Data processing method and system based on artificial intelligence, storage device and storage medium

Abstract

The invention discloses a data processing method and system based on artificial intelligence, a storage device and a storage medium, and relates to the technical field of storage. Inputting the data packet into a data processing model to generate a feature vector and a label so as to form structured metadata; globally unifying identifiers, organizing the feature vectors and the metadata into index structures, and mapping the feature vectors to an identifier list by each index structure; inputting the index structure into a data security verification model, and processing an abnormal data packet; inputting the index structure into a data processing model, and storing the index structure, the identifier of the original data item and the corresponding metadata in an index database; by accessing the index database to obtain the structured result and displaying the structured result to the user, the problem that the existing storage device is lack of intelligence is solved, and automatic classification, quick retrieval, secure encryption and efficient backup are realized.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of storage, in particular to a data processing method and system based on artificial intelligence, a storage device and a storage medium. BACKGROUND

[0002] The existing data storage devices such as U disks, mobile hard disks, hard disk boxes and NASs only provide passive data storage functions and lack intelligent data processing capabilities.

[0003] With the growth of big data and multimedia file types, users encounter the following problems in data retrieval, classification, backup and security protection, for example: when sorting files, users need to classify through artificial sorting, resulting in low efficiency of user file sorting and arrangement operations; secondly, in the file retrieval process, the overall speed is slow, and users are difficult to quickly and accurately locate the target file; moreover, there is a lack of intelligent encryption and privacy protection measures for data, if malicious files exist in the to-be-stored files and cannot be detected and processed in time, there is a risk of leakage of the files in the storage device during the storage process of new files; in addition, the storage device cannot support intelligent storage requirements such as incremental backup and repeated file detection, and the repeated files occupy redundant storage space in the storage device, which is not suitable for an environment with tight storage resources and is not conducive to the rational use of storage space. SUMMARY

[0004] In view of the technical defects in the background art, the application provides a data processing method and system based on artificial intelligence, a storage device and a storage medium, which solve the above technical problems and meet the actual needs, and the specific technical solutions are as follows: The application discloses a data processing method based on artificial intelligence, comprising the following steps: S1: receiving to-be-stored data with a plurality of original data items from an external device, and generating a plurality of data packets corresponding to the original data items after data preprocessing of the to-be-stored data; S2: inputting the data packets into a data processing model based on artificial intelligence and pre-trained to determine recognition and analysis, generating one or more feature vectors and one or more labels, and encapsulating the feature vectors and labels and inherent attributes extracted from the original data items to form structured metadata; S3: according to a predefined business rule and a data structure, performing global uniform identifier on an index database, and organizing the feature vectors and metadata into one or more index structures, and each index structure maps one or more feature vectors to an identifier list of original data items containing the feature vectors; S4: Input the index structure into the data processing model, and based on the processing result of the data processing model, persistently write the one or more index structures, the identifier of the original data item, and the corresponding metadata into the index database. S5: Detect data packets written to the index database using a data security verification model pre-trained based on artificial intelligence; process abnormal data packets according to the verification results of the data security verification model; and support at least one data encryption operation among encrypted storage, face recognition, and fingerprint recognition unlocking for the data packets. S6: The user obtains structured results by accessing the index database and displays them to the user.

[0005] As a further embodiment of the invention, in S3, the identifier of the original data item is generated using the SHA-1 hash algorithm, the MD5 hash algorithm, or the SHA-256 hash algorithm.

[0006] As a further embodiment of the invention, in S4, the index database scans the data to be stored and queries the database to see if the identifier of the original data item exists; If the identifier already exists, the original data item is determined to be duplicate data, the physical storage operation is skipped, and a reference to the stored data is returned; If the hash identifier does not exist, the original data is written to the index database.

[0007] As a further embodiment of the invention, in S6, the user inputs a natural language query through the Web UI or App, the query is converted into a semantic vector, and matched with the feature vector of the metadata stored in the index database. The query is then sorted based on semantic similarity and keywords, and the matching file path, title, and category are returned and displayed to the user.

[0008] This invention also discloses an artificial intelligence-based data processing system, which uses the above-described method for data processing, including: An interface module connects to an external device and receives data to be stored from the external device. The AI ​​processing module is connected to the interface module, receives the data to be stored transmitted from the interface module, and performs file recognition, classification, OCR, face recognition and semantic analysis. The storage module is connected to the AI ​​processing module to achieve data communication, and performs data processing operations on the data to be stored based on the data processing model to generate the index structure, identifier and metadata for subsequent search use; The security module is connected to the AI ​​processing module to achieve interaction, and performs data processing operations such as malicious file detection, data encryption, and facial recognition or fingerprint recognition unlocking on the data to be stored based on the data security verification model. The data indexing module is connected to the AI ​​processing module, receives data transmitted from the AI ​​processing module, and stores the file tags, metadata, and semantic vectors corresponding to the original data items in the index database based on SQLite or an embedded database; The user interaction module connects with the data index module to achieve data exchange. It allows users to access the data index module by inputting keywords and semantic search, quickly locate files in the index database, and display the results to the user.

[0009] The present invention also discloses a storage device, which includes a data transmission interface, a memory, a processor, and the aforementioned system. A plurality of the data transmission interfaces are connected to the processor. The processor includes an AI processing module, a security module, and a user interaction module. The processor is connected to the memory, which includes a storage module and a data indexing module and stores a computer program. When the processor executes the computer program, it implements the method described above.

[0010] As a further embodiment of the invention, the processor has an integrated NPU or AI SoC and adopts the UFS communication protocol standard.

[0011] As a further embodiment of the invention, the data transmission interface includes a USB 3.2 interface, a USB 4 interface, a Thunderbolt interface, a PCIe interface, and an Ethernet interface.

[0012] As a further embodiment of the invention, the memory includes at least one of HDD, SSD and NVMe SSD.

[0013] The present invention also discloses a storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0014] The superior effects of this invention are as follows: The data processing method of this invention includes generating data packets from the data to be stored; inputting the data packets into a data processing model to generate feature vectors and tags to form structured metadata; performing globally unified identifiers and organizing the feature vectors and metadata into an index structure, with each index structure mapping the feature vectors to an identifier list; inputting the index structure into a data security verification model to process abnormal data packets; inputting the index structure into a data processing model to store the index structure, the identifiers of the original data items, and the corresponding metadata in an index database; obtaining structured results by accessing the index database and displaying them to the user. Compared with the prior art, this invention can automatically complete the intelligent classification and tagging of files, improve file retrieval efficiency, provide AI-driven security protection and data encryption, save storage space through incremental backup and duplicate detection, and offer high flexibility in use. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating the data processing method of the present invention.

[0016] Figure 2 This is a flowchart illustrating the process of using the SHA-256 hash algorithm for data deduplication in S4 of the data processing method.

[0017] Figure 3 This is a flowchart illustrating the user access and result display process in the S6 data processing method.

[0018] Figure 4 This is a schematic diagram of the data processing system in this invention.

[0019] Figure 5 This is a schematic diagram of the storage device in this invention.

[0020] In the diagram, 1 is the interface module; 2 is the AI ​​processing module; 3 is the storage module; 4 is the security module; 5 is the data indexing module; 6 is the user interaction module; 10 is the data transmission interface; 20 is the memory; 30 is the processor; and 100 is the storage device. Detailed Implementation

[0021] The embodiments of the present invention will now be described in conjunction with the accompanying drawings and related examples: This invention discloses a data processing method based on artificial intelligence, such as... Figure 1 As shown, it includes the following steps: S1: Receive data to be stored from an external device containing multiple raw data items. The data to be stored is preprocessed to generate multiple data packets corresponding to the raw data items. S2: The data packet is input into a data processing model pre-trained based on artificial intelligence for identification and analysis, generating one or more feature vectors and one or more labels. The feature vectors, labels, and inherent attributes extracted from the original data items are encapsulated to form structured metadata. Common data packet types include image data and audio data. For image data objects, the data processing model used is a convolutional neural network model (such as VGG, ResNet, Inception), using its extracted visual feature vectors as the feature vectors of the data packet. For audio data objects, the data processing model used is an audio processing model (such as Whisper, Wav2Vec2.0, ECAPA-TDNN, x-vector, HuBERT), using its extracted acoustic feature vectors as the feature vectors of the data packet. S3: Based on predefined business rules and data structures, globally unify the identifiers of the index database, and organize the feature vectors and metadata into one or more index structures. Each index structure maps one or more feature vectors to a list of identifiers of the original data items containing the feature vectors. S4: Input the index structure into the data processing model, and based on the processing result of the data processing model, persistently write the one or more index structures, the identifier of the original data item, and the corresponding metadata into the index database. S5: The data packets written to the index database are detected using a pre-trained data security verification model based on artificial intelligence. Based on the verification results of the data security verification model, abnormal data packets are processed, and at least one data encryption operation is supported, including encrypted storage, facial recognition, and fingerprint recognition unlocking. The data security verification model used is a fully connected neural network, convolutional neural network, recurrent neural network / long short-term memory network, or Transformer model. The model output score is compared with a preset threshold to determine whether the data packet is abnormal or not. Abnormal data packets are blocked, and the index structure, original data item identifiers, and corresponding metadata of the blocked data packets are not written to the index database.

[0022] S6: The user obtains structured results by accessing the index database and displays them to the user.

[0023] Compared with existing technologies, the present invention has the following advantages: it automatically completes intelligent classification and tagging of files, improves file retrieval efficiency, and provides AI-driven security protection and data encryption protection for file storage.

[0024] Further explanation of the above method: In S3, the identifier of the original data item is generated using the SHA-1 hash algorithm, the MD5 hash algorithm, or the SHA-256 hash algorithm.

[0025] The process involves using a hash algorithm (such as MD5, SHA-1, or SHA-256) to calculate an identifier for the incoming file. This identifier is then applied to subsequent S5 file verification. Database matching involves comparing this hash value with known malware hash databases (such as VirusTotal's database or the vendor's threat intelligence database). If the hash value is found in the malware database, it's considered a match, immediately identifying the file as malicious. The corresponding data packets within the file are then blocked, isolated, or deleted. If the hash value is not found in the malware database, it's considered a miss, and the data security verification model continues to the next stage of analysis. The advantages of this file verification method are: extremely high speed (the comparison process is a simple string query with minimal computational overhead); high accuracy (once a match is found, the file is confirmed as known malware with zero false positives); and low resource consumption, making it ideal for first-round filtering in resource-constrained environments such as gateways and terminals.

[0026] Specifically, such as Figure 2 As shown, in S4, the index database scans the data to be stored and queries the database to see if the identifier of the original data item exists; If the identifier already exists, the original data item is determined to be duplicate data, the physical storage operation is skipped, and a reference to the stored data is returned; If the hash identifier does not exist, the original data is written to the index database.

[0027] If the SHA-256 hash algorithm is used, the specific steps are as follows: S10. Let the data packet result be: [Chunk_1, Chunk_2, ..., Chunk_N]; S20. Calculate the hash identifier of each data packet in parallel. For each data packet Chunk_i, calculate its hash value Hash_Chunk_i in parallel. Hash_Chunk_i = SHA-256(Chunk_i). S30. Match the hashes of each data packet in the index database, construct a list containing all Hash_Chunk_i, and perform a batch query in the index database; S40. Identify unique data packets and duplicate data packets, and classify the data packets into two categories based on the query results: Unique data packets: Data packets whose Hash_Chunk_i does not exist in the index; Duplicate packets: packets whose Hash_Chunk_i already exists in the index; S50. Store unique data packets and update the index. Only write the identified unique blocks to the physical storage and obtain their storage locations. Then, batch insert the (Hash_Chunk_i, Location_i) pairs of these unique data packets into the index database.

[0028] More specifically, such as Figure 3 As shown, in S6, the user inputs a natural language query through the Web UI or App, which is converted into a semantic vector and matched with the feature vector of the metadata stored in the index database. The query is then sorted based on semantic similarity and keywords, and the matching file path, title, and category are returned and displayed to the user.

[0029] A further implementation of the above method specifically includes the following steps: S100. The user enters a natural language query, such as "summary of last year's financial audit report", through the search box of the web user interface or mobile application. By receiving the query string, the data processing model is called to convert the natural language query into a query semantic vector. The data processing model used here is a pre-trained semantic model. This vector is located in the same vector space as the feature vector in the index database, thus making semantic similarity comparison possible. S200: Calculate the similarity between the obtained query semantic vector and the feature vectors of all files in the index database. The preferred similarity metric is cosine similarity, resulting in a similarity score between 0 and 1. S300: The system generates a final comprehensive relevance score by weighting and combining semantic similarity scores and keyword matching scores; for example, the comprehensive score = 0.7 * semantic similarity score + 0.3 * keyword matching score. This strategy ensures the core position of semantic matching while leveraging the advantage of keyword matching in precise terminology, thus improving the robustness of the ranking. S400: The system sorts all matching files in descending order based on their comprehensive relevance scores and selects the TopK (K=10) most relevant results from them; The S500 results presentation module extracts the file path, title, and category information for each file from the sorted results list. The system then displays this information to the user in a clear, user-friendly format, such as a list or card, on the web UI or app search results page. Each result item can be clicked directly to open or download the corresponding file via its file path.

[0030] This invention also discloses a data processing system based on artificial intelligence, such as... Figure 4 As shown, the system uses the above-described method for data processing, including: Interface module 1 connects to external devices and receives data to be stored from the external devices; it serves as a hub for connecting the system to external devices such as computers, mobile phones, cameras, and scanners, thereby receiving file data of various formats such as documents, pictures, audio, and video from external devices.

[0031] AI processing module 2 is connected to interface module 1, receives data to be stored transmitted from interface module 1, and performs file recognition, classification, OCR, face recognition, and semantic analysis. This module is the core analysis engine of the system and is specifically configured to perform one or more of the following processes, including automatically recognizing document types such as contracts, invoices, and resumes and classifying them according to their content, extracting text information from images or scanned documents, recognizing and analyzing facial features in images or videos, and understanding the theme, keywords, sentiment, and core semantics of text data through natural language processing technology.

[0032] Storage module 3 is connected to AI processing module 2 to achieve data communication, and performs data processing operations on the data to be stored based on the data processing model to generate the index structure, identifier and metadata for subsequent search; this module performs data modeling operations based on the analysis results generated by the data processing model of AI processing module 2 to generate the index structure, file identifier and metadata for subsequent fast retrieval. Security module 4 connects to and interacts with AI processing module 2, and performs data processing operations on the data to be stored based on the data security verification model, including malicious file detection, data encryption, and facial recognition or fingerprint recognition unlocking. Based on the preset data security verification model, this module performs one or more of the following operations: including scanning the received file for security threats such as viruses and Trojans, encrypting the data in storage module 3 to prevent unauthorized access, and integrating facial recognition or fingerprint recognition technology as a means of verifying system access permissions to ensure that only authorized users can access the data. The data indexing module 5 is connected to the AI ​​processing module 2, receives data transmitted from the AI ​​processing module 2, and stores the file tags, metadata, and semantic vectors corresponding to the original data items in the index database based on SQLite or an embedded database. This module is used to build an efficient data index and establish an index database based on a lightweight database. The content stored in the index database includes file tags generated by AI analysis, file metadata, and semantic vectors of text content. User interaction module 6 connects to the data index module 5 to achieve data communication. It supports users to access the data index module 5 by inputting keywords and semantic search, quickly locate files in the index database, and display the results to the user. This module provides users with an intuitive search interface and supports users to access the data index module 5 to quickly locate files in the index database in two main ways: first, through keyword search: the user inputs specific keywords, and the system quickly matches file tags or OCR text containing the keywords through the inverted index; second, through semantic search: the user inputs a descriptive statement, such as "find a contract about cooperation in artificial intelligence technology," and the system converts the user's query into a semantic vector, searches the index database for the file with the closest semantic vector, and finally displays the search results, such as file list and file preview, to the user clearly and orderly.

[0033] It should be noted that: the data to be stored enters the system through interface module 1, where it first undergoes in-depth content analysis by AI processing module 2, while security module 4 performs security verification in parallel. The results obtained from the analysis by AI processing module 2 are sent to storage module 3 to generate an index structure and metadata, and also to data index module 5 to be stored in the embedded index database. When a user initiates a search through user interaction module 6, the query request is sent to data index module 5, which uses a pre-built index structure and semantic vectors for rapid matching and sorting, ultimately returning and displaying the most relevant results to the user. This display can be done through the built-in screen or an external APP. This embodiment, through the collaborative work of the above modules, achieves automated and intelligent management of massive amounts of unstructured data, not only enhancing retrieval capabilities but also deeply integrating security mechanisms into the data processing flow, realizing the integration of data processing, security protection, and intelligent retrieval, and significantly improving the efficiency and security of data management.

[0034] The present invention also discloses a storage device, such as Figure 5 As shown, the storage device 100 includes a data transmission interface 10, a memory 20, a processor 30, and the aforementioned system. Several of the data transmission interfaces 10 are connected to the processor 30. The processor 30 includes an AI processing module 2, a security module 4, and a user interaction module 6. The processor 30 is connected to the memory. The memory 20 includes a storage module 3 and a data indexing module 5 and stores a computer program. When the processor 30 executes the computer program, it implements the method described above.

[0035] It should be noted that: the data transmission interface is used to connect with external devices such as computers and mobile phones to receive data transmitted from these devices. The received data is transmitted to the processor to execute artificial intelligence model inference tasks, such as image recognition, speech-to-text, and natural language processing. It also performs real-time, transparent hardware-level encryption (such as using the AES-256 algorithm) on the data written to the storage device and strictly authenticates device access permissions to prevent unauthorized access and data leakage. The user interaction module 6 manages the status indicators such as LED lights, small displays, and / or voice prompt units on the device. It is used to intuitively display the device's working status to the user, such as prompts like "AI processing in progress" and "encryption complete," storage capacity, network connection status, or interactive confirmation via voice / screen. For example, when storing photos, the AI ​​processing module 2 identifies the content in the photo and generates structured descriptive tags. These tags, along with metadata such as the photo's shooting time and geographical location, are sent to the data index module 5. The data index module 5 organizes these metadata and tags into an efficient and quickly queryable database and associates it with the original media file. This index database is also stored in a specific area of ​​the storage module 3.

[0036] This storage device enables "on-demand data processing" and near real-time intelligent retrieval of massive amounts of data through local AI processing and data indexing, greatly improving data utilization efficiency. Hardware-level encryption and security authentication ensure data security during static storage and transmission, effectively resisting external attacks. It integrates computing, storage, and security functions, reducing dependence on external computing resources. All sensitive data is processed on the device itself, avoiding the privacy leakage risks associated with cloud transmission. The intuitive user interaction module 6 allows users to clearly understand the device status, and intelligent data management reduces the burden of manually organizing and searching data.

[0037] The processor has a built-in NPU or AI SoC and adopts the UFS communication protocol standard. Chips using the UFS communication protocol standard as processors can achieve efficient data transmission, greatly improve multi-tasking capabilities and application loading experience, and have fast sequential and random read / write speeds. The processor has a built-in dedicated neural network processing unit or adopts a chip system that integrates multiple functional units, which has powerful fixed-point and high-precision floating-point arithmetic capabilities, and is used to execute artificial intelligence model inference tasks locally on the device, such as image recognition, speech-to-text, and natural language processing, without relying on cloud computing power.

[0038] Secondly, the data transmission interfaces include USB 3.2, USB 4, Thunderbolt, PCIe, and Ethernet interfaces. By integrating multiple mainstream and high-performance data transmission interfaces into a single device, it provides users with significant connectivity flexibility and convenience. This design can meet the high-speed data transmission needs of various application scenarios, from daily office work to professional content creation and high-performance computing, effectively avoiding performance bottlenecks and scalability issues caused by insufficient interface types. Furthermore, the collaborative operation between the interfaces maximizes the overall data throughput of the system.

[0039] Furthermore, the memory includes at least one of HDD, SSD, and NVMe SSD; this storage device incorporates a memory architecture of at least one of HDD, SSD, and NVMe, and can be widely used in personal computers, data center servers, workstations, network storage devices, embedded systems, and other electronic devices that require high-performance, high-capacity, or both storage solutions. By flexibly combining memories with different characteristics, system performance, cost, and power consumption can be effectively optimized.

[0040] This invention also discloses a storage medium storing a computer program that, when executed by a processor, implements the aforementioned method. If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The storage medium can be a USB flash drive, external hard drive, external hard drive enclosure, or NAS, etc., making its use more flexible.

[0041] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A data processing method based on artificial intelligence, characterized in that, Includes the following steps: S1: Receive data to be stored from an external device containing multiple raw data items. The data to be stored is preprocessed to generate multiple data packets corresponding to the raw data items. S2: Input the data packet into a data processing model pre-trained based on artificial intelligence for identification and analysis, generate one or more feature vectors and one or more labels, and encapsulate the feature vectors, labels and inherent attributes extracted from the original data items to form structured metadata; S3: Based on predefined business rules and data structures, globally unify the identifiers of the index database, and organize the feature vectors and metadata into one or more index structures. Each index structure maps one or more feature vectors to a list of identifiers of the original data items containing the feature vectors. S4: Input the index structure into the data processing model, and based on the processing result of the data processing model, persistently write the one or more index structures, the identifier of the original data item, and the corresponding metadata into the index database. S5: Detect data packets written to the index database using a data security verification model pre-trained based on artificial intelligence; process abnormal data packets according to the verification results of the data security verification model; and support at least one data encryption operation among encrypted storage, face recognition, and fingerprint recognition unlocking for the data packets. S6: The user obtains structured results by accessing the index database and displays them to the user.

2. The data processing method according to claim 1, characterized in that, In S3, the identifier of the original data item is generated using the SHA-1 hash algorithm, the MD5 hash algorithm, or the SHA-256 hash algorithm.

3. The data processing method according to claim 2, characterized in that, In S4, the index database scans the data to be stored and queries the database to see if the identifier of the original data item exists. If the identifier already exists, the original data item is determined to be duplicate data, the physical storage operation is skipped, and a reference to the stored data is returned; If the hash identifier does not exist, the original data is written to the index database.

4. The data processing method according to claim 1, characterized in that, In S6, users input natural language queries through the Web UI or App, which are then converted into semantic vectors and matched with feature vectors of metadata stored in the index database. Based on semantic similarity and keywords, the queries are retrieved and sorted, and the matching file paths, titles, and categories are returned and displayed to the user.

5. A data processing system based on artificial intelligence, wherein the system performs data processing using the method described in any one of claims 1 to 4, characterized in that, include: An interface module connects to an external device and receives data to be stored from the external device. The AI ​​processing module is connected to the interface module, receives the data to be stored transmitted from the interface module, and performs file recognition, classification, OCR, face recognition and semantic analysis. The storage module is connected to the AI ​​processing module to achieve data communication, and performs data processing operations on the data to be stored based on the data processing model to generate the index structure, identifier and metadata for subsequent search use; The security module is connected to the AI ​​processing module to achieve interaction, and performs data processing operations such as malicious file detection, data encryption, and facial recognition or fingerprint recognition unlocking on the data to be stored based on the data security verification model. The data indexing module is connected to the AI ​​processing module, receives data transmitted from the AI ​​processing module, and stores the file tags, metadata, and semantic vectors corresponding to the original data items in the index database based on SQLite or an embedded database; The user interaction module connects with the data index module to achieve data exchange. It allows users to access the data index module by inputting keywords and semantic search, quickly locate files in the index database, and display the results to the user.

6. A storage device comprising a data transfer interface, a memory, a processor, and the system of claim 5, characterized in that, Several data transmission interfaces are connected to a processor, which includes an AI processing module, a security module, and a user interaction module. The processor is connected to a memory, which includes a storage module and a data index module and stores a computer program. When the processor executes the computer program, it implements the method described above.

7. The storage device according to claim 6, characterized in that, The processor has an integrated NPU or AI SoC and uses the UFS communication protocol standard.

8. The storage device according to claim 6, characterized in that, The data transmission interfaces include USB 3.2, USB 4, Thunderbolt, PCIe, and Ethernet interfaces.

9. The storage device according to claim 6, characterized in that, The storage includes at least one of HDD, SSD and NVMe SSD.

10. A storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 4.

Images