Power battery information storage and retrieval device and method for differential privacy calculation vectorization storage

Abstract

The invention provides a power battery information storage and retrieval device and method for differential privacy calculation vectorization storage, and belongs to the technical field of new energy automobile power batteries. Carrying out vectorization processing on the collected power battery data, and carrying out differential privacy calculation and storage; based on a vector database, data retrieval and application are carried out, battery power consumption balance, temperature control and charging strategy optimization are achieved, or the method is used for battery life prediction and fault diagnosis; based on a differential privacy computing technology, a vectorization storage database and a vectorization query technology, through data storage and retrieval control of the power battery, accurate control of power battery echelon utilization is realized, data support is provided for battery full-life prediction management, and the power battery life prediction management efficiency is improved. The problems of storage, retrieval and privacy security protection of battery data are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of new energy vehicle power batteries, and specifically, relates to a power battery information storage and retrieval device and method with differential privacy computing vectorized storage. Background Art

[0002] The data storage and management of power battery data in new energy vehicles is a key technical field in the field of new energy vehicles. It involves multiple aspects such as battery monitoring, fault diagnosis, performance evaluation, and life prediction. The automotive BMS is the core of power battery data storage and management. It is responsible for real-time collection of battery voltage, current, temperature and other data, and processing and analyzing these data to ensure the safety and efficient operation of the battery. With the popularity of electric vehicles, a large amount of battery data is generated, which is eventually collected in the big data platform. Cloud computing provides a platform for data storage and processing, which can process and analyze these massive data to help optimize battery performance and life. Big data technology can be used to mine patterns and trends in data and provide decision support for battery research and development and maintenance. Power battery data storage management also needs to consider data security and privacy protection. Data encryption technology can ensure that stored data is not accessed by unauthorized persons, while security protocols can protect the security of data during transmission. However, how to store and utilize battery data under the premise of protecting privacy and security is the key problem that the present invention will solve.

[0003] Existing patents propose power battery data control methods from the perspectives of data storage, collection, and utilization. As the number of power batteries increases, the cascade utilization of power batteries will receive more and more attention, and the utilization of battery data will also provide a high-level intelligent control method for the recycling application of batteries.

[0004] The application and management of power batteries cannot be separated from the application of data. With the operation of new energy vehicles, a huge amount of battery data will be generated. The more data dimensions there are, the more accurate the prediction of the battery will be, resulting in a richer amount of stored data. Limited by the limitations of storage capacity and query speed, the increase in data volume will inevitably bring about storage and retrieval problems. Therefore, the storage, query and security protection of these data will bring new technical issues to the application of batteries.

[0005] With the rapid development of electric vehicles and energy storage systems, the amount of battery data has increased dramatically. Vector databases are able to process and analyze large-scale battery data sets. The performance and status of batteries can be efficiently represented in vector form. These vectors can capture multiple characteristics of batteries, such as charge and discharge curves, temperature changes, number of cycles, etc., thereby providing a rich data foundation for battery analysis and management, supporting the development and application of battery big data intelligent analysis platforms, helping R&D personnel to intuitively understand complex data sets and reveal patterns and relationships in the data. Vector databases can quickly retrieve the battery data vector that is most similar to a specific query vector, which is very useful for real-time monitoring of battery performance and fault diagnosis.

[0006] With the increase in new energy vehicles, battery operation data will increase exponentially. When the battery is recycled, it is not possible to obtain the first-hand operation data of the battery, so additional testing and verification are required, which requires a lot of time, equipment, personnel and other costs. Applying the power battery operation data to the development of battery recycling will save many links, improve efficiency and reduce energy consumption. At present, battery data is stored and managed by automobile manufacturers through the Internet of Vehicles data platform, which is very restrictive for the battery recycling of the entire society. Automobile manufacturers store a large amount of battery data and occupy a large amount of storage devices, which is also a huge burden for the data platform. Applying power battery operation data to commercial application development and giving full play to the value of data elements, data can only realize its value through circulation, so the query efficiency of data also directly affects its role. Summary of the invention

[0007] In view of the shortcomings of the prior art, the present invention proposes a power battery information storage and retrieval device and method with differential privacy computing and vectorized storage. Based on differential privacy computing technology, vectorized storage database and vectorized query technology, through the data storage and retrieval control of the power battery, precise control of the power battery cascade utilization is achieved, data support is provided for the battery life prediction management, and the three problems of battery data storage, retrieval and privacy security protection are solved.

[0008] The present invention is achieved through the following technical solutions:

[0009] A power battery information storage and retrieval device with differential privacy computing vectorized storage:

[0010] The power battery information storage and retrieval device includes a power battery data acquisition module, a first encryption protocol transmission module, a power battery data processing module, a second encryption protocol transmission module and an upper layer application module;

[0011] The power battery data acquisition module is responsible for collecting various raw data of the power battery, including voltage, current and temperature;

[0012] After data collection, the first encryption protocol transmission module encrypts the collected data and transmits the encrypted data to the power battery core processing module;

[0013] The power battery core processing module includes a power battery core processor, a data vectorization module, a differential privacy calculation module and a vectorization database, which are used for data vectorization, differential privacy processing and storing the processed data in the vector database;

[0014] The second encryption protocol transmission module encrypts the data to be transmitted again after the power battery core processing module processes the data;

[0015] The upper layer application module realizes battery power consumption balancing, temperature control, charging strategy optimization, or is used for battery life prediction and fault diagnosis based on the data received from the second encryption protocol transmission module.

[0016] Furthermore, the power battery core processor coordinates the work of the data vectorization module, the differential privacy calculation module and the vectorization database to control and manage the entire data processing process;

[0017] The data vectorization module decrypts the encrypted data received from the first encryption protocol transmission module and converts the original data into a vector form;

[0018] The differential privacy calculation module performs differential privacy processing on the vector data after the data vectorization module converts the data into vectors;

[0019] The vectorized database is used to store vectorized data processed by the differential privacy computing module and provide an indexing and retrieval mechanism.

[0020] A power battery information storage and retrieval method based on differential privacy computing vectorized storage:

[0021] The method specifically comprises the following steps:

[0022] Step 1: Collect power battery data and encrypt and transmit it;

[0023] Step 2: quantize the power battery data collected in step 1, perform differential privacy calculation and store it;

[0024] Step 3: Based on the vector database obtained in step 2, data retrieval and application are performed to achieve battery power consumption balancing, temperature control, charging strategy optimization, or for battery life prediction and fault diagnosis.

[0025] Furthermore, in step 1, raw data of the battery, including voltage, current, and temperature, is collected, and the collected data is encrypted through a first encryption protocol transmission module.

[0026] Further, in step 2, the encrypted data is first decrypted, and the raw data of the battery is converted into a vector form according to the data type;

[0027] Then, a differential privacy algorithm is applied to the vectorized data, the data is perturbed by adding Laplace noise or Gaussian noise, and the privacy budget parameters in the differential privacy algorithm are adjusted to balance the privacy protection and practicality of the data.

[0028] Finally, the vector data after differential privacy processing is stored in the vector database and an index is built.

[0029] Furthermore, in the vector database constructed in step 2, the distance between the query vector and each vector in the database is calculated through similarity measurement, and the data vector most similar to the query vector is quickly found through the index structure.

[0030] Further, in step 3, when a search request from a user is received, the request is first converted into a vector form, a search is performed in the vector database, the data vector most similar to the query vector is found, and a set of vectors with the smallest distance is returned as a result.

[0031] Furthermore, in step 3, privacy-preserving retrieval, approximate search, and multimodal and cross-modal retrieval are also included;

[0032] The privacy-preserving retrieval adds additional noise to the retrieval algorithm to protect the privacy during the retrieval process;

[0033] Use approximate search to speed up the retrieval process in large-scale datasets;

[0034] For retrieval requirements involving different types of data, different types of data are converted into representations in the same vector space, thereby achieving multimodal and cross-modal similarity search.

[0035] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0036] A computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by a processor, the steps of the above method are implemented.

[0037] Beneficial effects of the present invention

[0038] The present invention can efficiently store and simplify data:

[0039] Using vectorized storage technology, raw data (such as text, images, audio and other related battery data) is converted into vector representation in high-dimensional space through a specific algorithm. This vectorization method achieves maximum simplification of data and can more effectively utilize storage space and improve storage efficiency compared to traditional storage methods.

[0040] Vector databases use specific data structures (such as KD trees, ball trees, inverted indexes, etc.) to organize vector data. These structures provide support for efficient storage and can better cope with the storage needs of massive battery data.

[0041] The present invention can quickly search and accurately match:

[0042] Based on vectorized retrieval technology, it is possible to quickly retrieve the battery data vector that is most similar to a specific query vector. By measuring the similarity by the distance or angle between vectors, this retrieval method based on the vector space model is more efficient and accurate than traditional retrieval methods.

[0043] Common index structures (such as KD tree, ball tree, HNSW, etc.) are used to organize data, which improves retrieval efficiency and can quickly find the vector most similar to the query vector in large-scale data sets, meeting the needs of real-time query of battery data. For example, in real-time monitoring of battery performance and fault diagnosis scenarios, relevant data can be quickly obtained.

[0044] The present invention also supports multimodal and cross-modal retrieval:

[0045] Vectorized storage and retrieval can also support multimodal and cross-modal retrieval, that is, converting different types of data (such as battery-related text and image data) into representations in the same vector space, thereby enabling similarity searches across different data types. This is difficult to achieve in existing technologies and provides a more comprehensive means for the comprehensive analysis and utilization of battery data.

[0046] The present invention has strong privacy protection capabilities:

[0047] Differential privacy technology is used to encrypt data and add noise to the data or its query results to ensure that changes in a single data point will not significantly affect the overall output results, thereby protecting individual information from being leaked.

[0048] Differential privacy technology is applied in all stages of data storage and retrieval (including vectorized data processing, vector database construction, retrieval process and result output), which comprehensively protects the privacy of battery data and ensures the privacy security of the data usage process. This significantly improves the comprehensiveness and effectiveness of battery data privacy protection in existing technologies.

[0049] The present invention improves battery management efficiency and performance:

[0050] The performance and status of the battery can be efficiently represented in vector form. These vectors can capture various characteristics of the battery, such as charge and discharge curves, temperature changes, number of cycles, etc., providing a rich data foundation for battery analysis and management.

[0051] It supports the development and application of battery big data intelligent analysis platforms, helps researchers to intuitively understand complex data sets and reveal patterns and relationships in the data, thereby optimizing battery maintenance plans and replacement strategies, improving battery efficiency and safety, and achieving intelligent management of battery packs, including battery balancing, temperature control, and charging strategy optimization. Compared with existing technologies, it has made great progress in the intelligence and refinement of battery management.

[0052] To summarize, the present invention proposes an integrated system device and control method for storing, retrieving and protecting vehicle-mounted battery data based on vectorized storage and search methods combined with differential privacy technology. Vectorized storage can maximize the simplification and rapid retrieval of data, and differential privacy technology is used to encrypt data, which can effectively protect the privacy of battery data and ensure the privacy security of the data use process. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 This is a graph indexing method based on HNSW graph vector retrieval of the present invention.

[0054] Figure 2 Schematic diagram of the data vectorization principle of the present invention.

[0055] Figure 3 This is a schematic diagram of the differential privacy vectorized data storage and retrieval device system of the present invention.

[0056] Figure 4 This is a block diagram of the differential privacy vectorized data storage and retrieval device system of the present invention. DETAILED DESCRIPTION

[0057] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] The experimental methods used in the following examples are conventional methods unless otherwise specified. The materials, reagents, methods and instruments used are conventional materials, reagents, methods and instruments in the art unless otherwise specified, and can be obtained through commercial channels by those skilled in the art.

[0059] Combination Figures 1 to 4 ; This invention designs a differential privacy computing vectorized storage information storage and retrieval method:

[0060] Step 1, data preparation stage, data vectorization:

[0061] Various types of battery-related data (such as text, images, audio, etc.) are converted into numerical vectors through specific algorithms (for example, text data is converted into vectors through word embedding (such as Word2Vec, BERT); image data is extracted with feature vectors using convolutional neural networks CNN, etc.); these vectors can capture the key features of the data for subsequent storage and retrieval.

[0062] The principle of vectorized storage and retrieval is based on the Vector Space Model (VSM), which is a commonly used technology in information retrieval and machine learning. In this model, data such as text, images, and audio are converted into points in a high-dimensional space, namely vectors. These vectors represent the characteristics of the data through their coordinate values, and the distance or angle between vectors is used to measure the similarity of the data.

[0063] like Figure 2 The figure shows a schematic diagram of the principle of vectorized data: consisting of three coordinates (x, y, z), efficient data management and query can be achieved while protecting data privacy.

[0064] Step 2: In the data storage phase, the vectorized data in step 1 is differentially privately processed, then stored and indexed to facilitate retrieval;

[0065] Data vectorization storage:

[0066] Using feature extraction techniques or deep learning models, raw data is converted into numerical vectors and stored in databases, which are specially designed to handle high-dimensional vector data and provide efficient storage and indexing mechanisms. Vector databases use specific data structures (such as KD trees, ball trees, inverted indexes, etc.) to organize vector data for fast retrieval.

[0067] Reduce the set of candidate vectors: Similar to traditional text retrieval, vector retrieval also requires a certain index structure to avoid matching on the entire data. Traditional text retrieval uses an inverted index to filter out irrelevant documents, while vector retrieval bypasses irrelevant vectors to improve retrieval efficiency by establishing an index structure for vectors.

[0068] Differential privacy processing:

[0069] Apply differential privacy algorithms to vectorized data, such as perturbing the data by adding Laplace noise or Gaussian noise. Adjust the privacy budget parameters (such as ε) in the differential privacy algorithm to balance the privacy protection and practicality of the data.

[0070] Vector database construction:

[0071] Store the vector data after differential privacy processing in the vector database. Build indexes, such as using KD trees and inverted indexes, to optimize retrieval efficiency.

[0072] Reduce the complexity of single vector calculation: Traditional text retrieval usually adopts the funnel model when sorting. The upper layer calculation is relatively light, and the calculation becomes more and more sophisticated as the filtering progresses. However, as the filtering progresses, the number of documents that need to be calculated also decreases step by step. For high-dimensional vectors, the amount of calculation for a single vector is still relatively heavy. Usually, the vector is quantized and approximate calculations are performed on the vector, and finally the original vector is sorted on a very small data set.

[0073] Commonly used index structures can be classified as methods based on space partitioning. Each vector will only belong to a certain partitioned area. The biggest problem with these methods is that in order to improve recall, it is necessary to examine vectors in a large area, which leads to a surge in calculations. Graph-based methods can achieve this goal relatively well. The simplest idea of ​​graph methods is that the neighbors of neighbors are also neighbors. In this way, the nearest neighbor search is converted into a graph traversal. Due to its connectivity, some vectors can be examined in a targeted manner instead of by region, so the scope of vector examination can be greatly reduced. Based on the HNSW graph vector retrieval graph index method, the steps are: a. Select the entry point; b. Traverse the graph; c. Converge. The principle of the HNSW graph method, behind HNSW is actually the application of skip lists on graphs. Skip lists as a simple and efficient index structure are widely used in Redis, LevelDB and other systems, such as Figure 1 shown.

[0074] Step 3, in the data retrieval phase, query vectors are retrieved and relevant results are processed based on the standard metric vector similarity.

[0075] The similarity measure:

[0076] In a vector database, determine the distance metric on which the similarity search is based, such as Euclidean distance (L2 distance), cosine similarity, or Manhattan distance. These metrics define how to compare the similarity of two vectors in vector space.

[0077] The index and search:

[0078] The vector database builds indexes (such as KD tree, ball tree, HNSW, etc.) to organize data to support fast query. It receives retrieval requests from users and converts the requests into vector form. It performs retrieval in the vector database to find the data vector that is most similar to the query vector.

[0079] The query process:

[0080] When a user submits a query, the system first converts the query into a vector representation (if it is not already in vector form). The system then uses the index to quickly find the vectors that are most similar to the query vector. This typically involves calculating the distance between the query vector and each vector in the database, and returning the set of vectors with the smallest distance as the result.

[0081] The privacy-preserving search:

[0082] Apply differential privacy techniques, such as adding additional noise to the retrieval algorithm, to protect privacy during the retrieval process. Ensure the privacy of retrieval results and prevent the leakage of sensitive information.

[0083] The approximate search and optimization:

[0084] In large-scale datasets, accurate search is very time-consuming. Therefore, vector databases usually use approximate search algorithms (such as locality sensitive hashing (LSH) or product quantization) to speed up the retrieval process. These methods sacrifice a certain degree of accuracy in exchange for faster retrieval speed.

[0085] The multimodal and cross-modal retrieval:

[0086] For retrieval requirements involving different types of data (such as battery-related text and images), different types of data are converted into representations in the same vector space, thereby enabling similarity search across different data types.

[0087] Based on the above content, the present invention proposes a power battery information storage and retrieval device with differential privacy computing vectorized storage:

[0088] The power battery information storage and retrieval device includes a power battery data acquisition module, a first encryption protocol transmission module, a power battery data processing module, a second encryption protocol transmission module and an upper layer application module;

[0089] The power battery data acquisition module is responsible for collecting various raw data of the power battery, including voltage, current, temperature and other information. These data are the basic data source for the entire system to store and retrieve information on the power battery;

[0090] After data collection, the first encryption protocol transmission module encrypts the collected data to ensure the security and privacy of the data during transmission, and transmits the encrypted data to the power battery core processing module including core processing and other functions;

[0091] The power battery core processing module includes a power battery core processor, a data vectorization module, a differential privacy calculation module and a vectorization database, which are used for data vectorization, differential privacy processing and storing the processed data in the vector database;

[0092] The power battery core processor

[0093] Function: As the core control unit of this large module, it coordinates the work of the other three small modules (data vectorization module, differential privacy calculation module, and vectorized database). It controls and manages the entire data processing process, such as determining the flow of data, the processing order, and the parameter settings of each module.

[0094] Data vectorization module

[0095] Function: After decrypting the encrypted data received from the encryption protocol transmission module (if necessary), convert the original data into vector form. Through specific algorithms and techniques, different types of data (such as the battery data collected above) are converted into numerical vectors for subsequent storage and retrieval operations.

[0096] Differential privacy computing module

[0097] Function: After the data vectorization module converts the data into vectors, differential privacy processing is performed on the vector data. The privacy of the data is protected by adding noise and other methods to ensure that individual information is not leaked during data storage and retrieval. At the same time, the amount of noise added will be adjusted according to the set privacy budget parameters to balance the availability of data and the degree of privacy protection.

[0098] Vectorized Database

[0099] Function: Used to store vectorized data processed by the differential privacy computing module. This database is specially designed to process high-dimensional vector data and provide efficient indexing and retrieval mechanisms. It is a key component for fast retrieval and data storage, and can quickly find the data vector most similar to the query vector based on the user's retrieval request.

[0100] The second encryption protocol transmission module encrypts the data to be transmitted again after the power battery core processing module processes the data, thereby ensuring the security and privacy of the data during transmission to the upper-level application module.

[0101] The upper application module develops various applications based on the data received from the second encryption protocol transmission module (power battery information after a series of previous processing). For example, applications in the battery management system (BMS) are used to implement functions such as battery balancing, temperature control, and charging strategy optimization, or programs for application scenarios such as battery life prediction and fault diagnosis. These applications directly serve the management and maintenance of power batteries, and use the stored and retrieved information to provide users with valuable services and decision support.

[0102] The present invention also proposes a power battery information storage and retrieval method with differential privacy computing vectorized storage:

[0103] The method specifically comprises the following steps:

[0104] Step 1: Collect power battery data and encrypt and transmit it;

[0105] In step 1, the original data of the battery, including voltage, current, and temperature, is collected, and the collected data is encrypted through a first encryption protocol transmission module.

[0106] Step 2: quantize the power battery data collected in step 1, perform differential privacy calculation and store it;

[0107] In step 2, the encrypted data is first decrypted, and the raw data of the battery is converted into a vector form according to the data type;

[0108] Then, a differential privacy algorithm is applied to the vectorized data, the data is perturbed by adding Laplace noise or Gaussian noise, and the privacy budget parameters in the differential privacy algorithm are adjusted to balance the privacy protection and practicality of the data.

[0109] Finally, the vector data after differential privacy processing is stored in the vector database and an index is constructed, such as using KD tree, ball tree, inverted index, etc., to optimize the retrieval efficiency.

[0110] In the vector database constructed in step 2, the distance between the query vector and each vector in the database is calculated through similarity measurement. The data vector most similar to the query vector is quickly found through the index structure;

[0111] Step 3: Based on the vector database obtained in step 2, data retrieval and application are performed to achieve battery power consumption balancing, temperature control, charging strategy optimization, or for battery life prediction and fault diagnosis.

[0112] In step 3, when a search request is received from a user, the request is first converted into a vector form, a search is performed in the vector database, the data vector most similar to the query vector is found, and a set of vectors with the smallest distance is returned as a result.

[0113] In step 3, privacy-preserving retrieval, approximate search, and multimodal and cross-modal retrieval are also included;

[0114] The privacy-preserving retrieval adds additional noise to the retrieval algorithm to protect the privacy during the retrieval process;

[0115] Use approximate search to speed up the retrieval process in large-scale datasets;

[0116] For retrieval requirements involving different types of data, different types of data are converted into representations in the same vector space, thereby achieving multimodal and cross-modal similarity search.

[0117] An electronic device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0118] A computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by a processor, the steps of the above method are implemented.

[0119] The memory in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory, ROM, a programmable read-only memory, PROM, an erasable programmable read-only memory, EPROM, an electrically erasable programmable read-only memory, EEPROM, or a flash memory. The volatile memory may be a random access memory, RAM, which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory static RAM, SRAM, dynamic random access memory dynamic RAM, DRAM, synchronous dynamic random access memory synchronous DRAM, SDRAM, double data rate synchronous dynamic random access memory double data rate SDRAM, DDR SDRAM, enhanced synchronous dynamic random access memory enhanced SDRAM, ESDRAM, synchronous link dynamic random access memory synchlink DRAM, SLDRAM, and direct memory bus random access memory direct rambus RAM, DR RAM. It should be noted that memory of the methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0120] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center through a wired method such as coaxial cable, optical fiber, digital subscriber line digital subscriber line, DSL or wireless such as infrared, wireless, microwave, etc. to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media integrated. The available medium can be a magnetic medium such as a floppy disk, a hard disk, a tape, an optical medium such as a high-density digital video disc digital video disc, DVD, or a semiconductor medium such as a solid state hard disk solid state disc, SSD, etc.

[0121] In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in a processor or an instruction in the form of software. The steps of the method disclosed in conjunction with the embodiment of the present application can be directly embodied as a hardware processor for execution, or a combination of hardware and software modules in a processor for execution. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in a memory, and the processor reads the information in the memory and completes the steps of the above method in conjunction with its hardware. To avoid repetition, it is not described in detail here.

[0122] It should be noted that the processor in the embodiment of the present application can be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method embodiment can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor can be a general-purpose processor, a digital signal processor DSP, an application-specific integrated circuit ASIC, a field programmable gate array FPGA or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiment of the present application can be directly embodied as a hardware decoding processor to perform, or the hardware and software modules in the decoding processor can be combined to perform. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in a memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0123] The above is a detailed introduction to the power battery information storage and retrieval device and method for differential privacy computing vectorized storage proposed in the present invention, and the principles and implementation methods of the present invention are explained. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A power battery information storage and retrieval device with differential privacy computing vectorized storage, characterized in that: The power battery information storage and retrieval device includes a power battery data acquisition module, a first encryption protocol transmission module, a power battery data processing module, a second encryption protocol transmission module and an upper layer application module; The power battery data acquisition module is responsible for collecting various raw data of the power battery, including voltage, current and temperature; After data collection, the first encryption protocol transmission module encrypts the collected data and transmits the encrypted data to the power battery core processing module; The power battery core processing module includes a power battery core processor, a data vectorization module, a differential privacy calculation module and a vectorization database, which are used for data vectorization, differential privacy processing and storing the processed data in the vector database; The second encryption protocol transmission module encrypts the data to be transmitted again after the power battery core processing module processes the data; The upper layer application module realizes battery power consumption balancing, temperature control, charging strategy optimization, or is used for battery life prediction and fault diagnosis based on the data received from the second encryption protocol transmission module.

2. The power battery information storage and retrieval device according to claim 1, characterized in that: The power battery core processor coordinates the work of the data vectorization module, the differential privacy calculation module and the vectorization database, and controls and manages the entire data processing process; The data vectorization module decrypts the encrypted data received from the first encryption protocol transmission module and converts the original data into a vector form; The differential privacy calculation module performs differential privacy processing on the vector data after the data vectorization module converts the data into vectors; The vectorized database is used to store vectorized data processed by the differential privacy computing module and provide an indexing and retrieval mechanism.

3. A storage and retrieval method of a power battery information storage and retrieval device based on the differential privacy computing vectorized storage according to claim 1 or 2, characterized in that: The method specifically comprises the following steps: Step 1: Collect power battery data and encrypt and transmit it; Step 2: quantize the power battery data collected in step 1, perform differential privacy calculation and store it; Step 3: Based on the vector database obtained in step 2, data retrieval and application are performed to achieve battery power consumption balancing, temperature control, charging strategy optimization, or for battery life prediction and fault diagnosis.

4. The method for storing and retrieving power battery information according to claim 3, characterized in that: In step 1, the original data of the battery, including voltage, current, and temperature, is collected, and the collected data is encrypted through a first encryption protocol transmission module.

5. The method for storing and retrieving power battery information according to claim 4, characterized in that: In step 2, the encrypted data is first decrypted, and the raw data of the battery is converted into a vector form according to the data type; Then, a differential privacy algorithm is applied to the vectorized data. The data is perturbed by adding Laplace noise or Gaussian noise, and the privacy budget parameters in the differential privacy algorithm are adjusted to balance the privacy protection and practicality of the data. Finally, the vector data after differential privacy processing is stored in the vector database and an index is built.

6. The method for storing and retrieving power battery information according to claim 5, characterized in that: In the vector database constructed in step 2, the distance between the query vector and each vector in the database is calculated through similarity measurement, and the data vector most similar to the query vector is quickly found through the index structure.

7. The method for storing and retrieving power battery information according to claim 6, characterized in that: In step 3, when a search request is received from a user, the request is first converted into a vector form, a search is performed in the vector database, the data vector most similar to the query vector is found, and a set of vectors with the smallest distance is returned as a result.

8. The power battery information storage and retrieval method according to claim 7, characterized in that: In step 3, privacy-preserving retrieval, approximate search, and multimodal and cross-modal retrieval are also included; The privacy-preserving retrieval adds additional noise to the retrieval algorithm to protect the privacy during the retrieval process; Use approximate search to speed up the retrieval process in large-scale datasets; For retrieval requirements involving different types of data, different types of data are converted into representations in the same vector space, thereby achieving multimodal and cross-modal similarity search.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method described in any one of claims 3 to 8 are implemented.

10. A computer-readable storage medium for storing computer instructions, characterized in that: When the computer instructions are executed by a processor, the steps of the method described in any one of claims 3 to 8 are implemented.

Images