

[0029] In one embodiment, the system terminal inputs multiple data types into the integrated encoder, and calls corresponding multiple standard encoders from the integrated encoder. Each standard encoder performs differential encoding processing for a specific data type to ensure the accuracy and efficiency of data processing. After matching multiple standard encoders, the system terminal inputs multiple standard data sets to be processed into the corresponding standard encoders. These encoders are pre-configured according to the data type to adapt to their respective data characteristics. After the standard encoder receives the corresponding standard data, it calculates the difference between the current data and the predefined standard model. For example, for driving data, the standard encoder calculates the deviation between the actual speed and the standard speed, and identifies the acceleration, deceleration and other behaviors of the vehicle. For environmental data, the standard encoder evaluates the deviation between the current environmental state (such as temperature and humidity) and the set standard state to ensure that the impact of environmental changes is captured. Subsequently, based on the calculated deviation, the standard encoder generates a differential data set. This differential data set does not contain the complete original data, but reflects important change information. For example, for driving data, the output differential data set may only contain key data points such as speed changes, acceleration and braking events, highlighting the dynamic performance of the vehicle. In this way, multiple standard data sets can be effectively processed, key change information can be extracted, and data compression and efficient transmission can be guaranteed. The flexibility and specificity of this process ensures the accuracy and practicality of data coding.

[0030] Further, the present application provides calling multiple standard encoders based on the matching of the multiple data types, including:

[0031] On the cloud server, a first data type of the predetermined data type is randomly selected, and a first historical data set in a predetermined historical window is called with the first data type as a constraint; the first historical data set is divided according to predetermined attribute features to obtain a plurality of attribute data sets; feature clustering analysis is performed on the plurality of attribute data sets respectively to determine a plurality of high-frequency attribute data, and a first standard encoder is constructed based on the plurality of high-frequency attribute data; a first mapping between the first data type and the first standard encoder is established, an integrated encoder is constructed based on the first mapping, and the integrated encoder is decentralized to an edge node; at the edge node, the plurality of data types are input into the integrated encoder to obtain the plurality of standard encoders.

[0032] Preferably, in the cloud server, the system terminal randomly selects any one of the predetermined data types as the first data type. For example, driving data. Constrained by the first data type, the first historical data set within the predetermined historical window is called to obtain historical information related to this data type. Subsequently, according to the predetermined attribute characteristics, the first historical data set is divided. For driving data, these predetermined attribute characteristics include but are not limited to speed, acceleration, direction, and driving route, etc. Through this division, the historical data is subdivided into multiple attribute data sets for more in-depth analysis. Then, the system terminal randomly selects one attribute data set from the multiple attribute data sets as the first attribute data set. By performing the same data clustering on the first attribute data set and screening out the attribute data that meets the predetermined frequency threshold from the clustering results according to the predetermined frequency threshold, and then through weighted calculation, the first high-frequency attribute data is obtained. For other attribute data sets, the same operation is performed to obtain multiple high-frequency attribute data. These high-frequency attribute data may include, common speed ranges (such as 40-60 km / h), typical acceleration patterns (such as hard acceleration and hard braking events), common driving routes (such as main roads and highways), etc. Based on these high-frequency attribute data, the system terminal designs the first standard encoder, including an input layer, a logic layer, and an output layer. The logic layer includes a deviation calculation unit, a condition judgment unit, and a differential data generation unit. Among them, the deviation calculation unit is built-in with deviation calculation logic for calculating the deviation between the current value and the predefined standard model, such as the speed deviation between the actual speed and the standard speed, the acceleration deviation between the current acceleration and the standard acceleration, and the direction deviation between the current driving direction and the standard direction. The condition judgment unit is used to judge whether the trigger condition is reached according to the calculated deviation. For example, if the speed deviation exceeds the threshold, it is marked as an overspeed event; if the acceleration change exceeds a certain range, it needs to be recorded as an acceleration or braking event. The differential data generation unit is used to generate a differential data set according to the results of deviation calculation and condition judgment. For other predetermined data types, the corresponding standard encoder is constructed using the same method. Then, the first mapping relationship between the first data type and the first standard encoder, and the mapping relationships between other data types and their corresponding standard encoders are established. This mapping relationship ensures the association between a specific data type and its corresponding standard encoder. Based on this mapping, the system terminal integrates these standard encodings to construct an integrated encoder and deploys it to the edge node for local data processing. At the edge node, the system terminal inputs multiple data types into the integrated encoder to obtain multiple corresponding standard encoders. Each standard encoder will process the corresponding data type to ensure the accurate encoding and analysis of the data. This process not only improves the efficiency of data processing but also ensures that the standard encoder can dynamically adapt to different types of data.

[0033] Further, the present application provides obtaining a first historical data set, including:

[0034] Determine whether the data quantity of the first historical data set meets a predetermined index. If it does not meet the predetermined index, obtain the vehicle attribute characteristics of the target vehicle; use the vehicle attribute characteristics as comparison constraints to perform an associated search in the user sharing database, obtain a first sample data set, and use the first sample data set to compensate the first historical data set.

[0035] Optionally, the system terminal counts the number of data entries in the collected first historical data set and compares it with the predetermined index. The predetermined index can be the minimum sample quantity requirement. For example, at least 500 pieces of data are required for effective analysis. If the data quantity does not reach the predetermined index, the system terminal extracts relevant vehicle attribute characteristics from the target vehicle, and these characteristics include vehicle type, manufacturing year, driving habits, engine model, etc. Ensure that these characteristics can be effectively used for comparison and retrieval. Subsequently, use the obtained vehicle attribute characteristics as comparison constraints to perform an associated search in the user sharing database. The user sharing database is specially designed to store non-private data and contains historical data of similar vehicles or users. Retrieve through the attribute characteristics to find user data with similar characteristics to the target vehicle. This can improve the relevance and effectiveness of the samples. Then, according to the retrieval results, extract the relevant data set from the user sharing database to form a first sample data set. This sample data set contains driving data that can supplement the first historical data set. Then, use the first sample data set to compensate the first historical data set, that is, merge the data in the first sample data set with the first historical data set to increase the sample quantity. Then, adjust the first historical data set through weighted average or other statistical methods to make it more comprehensive and ensure the improvement of the accuracy of the encoder construction. Through the above process, even if the data quantity of the first historical data set is insufficient, the data can be effectively supplemented, the construction accuracy of the encoder can be improved, and the privacy information of the vehicle owner can be protected at the same time.

[0036] Further, the present application provides determining a plurality of high-frequency attribute data, including:

[0037] Randomly select a first attribute data set from the plurality of attribute data sets; perform the same data clustering on the first attribute data set, set the same data quantity as the data frequency, determine a plurality of same attribute data and a plurality of data frequencies; screen the same attribute data corresponding to the data frequencies that meet the predetermined frequency threshold to obtain a plurality of standard attribute data; set weight coefficients based on the data frequencies, weight the plurality of standard attribute data, obtain a first high-frequency attribute data, and add it to the plurality of high-frequency attribute data.

[0038] Optionally, the system terminal randomly selects one dataset from multiple attribute datasets as the first attribute dataset. This selection ensures the diversity of samples and contributes to subsequent analysis. Subsequently, clustering analysis is performed on the first attribute dataset to group the same data into one category. Each cluster represents data with the same characteristics. This process can calculate the number of data in each category and define it as the data frequency. Then, the same attribute data that meet the predetermined frequency threshold are filtered out, that is, only the data with a frequency higher than the set threshold are retained, so as to ensure that the selected data are representative in the analysis, forming multiple standard attribute datasets. Then, weight coefficients are set based on the data frequency corresponding to each standard attribute data. The higher the frequency of the attribute data, the greater the weight is given to highlight its importance in the dataset. Then, the weighted standard attribute data are integrated into the first high-frequency attribute data and added to the multiple high-frequency attribute data. This step ensures that the final high-frequency attribute data can reflect the main trends and characteristics in the dataset.

[0039] Furthermore, the present application provides a construction of an integrated encoder, and then further includes:

[0040] Obtain multiple standard datasets within a predetermined window, respectively perform data deviation degree analysis on the multiple standard datasets and multiple historical standard datasets in the previous adjacent window to determine multiple data deviation coefficients; respectively perform privacy difference degree analysis on the multiple standard datasets and multiple historical standard datasets to determine multiple privacy deviation coefficients; configure the encoder update period according to the multiple data deviation coefficients and multiple privacy deviation coefficients, and dynamically update the integrated encoder, wherein the data deviation coefficient and the privacy deviation coefficient are negatively correlated with the update period duration.

[0041] Optionally, the system terminal extracts multiple standard data sets according to a predetermined time window. These data sets represent driving data or other attribute data collected during that time period. Subsequently, data deviation analysis is performed on the multiple extracted standard data sets and the multiple historical standard data sets in the previous adjacent window. This analysis calculates the deviation of the standard data eigenvalue by using the ratio of the difference between the current standard data eigenvalue and the historical standard data eigenvalue to the historical standard data eigenvalue. Repeating this process, the deviations of all standard data eigenvalues are calculated, and the mean of these deviations is calculated to obtain the data deviation coefficient between the standard data set and the historical standard data set in the previous adjacent window. The same operation is performed on other standard data sets and their corresponding historical standard data sets to obtain multiple data deviation coefficients. Then, the system terminal extracts the standard privacy features and historical standard privacy features from the standard data set and the corresponding historical standard data set respectively and performs privacy data grouping. For example, classification can be performed according to features such as geographical location and user identity. For each feature group, the occurrence probability of each feature group is calculated, that is, the ratio of the number of samples in the feature group to the total number of samples is calculated. Then, the information entropy of the feature group is calculated. Among them, H is the information entropy, representing the uncertainty or chaos degree of information. The higher the entropy value, the richer the information and the lower the privacy risk. n is the number of feature groups. is the occurrence probability of the i-th feature group. Then, the information entropy difference operation is performed on the feature groups of the standard privacy features and the corresponding feature groups of the historical standard privacy features to obtain multiple information entropy deviations, and the mean of these information entropy deviations is calculated to obtain the privacy deviation coefficient between the standard data set and the historical standard data set in the previous adjacent window. The same operation is performed on other standard data sets and their corresponding historical standard data sets to obtain multiple privacy deviation coefficients. Finally, according to the data deviation coefficient and privacy deviation coefficient obtained from the analysis, the update period of the integrated encoder is configured. When the data deviation coefficient or privacy deviation coefficient is high, it means that the current data is quite different from the historical data. Therefore, the encoder needs to be updated more frequently to ensure the accuracy and effectiveness of the encoder. On the contrary, if the deviation coefficient is low, the update period can be extended to reduce unnecessary frequent updates. According to the configured update period, the integrated encoder is dynamically updated regularly, including updating the predefined standard model, updating the encoding processing logic, etc. This update process ensures that the encoder can adapt to data changes, optimize the effect of differential encoding processing, and further improve data privacy protection and processing accuracy.

[0042] Data encryption is performed on the multiple differential data sets according to the multiple encryption keys respectively, and the encrypted data is communicated and transmitted to the cloud server for storage and processing.

[0043] In one embodiment, for each differential data set, the system terminal extracts the corresponding key from the multiple generated encryption keys. Each differential data set uses a unique encryption key to enhance data security. Subsequently, using a specified encryption algorithm (such as AES, RSA, etc.), each differential data set is encrypted. Through encryption, the data content is converted into an unreadable format to ensure the protection of user privacy during transmission. After that, the encrypted differential data sets are packaged to ensure the integrity and consistency of the data during transmission. This process includes, but is not limited to, adding metadata such as timestamps, data types, etc. for subsequent processing. Then, the encrypted data packets are sent to the cloud server through a secure communication protocol (such as HTTPS, TLS). This step ensures that the data is not intercepted or tampered with during transmission. After the cloud server receives the encrypted data, it is securely stored in a dedicated database, and the system terminal can decrypt and analyze the stored data to support future data processing and decision-making without compromising user privacy and security.

[0044] Furthermore, the present application provides for communication and transmission to the cloud server for storage and processing, and then further includes:

[0045] In the cloud analysis platform, based on the staff permissions, establish a permission mapping association between the staff and the predetermined data types; based on the type-identifier mapping table, determine the matching unique identifier according to the permission mapping association to generate a decryption key, decrypt the differential data set to obtain the decrypted differential data set; according to the permission mapping association, call the matching standard encoder to decode the decrypted differential data set.

[0046] Optionally, in the cloud analysis platform, the system terminal establishes a permission mapping table according to the roles and responsibilities of the staff. This table clarifies the predetermined data types that each staff member can access and operate on, ensuring data security and compliance. Subsequently, using the defined type-identifier mapping table, determine the unique identifier associated with the staff through the permission mapping association. Generate the corresponding decryption key according to this identifier to ensure that only authorized personnel can decrypt the data. After that, use the generated decryption key to decrypt the differential data set stored in the cloud. This step restores the encrypted data to a readable format, enabling the staff to access the real data. After the decryption process is completed, the system terminal obtains a decrypted differential data set. These data will contain the change information of the original data, providing a basis for subsequent analysis. Then, according to the permission mapping association, call the standard encoder that matches the decrypted differential data set. This encoder is responsible for further decoding the decrypted data to restore more detailed and complete information. After decoding, this information can be used for analysis, report generation, decision-making, etc. operations to ensure the effective utilization of the data while still meeting the requirements of privacy protection.

[0047] In summary, the embodiments of the present application have at least the following technical effects:

[0048] In the embodiments of the present application, first, at the edge node of the target vehicle, the off-vehicle data set is classified and de-identified according to a predetermined strategy to obtain multiple data types and multiple standard data sets. Subsequently, based on the matching of these data types, multiple unique identifiers are determined and multiple encryption keys are generated. Then, multiple standard encoders are called to perform differential encoding processing on these standard data sets respectively, so as to obtain multiple differential data sets. After that, according to the generated encryption keys, these differential data sets are encrypted respectively, and the encrypted data is sent to the cloud server for storage and processing through communication transmission. In the generation of data types and standard data sets, the off-vehicle data set is identified and divided according to the predetermined data types to obtain multiple initial data sets such as driving data, vehicle status, environmental parameters, and location information. Then, according to these data types, a privacy feature recognition model is called to extract privacy features, and the privacy feature set is data masked according to the dynamic de-identification scheme, and finally multiple de-identified feature sets are obtained. These technical effects jointly solve the technical problems of high privacy leakage risk and low data transmission efficiency in the processing of privacy data in intelligent connected vehicles, and achieve the effects of ensuring the security of privacy data, improving the data transmission speed and processing efficiency by combining data privacy protection and real-time data processing.

[0049] Embodiment 2, based on the same inventive concept as the method for off-vehicle management of privacy data of an intelligent connected vehicle in the foregoing embodiment, as Figure 2As shown in the figure, the present application provides a privacy data off-vehicle management system for intelligent connected vehicles. The system includes: a data classification and processing module 1: The data classification and processing module 1 is used to classify and de-identify an off-vehicle data set according to a predetermined policy at the edge node of the target vehicle, so as to obtain multiple data types and multiple standard data sets; an encryption key generation module 2: The encryption key generation module 2 is used to match and determine multiple unique identifiers based on the multiple data types, and generate multiple encryption keys; a differential coding processing module 3: The differential coding processing module 3 is used to match and call multiple standard encoders based on the multiple data types, and perform differential coding processing on the multiple standard data sets respectively to obtain multiple differential data sets; a data encryption module 4: The data encryption module 4 is used to encrypt the multiple differential data sets respectively according to the multiple encryption keys, and communicate and transmit them to the cloud server for storage and processing; wherein, the data classification and processing module 1 includes: identifying and dividing the off-vehicle data set according to a predetermined data type to obtain multiple data types and multiple initial data sets, wherein the predetermined data type at least includes driving data, vehicle status, environmental parameters and location information; matching and calling multiple privacy feature recognition models according to the multiple data types, and extracting privacy features from the multiple initial data sets to obtain multiple privacy feature sets; masking the multiple privacy feature sets according to a dynamic de-identification scheme to obtain multiple de-identified feature sets, wherein the dynamic de-identification scheme is sent from the cloud server to the edge node and can be updated regularly; using the multiple de-identified feature sets to update the multiple initial data sets to obtain multiple standard data sets; wherein, the differential coding processing module 3 includes: at the cloud server, randomly selecting a first data type of the predetermined data type, and calling a first historical data set within a predetermined historical window with the first data type as a constraint; dividing the first historical data set according to a predetermined attribute feature to obtain multiple attribute data sets; performing feature clustering analysis on the multiple attribute data sets respectively to determine multiple high-frequency attribute data, and constructing a first standard encoder based on the multiple high-frequency attribute data; establishing a first mapping between the first data type and the first standard encoder, constructing an integrated encoder based on the first mapping, and sending the integrated encoder to the edge node; at the edge node, inputting the multiple data types into the integrated encoder to obtain the multiple standard encoders.

[0050] Further, the encryption key generation module 2 is further used to execute the following method:

[0051] Input the multiple data types into a type-identifier mapping table, and match to obtain multiple unique identifiers. The type-identifier mapping table is sent from the cloud server to the edge node and can be updated dynamically.

[0052] Further, the differential coding processing module 3 is further configured to execute the following method:

[0053] Determine whether the data quantity of the first historical data set meets a predetermined index. If it does not meet the predetermined index, obtain the vehicle attribute characteristics of the target vehicle; use the vehicle attribute characteristics as a comparison constraint to perform an associated search in the user sharing database to obtain a first sample data set, and use the first sample data set to compensate the first historical data set.

[0054] Further, the differential coding processing module 3 is further configured to execute the following method:

[0055] Randomly select a first attribute data set from the multiple attribute data sets; perform clustering of the same data on the first attribute data set, set the same data quantity as the data frequency, and determine multiple same attribute data and multiple data frequencies; screen the same attribute data corresponding to the data frequencies that meet a predetermined frequency threshold to obtain multiple standard attribute data; set weight coefficients based on the data frequencies, weight the multiple standard attribute data, obtain first high-frequency attribute data, and add it to the multiple high-frequency attribute data.

[0056] Further, the differential coding processing module 3 is further configured to execute the following method:

[0057] Obtain multiple standard data sets within a predetermined window, respectively perform data deviation degree analysis on the multiple standard data sets and multiple historical standard data sets in the previous adjacent window to determine multiple data deviation coefficients; respectively perform privacy difference degree analysis on the multiple standard data sets and multiple historical standard data sets to determine multiple privacy deviation coefficients; configure an encoder update period according to the multiple data deviation coefficients and multiple privacy deviation coefficients, and dynamically update the integrated encoder, where the data deviation coefficient and the privacy deviation coefficient are negatively correlated with the update period duration.

[0058] Further, the data encryption module 4 is further configured to execute the following method:

[0059] On the cloud analysis platform, establish a permission mapping association between the staff and a predetermined data type based on the staff's permissions; based on the type-identifier mapping table, determine a matching unique identifier to generate a decryption key according to the permission mapping association, decrypt the differential data set to obtain a decrypted differential data set; call a matching standard encoder according to the permission mapping association to decode the decrypted differential data set.

[0060] It should be noted that the above order of the embodiments of the present application is only for description and does not represent the superiority or inferiority of the embodiments. And the above description of specific embodiments of this specification has been made. The processes depicted in the drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0061] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included within the protection scope of the present application.

[0062] This specification and the drawings are only exemplary descriptions of the present application and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications.

Claims

1. A method for managing privacy data of an intelligent connected vehicle, characterized in that: include: At the edge node of the target vehicle, the off-vehicle data set is classified and de-identified according to a predetermined strategy to obtain multiple data types and multiple standard data sets; determining a plurality of unique identifiers based on the plurality of data type matches, and generating a plurality of encryption keys; Based on the matching of the multiple data types, multiple standard encoders are called to perform differential encoding processing on the multiple standard data sets respectively to obtain multiple differential data sets; Encrypting the plurality of differential data sets respectively according to the plurality of encryption keys, and transmitting the data to a cloud server for storage and processing; The off-vehicle data set is classified and de-identified according to a predetermined strategy to obtain multiple data types and multiple standard data sets, including: Identifying and dividing the off-vehicle data set according to predetermined data types to obtain multiple data types and multiple initial data sets, wherein the predetermined data types at least include driving data, vehicle status, environmental parameters and location information; Calling multiple privacy feature recognition models according to the matching of the multiple data types, extracting privacy features from the multiple initial data sets, and obtaining multiple privacy feature sets; Data shielding is performed on the multiple privacy feature sets according to a dynamic de-identification scheme to obtain multiple de-identification feature sets, wherein the dynamic de-identification scheme is decentralized from the cloud server to the edge node and can be updated regularly; Using the multiple de-identified feature sets, the multiple initial data sets are updated to obtain multiple standard data sets; The calling of multiple standard encoders based on the matching of the multiple data types includes: On the cloud server, randomly selecting a first data type of the predetermined data type, and calling a first historical data set within a predetermined historical window with the first data type as a constraint; Dividing the first historical data set according to predetermined attribute characteristics to obtain multiple attribute data sets; Performing feature clustering analysis on the multiple attribute data sets respectively to determine multiple high-frequency attribute data, and constructing a first standard encoder based on the multiple high-frequency attribute data; Establishing a first mapping between the first data type and the first standard encoder, constructing an integrated encoder based on the first mapping, and delegating the integrated encoder to an edge node; At the edge node, the multiple data types are input into the integrated encoder to obtain the multiple standard encoders.

2. A method for managing private data of an intelligent connected vehicle when leaving the vehicle according to claim 1, characterized in that: The multiple data types are input into a type-identifier mapping table, and multiple unique identifiers are obtained by matching. The type-identifier mapping table is decentralized from the cloud server to the edge node, and can be dynamically updated.

3. The method for managing private data of an intelligent connected vehicle off-vehicle according to claim 1, characterized in that: Get the first historical data set, including: Determining whether the amount of data in the first historical data set meets a predetermined indicator, and if not, obtaining vehicle attribute characteristics of the target vehicle; Using the vehicle attribute characteristics as comparison constraints, an associated search is performed in the user shared database to obtain a first sample data set, and the first sample data set is used to compensate the first historical data set, that is, the data in the first sample data set is merged with the first historical data set to increase the number of samples, and then the first historical data set is adjusted by weighted average.

4. The method for managing private data of an intelligent connected vehicle off-vehicle according to claim 1, characterized in that: Determine multiple high-frequency attribute data, including: Randomly selecting a first attribute data set from the multiple attribute data sets; Clustering the same data on the first attribute data set, setting the number of the same data as the data frequency, and determining a plurality of the same attribute data and a plurality of the data frequencies; Filtering the same attribute data corresponding to the data frequency that meets the predetermined frequency threshold to obtain multiple standard attribute data; A weight coefficient is set based on the data frequency, and the plurality of standard attribute data are weighted to obtain first high-frequency attribute data, which is added to the plurality of high-frequency attribute data.

5. The method for managing private data of an intelligent connected vehicle off-vehicle according to claim 1, characterized in that: Build the integrated encoder, which then also includes: Acquire multiple standard data sets within a predetermined window, perform data deviation analysis on the multiple standard data sets and multiple historical standard data sets within a previous adjacent window, and determine multiple data deviation coefficients; Performing privacy difference analysis on the multiple standard data sets and the multiple historical standard data sets respectively to determine multiple privacy deviation coefficients; The encoder update period is configured according to the multiple data deviation coefficients and the multiple privacy deviation coefficients, and the integrated encoder is dynamically updated, wherein the data deviation coefficient and the privacy deviation coefficient are negatively correlated with the update period duration.

6. The method for managing private data of an intelligent connected vehicle off-vehicle according to claim 2, characterized in that: The communication is transmitted to the cloud server for storage and processing, followed by: On the cloud analysis platform, establish permission mapping associations between staff members and predetermined data types based on staff members’ permissions; Based on the type-identifier mapping table, a matching unique identifier is determined according to the permission mapping association to generate a decryption key, and the differential data set is decrypted to obtain a decrypted differential data set; A matching standard encoder is called according to the permission mapping association to decode the decrypted differential data set.

7. A privacy data off-vehicle management system for an intelligent connected vehicle, characterized in that: The steps for implementing the method for managing privacy data off-vehicle of an intelligent connected vehicle as described in any one of claims 1 to 6 include: Data classification and processing module: At the edge node of the target vehicle, the off-vehicle data set is classified and de-identified according to a predetermined strategy to obtain multiple data types and multiple standard data sets; An encryption key generation module: determining multiple unique identifiers based on the matching of the multiple data types, and generating multiple encryption keys; A differential encoding processing module: calling multiple standard encoders based on the matching of the multiple data types, performing differential encoding processing on the multiple standard data sets respectively, to obtain multiple differential data sets; Data encryption module: encrypts the plurality of differential data sets according to the plurality of encryption keys respectively, and transmits the data to the cloud server for storage and processing.

Images