Safety control method for smart phone maintenance based on big data

Abstract

The invention relates to the technical field of intelligent equipment maintenance, and discloses a security control method for intelligent mobile phone maintenance based on big data, which comprises the following steps: step 1, data acquisition and primary processing: acquiring hardware state information and software operation information of equipment through a diagnosis interface of an intelligent mobile phone, the hardware state information comprises a battery health condition, CPU temperature and sensor performance, the software operation information comprises a system version, an application crash log and an abnormal alarm record, and meanwhile, recording timestamps of all pieces of acquisition information to form a data set. According to the method, hardware and software feature information is extracted through historical equipment states and maintenance records, a deep learning model is combined to achieve accurate judgment of fault types, a predefined maintenance scheme is matched, meanwhile, the fault severity degree is described by marking the equipment health state level, high-intelligence support is provided for maintenance priority judgment and resource scheduling, and the maintenance efficiency is improved. And an automatic and intelligent diagnosis process is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent device maintenance, and in particular to a safety management and control method for intelligent phone maintenance based on big data. Background Art

[0002] With the widespread popularity of smartphones and their enhanced functions, mobile phones have become an indispensable tool in people's daily lives. However, due to the high frequency of use and heavy operating load of the devices, the failure rate of smartphones has increased simultaneously. Common failures include hardware problems and software problems. In order to ensure user experience, maintenance services have become an indispensable link. However, the existing maintenance service process has many problems in terms of technical means, data management, and user experience, which seriously restrict the improvement of maintenance service quality. The specific manifestations are as follows:

[0003] In the traditional mobile phone repair model, the repair process and operation details lack transparency to users, and the authenticity of the repair content and hardware replacement cannot be guaranteed. For example, repair personnel will replace the original hardware with inferior products or exaggerate the fault problems, which leads to users' distrust of repair services. In addition, users usually cannot understand the progress of the repair in real time, especially in complex fault scenarios, where there is a lack of effective communication on the repair steps and expected completion time.

[0004] Data is generated during the mobile phone repair process, including device status data, repair records, and hardware replacement information. Effective data management and protection is an important foundation for repair services, but existing repair services usually store repair information in the form of paper records or independent electronic forms. Data is easily lost, tampered with, or falsified, and the authenticity and integrity of repair history records cannot be guaranteed. At the same time, the repair process involves user privacy data, lacks data encryption and permission management, and user data security risks are prominent.

[0005] The complexity of smartphone hardware and software continues to increase, resulting in the traditional fault diagnosis method that relies on manual experience gradually failing to meet the needs. Maintenance personnel need to spend time checking the status of the equipment and determining the cause of the fault, which in turn affects the maintenance efficiency. At the same time, the maintenance system lacks an intelligent resource scheduling mechanism and cannot reasonably arrange the maintenance task priority, further extending the maintenance cycle.

[0006] Smartphones generate massive amounts of status data and log information during operation, including the health status of the device and potential fault information. However, traditional maintenance systems have limited capabilities in data collection, storage, and analysis, and are unable to fully tap the value of data for intelligent maintenance optimization. In addition, with the expansion of the scale of maintenance equipment, traditional single-machine storage and analysis methods cannot meet the real-time processing needs of large-scale data.

[0007] Users are in a passive state during the maintenance service process, unable to grasp the specific circumstances of equipment failures and unable to actively participate in decision-making in the maintenance process. At the same time, the existing maintenance system lacks health status monitoring and fault warning functions, so users cannot be informed of potential problems in advance and take corresponding measures, and the service life and performance maintenance of the equipment cannot be effectively guaranteed.

[0008] Therefore, those skilled in the art provide a safety management and control method for smartphone maintenance based on big data, which can comprehensively solve the problems of transparency, data security, diagnostic efficiency, large-scale data processing capabilities and insufficient user experience in the maintenance process. Summary of the invention

[0009] In view of the deficiencies in the prior art, the present invention provides a safety management and control method for smartphone maintenance based on big data, which can comprehensively solve the problems of transparency, data security, diagnostic efficiency, large-scale data processing capabilities and insufficient user experience in the maintenance process.

[0010] To achieve the above objectives, the present invention is implemented through the following technical solutions: a safety control method for smartphone maintenance based on big data, comprising:

[0011] Step 1: Data collection and preliminary processing: The hardware status information and software operation information of the device are collected through the diagnostic interface of the smartphone. The hardware status information includes battery health, CPU temperature and sensor performance, and the software operation information includes system version, application crash log and abnormal alarm record. At the same time, the timestamp of each collected information is recorded to form a data set, which is formatted and stored as unified device status data to remove redundant information and obvious abnormal values ​​in the collection process;

[0012] Step 2: Storing maintenance data: uploading the collected device status data to the distributed storage system and associating it with the maintenance record data of the smartphone. The maintenance record data includes maintenance steps, maintenance personnel identification, timestamp, and hardware replacement information. The associated device status data and maintenance record data are stored as a time series structure in the distributed file system for subsequent retrieval and analysis.

[0013] Step 3: Fault diagnosis: Based on the time series structure data stored in the distributed storage system, feature information of hardware and software is extracted according to the historical equipment status and maintenance records. The feature information is used to identify the fault type, and the predefined maintenance plan is matched in combination with the current equipment status data and the fault type. At the same time, the equipment health status level is marked to describe the severity of the equipment fault.

[0014] Step 4: Blockchain storage: Generate a block with the generated maintenance plan and health status level together with the time series data of the equipment. The block contains the equipment identification, data hash value and timestamp, and records the data in the blockchain distributed ledger through blockchain technology. Each block records the hash value of the previous block and the hash value of the current block.

[0015] Step 5: User rights management: set the access rights of users, maintenance personnel and system administrators through a hierarchical authorization mechanism. Users can view blockchain data related to the equipment, maintenance personnel can access the equipment history records that cannot be changed by the current equipment failure type and maintenance plan, and the system administrator is responsible for maintaining and auditing the blockchain ledger.

[0016] Step 6: User interaction and maintenance monitoring: The equipment failure type, maintenance plan, and health status level recorded in the blockchain are visualized and displayed to the user through the user terminal application. The function of real-time monitoring of maintenance steps is also provided. The real-time monitoring content includes the completed maintenance steps and the total estimated maintenance time, so as to provide early warning reminders according to the equipment health status level.

[0017] Preferably, the hardware status information in step 1 further includes the voltage, current, number of charging cycles and operating status of key sensors of the device, and the software operation information further includes the kernel version of the operating system, the resource occupancy ratio of the application and the background service startup record, and the above information is stored in JSON format to support distributed data query.

[0018] Preferably, the data collected in step 1 is used to remove outliers by the following formula:

[0019] x∈[μ-kσ,μ+kσ],

[0020] Among them, x is the value of an indicator of the data collected by the device; μ is the mean of the indicator; σ is the standard deviation; the value range of k is 2 to 3, which is used to balance the sensitivity of outlier detection.

[0021] Preferably, after the equipment status data and maintenance record data in step 2 are stored as a time series structure, they are further compressed using the following formula:

[0022]

[0023] Among them, C t is the compressed value of the time series, D i is the i-th data, and n is the total number of data.

[0024] Preferably, the equipment health status level in step 3 is calculated by the following formula:

[0025]

[0026] Among them, S h is the health status level, with a value range of [0, 1],

[0027] Failure Count is the number of observed failure events.

[0028] Total Observations is the total number of detections.

[0029] The fault diagnosis in step 3 is implemented through a deep learning model, which is composed of a convolutional neural network and a long short-term memory network. The convolutional neural network is used to extract the spatial features of the equipment data, and the long short-term memory network is used to analyze the time series pattern and finally output the fault type.

[0030] Preferably, the block generated in step 4 includes the following contents: summary information of the current maintenance plan, health status level, hash value H(B) of the previous block, k-1 ), current data hash value H(D k ) and block generation time T k , where the hash value is calculated using the following formula:

[0031] H(D k )=SHA256(D k +H(B k-1 )+T k ),

[0032] SHA256 is a cryptographically secure hash algorithm that outputs a fixed-length 256-bit hash value. k is the data content of the current block k;

[0033] The blockchain in step 4 adopts a consortium chain architecture, and the blockchain nodes are divided into data recording nodes, audit nodes and read-only nodes. Among them, the data recording nodes are responsible for writing blockchain data, the audit nodes are responsible for verifying data integrity, and the read-only nodes are used for user queries.

[0034] Preferably, in step 6, the user terminal application provides real-time updates of the device health status level and provides early warning prompts according to the following rules:

[0035] If S h When ≤0.3, an emergency maintenance reminder is generated;

[0036] If 0.3<S h When <0.6, a suggested maintenance reminder is generated;

[0037] If S h When <0.6, no reminder will be given.

[0038] Preferably, the blockchain storage in step 4 adopts a layered storage architecture, wherein:

[0039] The first layer is the fast access layer, which is used to store the complete data of the most recent m blocks;

[0040] The second layer is the historical archive layer, which is used to store hash summary information exceeding m blocks;

[0041] Data access priority determines the access path using the following formula:

[0042]

[0043] Among them, P(d) is the data access path, k is the current block number, L1 represents the fast access layer, L2 represents the historical archive layer, and m is the set block number threshold.

[0044] Preferably, the loss function of the deep learning model in step 3 is defined as cross entropy loss:

[0045]

[0046] Among them, L is the loss value, N is the number of samples, M is the number of fault categories, and y ij Indicates the actual value that sample I belongs to category J, Indicates the predicted value that sample I belongs to category J.

[0047] Preferably, in step 6, the user terminal application provides a real-time maintenance progress prediction function, and the progress prediction is calculated by the following formula:

[0048]

[0049] in, is the estimated total maintenance time, T c is the current maintenance time, P t The current maintenance completion progress is calculated as follows:

[0050]

[0051] Among them, S c is the number of completed maintenance steps, S t is the total number of maintenance steps.

[0052] The present invention provides a safety control method for smartphone maintenance based on big data.

[0053] Beneficial effects:

[0054] 1. The present invention extracts hardware and software feature information through historical equipment status and maintenance records, combines deep learning models to accurately identify fault types, and matches predefined maintenance plans. At the same time, by marking the equipment health status level to describe the severity of the fault, it provides highly intelligent support for maintenance priority judgment and resource scheduling, and realizes an automated and intelligent diagnostic process.

[0055] 2. The present invention uses blockchain technology to generate blocks of maintenance plans, health status levels and time series data, and records them in a distributed ledger. The chain storage structure and encrypted hash algorithm of the blockchain ensure the integrity and non-tamperability of the maintenance data, enhance the reliability of the maintenance process, and solve the problem that traditional maintenance records are easily tampered with.

[0056] 3. The present invention can monitor the maintenance progress in real time through the terminal application, including the dynamic update of the completed steps, estimated completion time and health status level, and the early warning reminder function based on the health status level, so that users can understand the equipment status in time and take corresponding measures, thereby significantly improving the user's transparency and participation in the maintenance process.

[0057] 4. The present invention supports the storage, query and analysis of massive equipment data through the combination of distributed storage system and blockchain. The data storage adopts JSON format and time series structure, which makes data management more efficient. At the same time, it provides a technical basis for subsequent big data mining and maintenance optimization, and then shows superiority in processing large-scale equipment data. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0059] In order to make the technical personnel in the technical field understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in combination with the drawings in the embodiment of the present invention. Obviously, the described embodiment is a partial embodiment of the present invention, not a complete embodiment. Based on the embodiment of the present invention, other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present invention.

[0060] The present invention is described in detail below in conjunction with the accompanying drawings:

[0061] Example:

[0062] Please see attached Figure 1 The embodiment of the present invention provides a safety control method for smart phone maintenance based on big data, including:

[0063] Step 1: Data collection and preliminary processing: The hardware status information and software operation information of the device are collected through the diagnostic interface of the smartphone. The hardware status information includes battery health, CPU temperature and sensor performance, and the software operation information includes system version, application crash log and abnormal alarm record. At the same time, the timestamp of each collected information is recorded to form a data set, which is formatted and stored as unified device status data to remove redundant information and obvious abnormal values ​​in the collection process;

[0064] Step 2: Storing maintenance data: uploading the collected device status data to the distributed storage system and associating it with the maintenance record data of the smartphone. The maintenance record data includes maintenance steps, maintenance personnel identification, timestamp, and hardware replacement information. The associated device status data and maintenance record data are stored as a time series structure in the distributed file system for subsequent retrieval and analysis.

[0065] Step 3: Fault diagnosis: Based on the time series structure data stored in the distributed storage system, feature information of hardware and software is extracted according to the historical equipment status and maintenance records. The feature information is used to identify the fault type, and the predefined maintenance plan is matched in combination with the current equipment status data and the fault type. At the same time, the equipment health status level is marked to describe the severity of the equipment fault.

[0066] Step 4: Blockchain storage: Generate a block with the generated maintenance plan and health status level together with the time series data of the equipment. The block contains the equipment identification, data hash value and timestamp, and records the data in the blockchain distributed ledger through blockchain technology. Each block records the hash value of the previous block and the hash value of the current block.

[0067] Step 5: User rights management: set the access rights of users, maintenance personnel and system administrators through a hierarchical authorization mechanism. Users can view blockchain data related to the equipment, maintenance personnel can access the equipment history records that cannot be changed by the current equipment failure type and maintenance plan, and the system administrator is responsible for maintaining and auditing the blockchain ledger.

[0068] Step 6: User interaction and maintenance monitoring: The equipment failure type, maintenance plan, and health status level recorded in the blockchain are visualized and displayed to the user through the user terminal application. The function of real-time monitoring of maintenance steps is also provided. The real-time monitoring content includes the completed maintenance steps and the total estimated maintenance time, so as to provide early warning reminders according to the equipment health status level.

[0069] The hardware status information and software operation information are collected through the diagnostic interface, and the data is uniformly formatted and cleaned in combination with the timestamp to remove redundant information and obvious abnormal values, ensuring the high quality and integrity of the data and providing a reliable data basis for subsequent analysis;

[0070] Optimize data management and storage efficiency. Associate the collected equipment status data with maintenance record data and store them in a time series structure in a distributed storage system to achieve efficient organization and management of data.

[0071] Realize intelligent fault diagnosis Based on historical data in the distributed storage system, use deep learning models to extract hardware and software features, accurately identify fault types and generate maintenance plans, and combine the health status level annotation function to intuitively describe the severity of equipment failures, providing a scientific basis for maintenance priority judgment and resource scheduling;

[0072] Improve the security and credibility of maintenance data. Blockchain technology is used to generate blocks of maintenance plans, health status levels and time series data, and record them in a distributed ledger. The chain storage structure and encrypted hash algorithm of blockchain ensure the integrity and immutability of data, solving the problem that traditional maintenance records are easily modified or forged.

[0073] Enhance the transparency of the maintenance process. The terminal application displays maintenance steps, health status level, and estimated completion time in real time. Users can clearly understand the maintenance progress and avoid information asymmetry. The transparency of the maintenance process increases users' trust in and participation in maintenance services.

[0074] Proactive health monitoring and early warning Based on the dynamic update function of the equipment health status level, users can know the equipment status in time and take corresponding measures. If the health status level is low, the system will provide early warning reminders to avoid further damage to the equipment and extend the service life of the equipment;

[0075] Improve maintenance efficiency and service experience. Intelligent analysis and real-time update of maintenance data optimize the maintenance process, rationally allocate maintenance resources, and improve overall efficiency. At the same time, real-time progress monitoring and interactive feedback functions enhance user experience and make the maintenance process more convenient and transparent.

[0076] Support large-scale data processing and expansion Through the combination of distributed storage system and blockchain, the system can efficiently process large-scale equipment data; the time-series storage and JSON formatting design of data make data analysis and expansion more convenient, providing a solid technical foundation for future big data mining and maintenance optimization;

[0077] Providing multi-level permissions to ensure user data security Through the hierarchical authorization mechanism, different roles can access data related to their permissions to avoid sensitive data leakage; maintenance personnel can view the fault type and solution of the current equipment, while the administrator is responsible for account book maintenance, further ensuring data security;

[0078] Build an intelligent, transparent and secure maintenance ecosystem. This invention integrates technologies such as big data, artificial intelligence and blockchain to build a comprehensive solution for the smartphone maintenance industry that integrates transparency, intelligence and security. The integrity and innovation of the system effectively improve the industry's service level and provide strong support for the development of the maintenance industry.

[0079] The hardware status information in step 1 further includes the voltage, current, number of charging cycles and operating status of key sensors of the device, and the software operation information further includes the kernel version of the operating system, the resource usage ratio of the application and the background service startup record, and the above information is stored in JSON format to support distributed data query.

[0080] The data collected in step 1 is used to remove outliers using the following formula:

[0081] x∈[μ-kσ,μ+kσ],

[0082] Among them, x is the value of an indicator of the data collected by the device; μ is the mean of the indicator; σ is the standard deviation; the value range of k is 2 to 3, which is used to balance the sensitivity of outlier detection.

[0083] The formula is used to remove outliers in the data to avoid the risk of extreme values ​​affecting subsequent analysis results. The indicator values ​​are limited to a reasonable range to ensure the authenticity and reliability of the collected data.

[0084] The value range of parameter k in the formula (2 to 3) provides flexible sensitivity adjustment capability for outlier detection. For different devices or scenarios, the outlier filtering criteria can be optimized by adjusting the k value to adapt to diverse data characteristics.

[0085] By judging abnormal values ​​through standard deviation, we can effectively filter out erroneous readings during sensor acquisition or accidental abnormalities during equipment operation. This can reduce the interference of noise data on subsequent analysis of feature extraction and fault diagnosis, and improve the accuracy of data analysis.

[0086] After removing outliers, the data distribution conforms to the normal state characteristics of equipment operation. The deep learning model or diagnostic algorithm built based on the data can reflect the true state of the equipment and improve the accuracy and reliability of fault diagnosis.

[0087] By applying the outlier removal formula, the present invention achieves efficient filtering and cleaning of equipment collected data, providing an accurate and stable basic guarantee for the storage, analysis and intelligent application of subsequent maintenance data.

[0088] After the equipment status data and maintenance record data in step 2 are stored as a time series structure, they are further compressed using the following formula:

[0089]

[0090] Among them, C t is the compressed value of the time series, D i is the i-th data, and n is the total number of data.

[0091] Through the formula, the time series data is compressed into a single representative value C t , effectively reducing the storage space occupied. Compared with storing original data, the compressed data volume is smaller, optimizing the resource usage of the distributed storage system.

[0092] The compressed time series data structure simplifies the subsequent retrieval and analysis process. When querying and analyzing are required, the compressed value C is directly used. t It can quickly obtain the overall trend of time series, reduce the complexity of data calculation, and improve analysis efficiency.

[0093] In a distributed storage environment, the compressed time series data is smaller in size, and the bandwidth and time costs required for transmission are reduced. This is particularly important for cross-node data access and remote diagnostic services, and helps improve the response speed of the maintenance system.

[0094] The original time series data contains redundant information that is repeated or fluctuates, and the information has limited value for subsequent analysis. Through data compression, redundant data is removed, the expressiveness of key features is enhanced, and the analysis model is focused on identifying actual faults or trends.

[0095] Through the compression formula, the present invention optimizes data storage and processing while maintaining the core information of time series data, providing efficient and streamlined data management capabilities for the smartphone maintenance system, and laying the foundation for subsequent intelligent analysis and interactive functions.

[0096] The device health status level in step 3 is calculated using the following formula:

[0097]

[0098] Among them, S h is the health status level, with a value range of [0, 1],

[0099] Failure Count is the number of observed failure events.

[0100] Total Observations is the total number of detections.

[0101] The fault diagnosis in step 3 is implemented through a deep learning model, which consists of a convolutional neural network and a long short-term memory network. The convolutional neural network is used to extract the spatial features of the equipment data, and the long short-term memory network is used to analyze the time series pattern and finally output the fault type.

[0102] The equipment health status level is calculated by the formula. The value range of the health status level is [0, 1], which can directly reflect the current operating status of the equipment. Avoid serious losses caused by fault expansion. The health status level combines the number of fault events and the total number of detections to quantify the failure frequency and stability of the equipment. Maintenance personnel can judge the urgency and priority of the fault based on the health status level, providing a scientific basis for the scheduling of maintenance resources and task allocation.

[0103] Convolutional neural networks are used to extract the spatial features of device data to ensure in-depth mining of hardware and software features. At the same time, long-term and short-term memory networks are combined to analyze time series patterns, so that the diagnostic model can identify dynamic changes in device status, accurately capture potential fault trends, and improve the accuracy of fault type identification.

[0104] The deep learning model jointly analyzes the equipment hardware status information and software operation information to avoid diagnostic bias caused by a single feature. Multi-dimensional feature fusion improves the comprehensiveness of fault identification. After the fault type is output by the diagnostic model, the system automatically matches the predefined maintenance plan to reduce manual intervention and improve maintenance efficiency. The maintenance plan is combined with the health status level to further provide a basis for optimizing the maintenance strategy.

[0105] Through the intelligent diagnosis technology of equipment health status level calculation and deep learning model, the present invention realizes quantitative evaluation of equipment operation status, automatic fault identification and matching of maintenance plans, significantly improves the efficiency, accuracy and transparency of the maintenance process, and lays a technical foundation for intelligent maintenance services.

[0106] The block generated in step 4 contains the following: summary information of the current maintenance plan, health status level, hash value H(B) of the previous block, and the k-1 ), current data hash value H(D k ) and block generation time T k , where the hash value is calculated using the following formula:

[0107] H(D k )=SHA256(D k +H(B k-1 )+T k ),

[0108] SHA256 is a cryptographically secure hash algorithm that outputs a fixed-length 256-bit hash value. k is the data content of the current block k;

[0109] The blockchain in step 4 adopts a consortium chain architecture. The blockchain nodes are divided into data recording nodes, audit nodes and read-only nodes. Among them, the data recording nodes are responsible for writing blockchain data, the audit nodes are responsible for verifying data integrity, and the read-only nodes are used for user queries.

[0110] The hash value of the current block is generated through a formula, and the data content, the hash value of the previous block and the timestamp are combined and calculated using a cryptographically secure hash algorithm to ensure the integrity and immutability of blockchain data. Any tampering with the data will result in an inconsistent hash value, ensuring the authenticity of the maintenance record.

[0111] The block contains maintenance plan summary information and health status level, and the chain structure ensures the traceability of maintenance history. Users, maintenance personnel and administrators verify the validity of each record through the blockchain to avoid falsification of maintenance records and improve users' trust in maintenance services.

[0112] Through the alliance chain architecture design, blockchain nodes are divided into data recording nodes, audit nodes and read-only nodes, realizing clear division of responsibilities. Data recording nodes are responsible for data writing, audit nodes ensure data consistency and integrity, read-only nodes support user query needs, and distributed storage methods improve data processing efficiency and reliability.

[0113] Blockchain uses a distributed ledger storage mechanism to avoid data loss or system unavailability caused by single point failure in traditional centralized storage systems. The multi-node structure in the alliance chain further improves data invulnerability and system stability.

[0114] The integrity of blockchain data is verified through audit nodes to ensure that each record complies with storage rules and has not been tampered with. Audit nodes use hash value comparison to quickly detect data anomalies and improve verification efficiency.

[0115] Through the design of block generation mechanism and alliance chain architecture, the present invention realizes the secure storage and efficient management of maintenance data, provides a transparent and reliable solution for the smartphone maintenance industry, and at the same time ensures the efficiency and scalability of large-scale data storage and query.

[0116] In step 6, the user terminal application provides real-time updates of the device health status level and provides early warning prompts through the following rules:

[0117] If S h When ≤0.3, an emergency maintenance reminder is generated;

[0118] If 0.3<S h When <0.6, a suggested maintenance reminder is generated;

[0119] If S h When <0.6, no reminder will be given.

[0120] Through real-time monitoring and a reasonable early warning mechanism, users can carry out timely repairs on the equipment before major failures occur, avoiding permanent damage to the equipment due to minor problems not being handled in time, and extending the service life of the equipment.

[0121] The early warning rules provide clear reminder conditions, allowing users to flexibly arrange maintenance time according to the health status of the equipment, without the need for emergency maintenance after a fault occurs, reducing the inconvenience in life or work caused by sudden maintenance.

[0122] Real-time updates of health status and early warning functions help users avoid critical mission interruptions or data loss due to equipment failures, and can effectively reduce the risks caused by equipment hardware crashes or unrecoverable data.

[0123] By generating early warning reminders in a timely manner, users can perform repairs before equipment failures further deteriorate, reducing the high costs caused by delayed repairs.

[0124] Through real-time updates of health status levels and scientific graded warning rules, the present invention realizes active and intelligent equipment health management, improves users' ability to control equipment operating status, optimizes the configuration and use of maintenance resources, and provides strong support for the long-term and stable operation of equipment.

[0125] The blockchain storage in step 4 adopts a layered storage architecture, where:

[0126] The first layer is the fast access layer, which is used to store the complete data of the most recent m blocks;

[0127] The second layer is the historical archive layer, which is used to store hash summary information exceeding m blocks;

[0128] Data access priority determines the access path using the following formula:

[0129]

[0130] Among them, P(d) is the data access path, k is the current block number, L1 represents the fast access layer, L2 represents the historical archive layer, and m is the set block number threshold.

[0131] Complete data is stored in the fast access layer L1 to ensure that users can instantly access frequently used data. The hash summary information stored in the historical archive layer L2 can still ensure data integrity and security through the on-chain verification mechanism.

[0132] The data in the historical archive layer is stored as hash summaries to avoid traversing and querying large-scale historical data one by one. The historical data is verified by quickly comparing the hash values, which significantly reduces the query overhead and time complexity.

[0133] As the scale of blockchain storage data expands, the layered storage architecture compresses earlier blocks into hash summary form, enabling the system to efficiently process long-term accumulated maintenance data and improve the scalability of blockchain for large-scale data.

[0134] Through the layered storage architecture, the present invention achieves efficient data storage and access management while ensuring the security and integrity of blockchain data, providing strong support for the real-time processing and long-term storage of large-scale maintenance data.

[0135] The loss function of the deep learning model in step 3 is defined as the cross entropy loss:

[0136]

[0137] Among them, L is the loss value, N is the number of samples, M is the number of fault categories, and y ij Indicates the actual value that sample I belongs to category J, Indicates the predicted value that sample I belongs to category J.

[0138] The cross entropy loss function guides the parameter optimization of the deep learning model by measuring the difference between the model's predicted probability and the actual category, making the predicted probability output by the model closer to the true distribution. For multi-classification tasks, cross entropy can effectively improve the accuracy of classification.

[0139] The cross entropy loss function is used for multi-category classification tasks and can simultaneously handle multiple hardware and software failure types in smartphone repair, providing strong support for the accurate identification of complex failure modes.

[0140] The cross entropy loss function penalizes and guides the probability distribution of the model output, making the model learning process more efficient, quickly converging to the optimal state, and reducing training time. It is very important for smartphone fault diagnosis with large-scale data sets.

[0141] The cross entropy loss function uses logarithmic operations and is sensitive to small changes in model predictions, making the model optimization process smoother. The trained model can maintain high classification stability among different equipment fault types.

[0142] Through the cross-entropy loss function, the deep learning model of the present invention can efficiently and accurately handle multi-category fault classification tasks of smartphones, enhance the reliability and adaptability of the intelligent diagnosis system, and provide important technical support for optimizing the smartphone maintenance process.

[0143] In step 6, the user terminal application provides a real-time maintenance progress prediction function, and the progress prediction is calculated using the following formula:

[0144]

[0145] in, is the estimated total maintenance time, T c is the current maintenance time, P t The current maintenance completion progress is calculated as follows:

[0146]

[0147] Among them, S c is the number of completed maintenance steps, S t is the total number of maintenance steps.

[0148] By calculating the estimated total maintenance time and combining it with the elapsed maintenance time and the current maintenance completion progress, users can intuitively understand the estimated completion time of the maintenance task, effectively reducing the uncertainty of the maintenance time and improving the user's sense of control over the maintenance process.

[0149] By calculating the current maintenance completion progress and reflecting the ratio of the number of completed maintenance steps to the total number of maintenance steps in real time, users can dynamically track the maintenance progress and avoid information asymmetry.

[0150] Real-time display of estimated total maintenance time and progress information alleviates user dissatisfaction or anxiety caused by unclear maintenance cycles. Transparent maintenance time management makes users more confident in the service process and significantly improves user experience and satisfaction.

[0151] For maintenance personnel or maintenance centers, this function can help plan maintenance task priorities and resource allocation. For example, maintenance personnel can adjust their work rhythm through total time forecast to ensure that high-priority tasks are completed on time, thereby improving overall maintenance efficiency.

[0152] Through the real-time maintenance progress prediction function, the present invention provides users and maintenance personnel with a clear time management tool, improves the transparency and efficiency of the maintenance process, improves user experience and resource scheduling capabilities, and provides efficient and intelligent time prediction support for smartphone maintenance services.

[0153] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A safety control method for smartphone maintenance based on big data, characterized in that: include: Step 1: Data collection and preliminary processing: The hardware status information and software operation information of the device are collected through the diagnostic interface of the smartphone. The hardware status information includes battery health, CPU temperature and sensor performance, and the software operation information includes system version, application crash log and abnormal alarm record. At the same time, the timestamp of each collected information is recorded to form a data set, which is formatted and stored as unified device status data to remove redundant information and obvious abnormal values ​​in the collection process; Step 2: Storing maintenance data: uploading the collected device status data to the distributed storage system and associating it with the maintenance record data of the smartphone. The maintenance record data includes maintenance steps, maintenance personnel identification, timestamp, and hardware replacement information. The associated device status data and maintenance record data are stored as a time series structure in the distributed file system for subsequent retrieval and analysis. Step 3: Fault diagnosis: Based on the time series structure data stored in the distributed storage system, feature information of hardware and software is extracted according to the historical equipment status and maintenance records. The feature information is used to identify the fault type, and the predefined maintenance plan is matched in combination with the current equipment status data and the fault type. At the same time, the equipment health status level is marked to describe the severity of the equipment fault. Step 4: Blockchain storage: Generate a block with the generated maintenance plan and health status level together with the time series data of the equipment. The block contains the equipment identification, data hash value and timestamp, and records the data in the blockchain distributed ledger through blockchain technology. Each block records the hash value of the previous block and the hash value of the current block. Step 5: User rights management: set the access rights of users, maintenance personnel and system administrators through a hierarchical authorization mechanism. Users can view blockchain data related to the equipment, maintenance personnel can access the equipment history records that cannot be changed by the current equipment failure type and maintenance plan, and the system administrator is responsible for maintaining and auditing the blockchain ledger. Step 6: User interaction and maintenance monitoring: The equipment failure type, maintenance plan, and health status level recorded in the blockchain are visualized and displayed to the user through the user terminal application. The function of real-time monitoring of maintenance steps is also provided. The real-time monitoring content includes the completed maintenance steps and the total estimated maintenance time, so as to provide early warning reminders according to the equipment health status level.

2. According to the big data-based safety management and control method for smartphone maintenance according to claim 1, it is characterized in that: The hardware status information in step 1 further includes the voltage, current, number of charging cycles and operating status of key sensors of the device, and the software operation information further includes the kernel version of the operating system, the resource occupancy ratio of the application and the background service startup record, and the above information is stored in JSON format to support distributed data query.

3. According to the big data-based safety management and control method for smartphone maintenance according to claim 2, it is characterized in that: The data collected in step 1 are used to remove outliers by the following formula: x∈[μ-kσ,μ+kσ], Among them, x is the value of an indicator of the data collected by the device; μ is the mean of the indicator; σ is the standard deviation; the value range of k is 2 to 3, which is used to balance the sensitivity of outlier detection.

4. According to the big data-based safety management and control method for smartphone maintenance according to claim 1, it is characterized in that: After the equipment status data and maintenance record data in step 2 are stored as a time series structure, they are further compressed using the following formula: Among them, C t is the compressed value of the time series, D i is the i-th data, and n is the total number of data.

5. According to the big data-based safety management and control method for smartphone maintenance according to claim 1, it is characterized in that: The equipment health status level in step 3 is calculated by the following formula: Among them, S h is the health status level, with a value range of [0, 1], Failure Count is the number of observed failure events. Total Observations is the total number of detections. The fault diagnosis in step 3 is implemented through a deep learning model, which is composed of a convolutional neural network and a long short-term memory network. The convolutional neural network is used to extract the spatial features of the equipment data, and the long short-term memory network is used to analyze the time series pattern and finally output the fault type.

6. The method for safety control of smartphone maintenance based on big data according to claim 1, characterized in that: The block generated in step 4 contains the following contents: summary information of the current maintenance plan, health status level, hash value H(B) of the previous block, k-1 ), current data hash value H(D k ) and block generation time T k , where the hash value is calculated using the following formula: H(D k )=SHA256(D k +H(B k-1 )+T k ), SHA256 is a cryptographically secure hash algorithm that outputs a fixed-length 256-bit hash value. k is the data content of the current block k; The blockchain in step 4 adopts a consortium chain architecture, and the blockchain nodes are divided into data recording nodes, audit nodes and read-only nodes. Among them, the data recording nodes are responsible for writing blockchain data, the audit nodes are responsible for verifying data integrity, and the read-only nodes are used for user queries.

7. The method for safety control of smartphone maintenance based on big data according to claim 5, characterized in that: In step 6, the user terminal application provides real-time updates of the device health status level and provides early warning prompts according to the following rules: If S h When ≤0.3, an emergency maintenance reminder is generated; If 0.3<S h When <0.6, a suggested maintenance reminder is generated; If S h When <0.6, no reminder will be given.

8. The method for safety control of smartphone maintenance based on big data according to claim 6, characterized in that: The blockchain storage in step 4 adopts a layered storage architecture, where: The first layer is the fast access layer, which is used to store the complete data of the most recent m blocks; The second layer is the historical archive layer, which is used to store hash summary information exceeding m blocks; Data access priority determines the access path using the following formula: Among them, P(d) is the data access path, k is the current block number, L1 represents the fast access layer, L2 represents the historical archive layer, and m is the set block number threshold.

9. The method for safety control of smartphone maintenance based on big data according to claim 5, characterized in that: The loss function of the deep learning model in step 3 is defined as the cross entropy loss: Among them, L is the loss value, N is the number of samples, M is the number of fault categories, and y ij Indicates the actual value that sample I belongs to category J, Indicates the predicted value that sample I belongs to category J.

10. The method for safety control of smartphone maintenance based on big data according to claim 7, characterized in that: In step 6, the user terminal application provides a real-time maintenance progress prediction function, and the progress prediction is calculated by the following formula: in, is the estimated total maintenance time, T c is the current maintenance time, P t The current maintenance completion progress is calculated as follows: Among them, S c is the number of completed maintenance steps, S t is the total number of maintenance steps.

Images