A terminal device performance data intelligent analysis method, system, device and medium

Abstract

The application provides a terminal device performance data intelligent analysis method, system, device and medium, and belongs to the technical field of data acquisition and analysis. The method comprises the following steps: a cloud server manages data acquisition rules, and a terminal device acquires association rules through an HTTP protocol. The device acquires performance data in real time according to the rules with the help of a monitoring Agent program, generates standardized event records through local processing, and reports the standardized event records to a cloud message middleware after compression through an MQTT protocol. A cloud data processing service adopts hierarchical processing, is first stored in a local SQLite database, is stored in a MySQL database after repeated alarm aggregation through data cleaning, is synchronized to Elasticsearch to establish an index, is combined with Kibana to be displayed and is analyzed in depth through TensorFlow, prediction and early warning information is generated, a structured analysis report is finally generated based on the result, and efficient intelligent analysis of terminal device performance data is realized.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of data collection and analysis, and more particularly relates to a terminal device performance data intelligent analysis system, method, device and medium. BACKGROUND

[0002] With the rapid development of the Internet of Things and intelligent display technology, the deployment scale of terminal devices such as central control screens, intelligent gateways and outdoor large screens presents an exponential growth. These devices are widely distributed and have complex environments, from chain store digital signage to transportation hub information large screens, and the device points are scattered and the deployment environments are various. In the traditional operation and maintenance mode, when the device has faults such as playing lag, switching inconvenience, CPU occupancy rate exceeding 90% or insufficient memory, complete reliance on manual on-site troubleshooting and processing leads to long delay of fault response for several hours or even several days, sharp increase of operation and maintenance cost, and inability to realize unified health state management of large-scale devices.

[0003] The existing technical solutions mostly adopt a passive response mechanism, only notifying operation and maintenance personnel through simple alarm after the fault occurs, lacking early prediction and dynamic protection capability for performance abnormalities. For example, the hardware performance degradation of outdoor large screens due to environmental temperature changes, or the playing abnormality caused by compatibility problems of specific video decoders, cannot be effectively identified and warned through the established rules. This passive monitoring makes the fault influence range often continue to expand, causing the business interruption time to be prolonged and the commercial loss risk to be significantly increased.

[0004] In addition, in the face of high-frequency and multi-dimensional performance data generated by a large number of devices in operation, the traditional monitoring system based on a relational database has obvious bottlenecks. The device logs and performance indicators generated in a single day can reach TB level, and the traditional architecture cannot support real-time data processing and multi-dimensional correlation analysis. The insufficient data storage and computing capability leads to inability to accurately extract effective information from historical data, and makes it more difficult to realize fault root cause positioning or performance degradation trend judgment based on multi-index fusion. For example, the playing data of 200 4K screens in a chain cinema needs to analyze more than 20 associated indicators in real time, and the traditional technical solution cannot fully meet such demand, resulting in lack of data support and decision basis for overall risk control at the system level, forming a significant data analysis blind area. SUMMARY

[0005] In view of the above problems, the present application aims to provide a terminal device performance data intelligent analysis method, system, device and medium, which realizes terminal data collection and compression through edge computing, combines cloud multi-level cache and machine learning analysis, constructs a complete operation and maintenance system from real-time monitoring to prediction and warning, and significantly improves the device management efficiency and reduces the business interruption risk.

[0006] To achieve the above-mentioned purpose, the present application realizes the following technical solutions:

[0007] In a first aspect, the embodiments of the present application provide a terminal device performance data intelligent analysis method, comprising:

[0008] The cloud server manages the data collection rules, and the terminal device acquires the data collection rules associated with the device properties from the cloud server through the HTTP protocol;

[0009] The terminal device collects system performance data in real time according to the data collection rules through the monitoring Agent program, performs calculation and judgment locally to generate standardized event records, generates device data after processing by the compression algorithm, and reports to the cloud message middleware asynchronously through the MQTT protocol;

[0010] The cloud data processing service consumes the device data in the message middleware, adopts a hierarchical processing mechanism, stores the device data in the local SQLite database first, aggregates and processes repeated alarms through the data cleaning pipeline, stores the unique and effective event records in the MySQL database after generating the records;

[0011] The data synchronization tool synchronizes the effective event records in the MySQL database to the Elasticsearch search engine to establish the device performance index analysis index;

[0012] Based on the device performance index analysis index, the data is displayed in combination with the Kibana data visualization platform, deep analysis is performed by using the TensorFlow machine learning framework, the device performance law is identified, and prediction and early warning information is generated as the analysis result;

[0013] Based on the analysis result, a structured analysis report is generated.

[0014] In an optional embodiment, the cloud server manages the data collection rules, and the terminal device acquires the data collection rules associated with the device properties from the cloud server through the HTTP protocol, comprising:

[0015] The cloud server creates the data collection rules, and the data collection rules include the CPU usage threshold, the memory occupancy threshold, the frame rate threshold, and the decoding time threshold;

[0016] The data collection rules are associated and bound with the device properties, and the data collection rules are published; the device properties include the device model, the device name, and the device unique identifier;

[0017] When the terminal device starts, the published data collection rules matching the device properties are acquired from the Nginx reverse proxy server through the HTTP request; the terminal device includes but is not limited to the central control screen, the intelligent gateway, and the outdoor large screen.

[0018] In an optional embodiment, the terminal device collects system performance data in real time according to the data collection rule through the monitoring Agent program, performs calculation and judgment locally to generate a standardized event record, generates device data after processing by a compression algorithm, and reports to the cloud message middleware asynchronously through the MQTT protocol, including:

[0019] On the terminal device, the monitoring Agent program loaded with the data collection rule is locally deployed to monitor system performance indicators in real time and perform localized calculation and judgment according to the corresponding threshold in the data collection rule, identify abnormal system performance indicators, determine the type of abnormality, and generate a standardized event record based on the device identifier of the terminal device; the system performance indicators include CPU usage, memory occupancy, frame rate, and decoding time consumption;

[0020] The standardized event record is compressed and processed by using the LZ4 compression algorithm to generate device data;

[0021] The device data is published to the cloud RabbitMQ message middleware through the Paho MQTT client, and a heartbeat packet containing device state information is sent regularly.

[0022] In an optional embodiment, the cloud data processing service consumes the device data in the message middleware, adopts a hierarchical processing mechanism, stores the device data in a local SQLite database first, and then stores the unique and valid event record in a MySQL database after aggregation processing of repeated alarms through a data cleaning pipeline, including:

[0023] The cloud data processing service obtains device data from the RabbitMQ message middleware and writes the device data into a local SQLite relational database for temporary caching;

[0024] An Apache Kafka data stream processing pipeline is established to clean and aggregate the device data in the SQLite relational database based on the device identifier, the type of abnormality, and the time window dimension;

[0025] The device data of the same abnormal type generated by the same device within a preset time window is merged into one valid event record;

[0026] The valid event record after deduplication is stored persistently in a MySQL relational database.

[0027] In an optional embodiment, the data synchronization tool synchronizes the valid event record in the MySQL database to the Elasticsearch search engine to establish an index for device performance indicator analysis, including:

[0028] Incrementally extracting valid event records from a MySQL relational database through a Logstash data collection engine;

[0029] Converting and mapping fields of the extracted valid event records;

[0030] Batch writing the processed valid event records into a specified index of an Elasticsearch distributed search engine.

[0031] In an optional implementation, the index based on the device performance indicators is analyzed, data is displayed in combination with a Kibana data visualization platform, and a TensorFlow machine learning framework is used for deep analysis to identify device performance rules and generate prediction and early warning information as analysis results, including:

[0032] Based on the index of device performance indicators, valid event records are extracted from the Elasticsearch distributed search engine, device data is read, and a Kibana data visualization platform is used to build multi-dimensional visual dashboards and generate visual charts;

[0033] The TensorFlow machine learning framework is applied to analyze the device data, a correlation rule algorithm is used to mine the spatio-temporal correlation between CPU usage and card lag, an LSTM time series prediction algorithm is used to predict the device hardware aging trend based on the historical device data of the past 30 days, and hardware aging risk warning information is generated as the analysis result.

[0034] In an optional implementation, based on the analysis results, a structured analysis report is generated, including:

[0035] Periodically reading the analysis results, automatically calling a JasperReports report engine service through a RESTful API interface, combining the analysis results with the visual charts generated by the Kibana data visualization platform, generating a structured analysis report, and outputting it in PDF format.

[0036] In a second aspect, the embodiments of the present application also provide a terminal device performance data intelligent analysis system, including:

[0037] A cloud rule management module deployed on a cloud server, used to create, store and manage data collection rules, and issue rule configurations associated with device attributes to terminal devices through an HTTP protocol;

[0038] A device-side data collection module deployed on a terminal device, used to load data collection rules, collect system performance data in real time, calculate and judge locally to generate standardized event records, and report them asynchronously after processing through a compression algorithm via an MQTT communication protocol;

[0039] A message middleware module, using RabbitMQ message middleware, is configured to receive and cache device data reported by terminal devices;

[0040] A cloud data processing module, including a SQLite relational database, an Apache Kafka data stream processing platform and a MySQL relational database, is configured to consume device data from the message middleware module, and perform hierarchical processing on the device data, including temporary caching, stream cleaning, repeated alarm aggregation and effective event record persistent storage;

[0041] A data synchronization and analysis module, including a Logstash data collection engine, an Elasticsearch distributed search engine and a TensorFlow machine learning framework, is configured to synchronize effective event records in the MySQL to the search engine, establish device performance index analysis indexes, and perform deep analysis and prediction;

[0042] A visualization and report generation module, including a Kibana data visualization platform and a JasperReports report engine, is configured to perform multi-dimensional visual display based on analysis results, and automatically generate structured analysis reports including charts and statistical conclusions.

[0043] In a third aspect, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor executes the program to implement the steps of the terminal device performance data intelligent analysis method according to any one of the preceding aspects.

[0044] In a fourth aspect, a storage medium is provided, having a computer program stored thereon, and the computer program is executable on a processor to implement the steps of the terminal device performance data intelligent analysis method according to any one of the preceding aspects.

[0045] As can be seen from the above technical solutions, the present application has the following advantages:

[0046] In the terminal device performance data intelligent analysis method provided by the present application, a cloud collaborative intelligent analysis system is constructed to realize full-process automation and intelligentization of terminal device performance management. The present method relies on edge computing to complete real-time data collection and preliminary judgment at the device end, significantly reducing network transmission pressure; through multi-level caching and stream cleaning mechanism, massive data is efficiently processed, and abnormal events are accurately aggregated; combined with the Elasticsearch search engine and machine learning algorithm, performance rules are deeply mined to realize the leap from real-time monitoring to risk prediction; finally, structured reports combining data visualization and deep analysis are automatically generated, greatly improving the operation and maintenance efficiency and reliability of large-scale terminal devices, effectively reducing the cost of manual intervention and the risk of business interruption.

[0047] The application realizes the automatic centralized management and control of large-scale terminal devices by constructing a cloud collaborative intelligent management architecture. The unified rule configuration and distribution mechanism enables the terminal devices to independently obtain the collection strategy and execute the monitoring tasks, significantly improving the operation and maintenance efficiency and effectively solving the management difficulties caused by the rapid increase of device scale.

[0048] The application adopts the combination of edge computing and intelligent compression transmission, greatly reduces the network bandwidth occupation by pre-processing and compression of data at the device end. At the same time, the cloud end adopts multi-level cache and data cleaning mechanism, effectively reduces the storage resource consumption, and realizes the efficient processing of massive device data.

[0049] The application establishes a complete data processing and analysis pipeline, realizes the deep mining and intelligent analysis of device performance data. Through the combination of distributed search engine and machine learning algorithm, valuable information can be quickly extracted from massive data, the internal law and potential problems of device performance can be found, and the data analysis ability and decision support ability are greatly enhanced.

[0050] Based on advanced machine learning algorithms, the application realizes the accurate prediction and early warning of device performance anomalies. By analyzing historical data trends and real-time running status, potential fault risks can be found in time, and warning information can be generated, so that device operation and maintenance can be changed from passive response to active prevention, effectively avoiding possible system failures.

[0051] The application builds a complete monitoring, analysis and early warning closed loop, and comprehensively improves the reliability and business continuity of device operation. From data collection to intelligent analysis, from anomaly warning to report generation, a complete operation and maintenance guarantee system is formed, which provides all-round protection for the stable operation of large-scale device groups, and significantly reduces the risk of business interruption. BRIEF DESCRIPTION OF DRAWINGS

[0052] In order to more clearly illustrate the technical solutions of the present application, the drawings needed in the description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0053] Figure 1 The flowchart of the terminal device performance data intelligent analysis method provided by the present application.

[0054] Figure 2 The structure diagram of the terminal device performance data intelligent analysis system provided by the present application.

[0055] Figure 3A structural schematic diagram of an electronic device provided for the present application is shown. DETAILED DESCRIPTION

[0056] The various embodiments of the present disclosure will be described more fully with reference to the accompanying drawings. The various embodiments of the present disclosure can have various modifications, and all should not be construed as being limited to the specific embodiments disclosed herein. It should be understood that the various embodiments of the present disclosure are not intended to limit the present disclosure to particular embodiments described herein, but include all solutions falling within the spirit and scope of the various embodiments of the present disclosure.

[0057] Hereinafter, the term "include" or "may include" used in the various embodiments of the present disclosure indicates the presence of disclosed functions, operations, or elements, and does not limit the addition of one or more functions, operations, or elements. Also, as used in the various embodiments of the present disclosure, the terms "include", "have", and their conjugates merely indicate the presence of specific features, numbers, steps, operations, elements, components, or combinations thereof, and should not be construed as excluding the presence or addition of one or more other features, numbers, steps, operations, elements, components, or combinations thereof.

[0058] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0059] Please refer to Figure 1 A method flowchart of a terminal device performance data intelligent analysis method in an embodiment is shown, and the method includes:

[0060] S1: The cloud server manages data collection rules, and the terminal device acquires the data collection rules associated with its own device attributes from the cloud server through the HTTP protocol.

[0061] In the specific embodiments, the data collection rules are created through a special configuration interface in the management background of the cloud server. For example, the CPU usage threshold is set to 80%, the memory occupancy threshold is set to 70%, the frame rate threshold is set to 25fps, and the decoding time threshold is set to 50ms. The setting of these thresholds is based on the consideration of common terminal device performance standards and actual business needs to ensure that performance abnormal situations can be effectively captured.

[0062] Then the data acquisition rules are associated with the device attributes. Specifically, in the management background, the device model, device name, and device unique identifier are selected through a drop-down menu or search box to achieve precise matching of the rules with specific devices. After binding, click the publish button to publish the data acquisition rules to the server so that they can be accessed by terminal devices.

[0063] When the terminal device (such as a central control screen, smart gateway, or outdoor large screen) starts, its built-in network module automatically sends a request to the Nginx reverse proxy server through HTTP to acquire data acquisition rules. The request carries the device's own attribute information, such as device model, name, and unique identifier.

[0064] After receiving the request, the Nginx reverse proxy server matches the device attribute information in the server-stored published data acquisition rules. If a matching rule is found, it is returned to the terminal device; if no matching rule is found, a default rule or a prompt message can be returned to the terminal device, indicating that the device failed to acquire a dedicated rule.

[0065] S2: The terminal device acquires system performance data in real time according to the data acquisition rules through the monitoring Agent program, calculates and generates standardized event records locally, processes device data through compression algorithms, and reports them to the cloud message middleware asynchronously through the MQTT protocol.

[0066] In the specific implementation, a monitoring Agent program is deployed locally on the terminal device. This program is designed based on edge computing architecture and aims to acquire key performance indicators in real time while minimizing the impact on the device's own performance. When the device starts and acquires data acquisition rules, the monitoring Agent program loads these rules immediately.

[0067] First, the monitoring Agent program monitors system performance indicators in real time, such as CPU usage, memory occupancy, frame rate, and decoding time, through system interfaces or specific monitoring tools. For frame rate monitoring, real-time frame rate data can be obtained by interacting with the display driver or video playback module; decoding time can be obtained by adding timing logic at key nodes in the decoding module.

[0068] Then, according to the corresponding threshold in the data collection rule, the localization calculation and judgment are performed. For example, when the CPU usage is monitored to exceed 80%, it is judged that the CPU usage is abnormal; when the memory occupancy rate exceeds 70%, it is judged that the memory occupancy is abnormal, etc. After identifying the abnormal system performance indicators, the type of the abnormality is determined, such as "CPU usage is too high", "memory occupancy is too high", etc., and a standardized event record is generated based on the device identifier of the terminal device. The record contains device identifier, abnormal type, abnormal time, specific performance indicator value at the time of abnormality, etc.

[0069] At this time, the standardized event record is compressed by using the LZ4 compression algorithm. The LZ4 compression algorithm has the characteristics of fast compression speed and moderate compression ratio, and is suitable for fast processing of data in the case of limited resources of the terminal device. By calling the LZ4 algorithm library, the standardized event record is converted into compressed device data. Through actual test, this compression mode can reduce data flow by 70%, effectively avoiding that the monitoring program itself becomes a performance burden.

[0070] Finally, the PahoMQTT client is configured on the terminal device, and the compressed device data is published to the cloud RabbitMQ message middleware. The MQTT protocol is a lightweight message transmission protocol, which is suitable for use on devices with limited network conditions. When publishing data, a suitable QoS (Quality of Service) level is set to ensure the reliability of data transmission.

[0071] At the same time, the monitoring Agent program sends a heartbeat packet containing device state information to the cloud regularly (for example, every 30 seconds). The heartbeat packet contains the current running state of the device (such as normal operation, abnormal operation, etc.), power information (for mobile devices), network connection state, etc., so that the cloud can timely understand the device state.

[0072] S3: The cloud data processing service consumes the device data in the message middleware, adopts a hierarchical processing mechanism, first stores the device data in the local SQLite database, and then aggregates the repeated alarms through the data cleaning pipeline to generate a unique effective event record and store it in the MySQL database.

[0073] In the specific embodiment, the cloud data processing service obtains the device data from the RabbitMQ message middleware through a special message consumption interface. After obtaining the data, the device data is written into the local SQLite relational database for temporary caching. The local SQLite database is used to store data for nearly 3 hours, which is based on the balance between data processing timeliness and storage resources. When writing data, a suitable table structure is designed according to the data structure to ensure that the data can be accurately and efficiently stored.

[0074] Then, an Apache Kafka data stream processing pipeline is established, which cleans and aggregates device data in the SQLite relational database based on device identification, anomaly type, and time window dimensions. For example, set the time window to 5 minutes, that is, process the device data generated by the same device within 5 minutes.

[0075] During data cleaning, data is filtered based on preset rules, such as removing format errors or incomplete data. For aggregation operations, device data of the same device generated within a preset time window and of the same anomaly type are combined into one valid event record. For example, if a device has 5 consecutive "CPU usage too high" anomaly records within 5 minutes, it is combined into one record containing the first occurrence time, last occurrence time, and occurrence number.

[0076] Finally, the deduplicated valid event records are stored in a MySQL relational database. MySQL database has good data management and query performance, and is suitable for long-term storage of large amounts of data. During storage, a reasonable database table structure and index are designed according to data characteristics and query requirements to facilitate subsequent data query and analysis.

[0077] S4: Synchronize the valid event records in the MySQL database to the Elasticsearch search engine through a data synchronization tool to establish a device performance index analysis index.

[0078] In the specific implementation, first, the Logstash data collection engine is used to incrementally extract valid event records from the MySQL relational database. The Logstash configuration file sets the extraction time interval (such as every 10 minutes) and the extraction conditions (such as only extracting data added or modified since the last extraction), ensuring the accuracy and efficiency of data extraction.

[0079] Then, the extracted valid event records are processed for format conversion and field mapping. For example, the date format in the MySQL database is converted to a format recognized by Elasticsearch, and the field name is mapped to a name that conforms to the Elasticsearch index specification. This step ensures that the data can be correctly written to Elasticsearch and facilitates subsequent query and analysis.

[0080] Finally, the processed valid event records are batch written to the specified index of the Elasticsearch distributed search engine. During writing, according to the cluster configuration and performance optimization principles of Elasticsearch, the size and concurrency of batch writing are reasonably set to improve writing efficiency and avoid excessive pressure on the Elasticsearch cluster.

[0081] S5: Based on the device performance index analysis index, combined with the Kibana data visualization platform for data display, and using the TensorFlow machine learning framework for deep analysis, identifying the device performance law and generating prediction and warning information as the analysis result.

[0082] In the specific embodiment, based on the device performance index analysis index, the effective event record is extracted in the Elasticsearch distributed search engine, and the device data therein is read. The Kibana data visualization platform is used to build multi-dimensional visual dashboards, such as creating a device health heat map to represent the health status of different devices with different colors and shades; generating a failure prediction curve to show the trend of device performance over time and the predicted failure time point. Through these visual charts, the operation and maintenance personnel can intuitively understand the running status of the device.

[0083] Then, the TensorFlow machine learning framework is applied to analyze the device data. Specifically: the association rule algorithm is used to mine the spatio-temporal association between CPU usage and playback stuttering, for example, by analyzing a large amount of device data, it is found that when the CPU usage exceeds a certain threshold in a certain time period, the probability of playback stuttering increases significantly, thereby providing a basis for optimizing device performance; the LSTM time series prediction algorithm is used to predict the device hardware aging trend based on the last 30 days of historical device data.

[0084] The LSTM algorithm can effectively process time series data, predict the trend of device hardware performance indicators by learning from historical data, and provide 8-hour early warning of hardware aging risk, generate hardware aging risk warning information as the analysis result to provide to the operation and maintenance personnel, so as to take timely maintenance measures.

[0085] S6: Based on the analysis result, generate a structured analysis report.

[0086] In the specific embodiment, first, set a timing task (for example, every day at 2 o'clock in the morning), and periodically read the analysis results generated in step S5 through a special script or program, including hardware aging risk warning information, device performance law, etc.

[0087] Then, the RESTful API interface is used to automatically call the JasperReports report engine service, and the analysis results and the visual charts generated by the Kibana data visualization platform are combined. During the calling process, according to the template design specification of JasperReports, the analysis result data is filled into the corresponding report template position, and the visual chart is embedded in the report in the form of a picture.

[0088] Finally, a structured analysis report is generated and output in PDF format. The output PDF report contains detailed information on device performance analysis, visual charts, and risk warning information, providing comprehensive and intuitive device performance analysis materials for device management and maintenance personnel, significantly reducing the difficulty of device management and maintenance.

[0089] In this embodiment, an edge computing architecture is used to deploy Agent programs on terminals to collect key indicators, and a heartbeat packet and compression algorithm are combined to effectively reduce data traffic and reduce the performance burden of monitoring programs on devices. Through distributed data platform construction, a multi-level cache mechanism is established for local SQLite near 3-hour data and cloud ES cluster historical data, and a data cleaning pipeline is designed to automatically filter duplicate alarms. Based on Elasticsearch, an intelligent analysis engine is built, Kibana is used to realize visual display of device health degree heat map, and machine learning algorithm is used to identify CPU occupation and playback cardon association rules, to give an 8-hour early warning of hardware aging risk. The overall difficulty of device management and maintenance is significantly reduced, and efficient collection, accurate analysis and intelligent presentation of device performance data are realized.

[0090] As shown in Figure 2 The following is an embodiment of a terminal device performance data intelligent analysis system provided by the embodiments of the present disclosure. The system and the terminal device performance data intelligent analysis method of the above embodiments belong to the same inventive concept. Details not described in the embodiment of the terminal device performance data intelligent analysis system can be referred to the above embodiment of the terminal device performance data intelligent analysis method.

[0091] A terminal device performance data intelligent analysis system comprises:

[0092] A cloud rule management module is deployed on a cloud server, used to create, store and manage data collection rules, and issue rule configurations associated with device attributes to terminal devices through HTTP protocol.

[0093] A device-side data collection module is deployed on a terminal device, used to load data collection rules, collect system performance data in real time, perform calculation and judgment locally and generate standardized event records, and report after processing by compression algorithm through MQTT communication protocol.

[0094] A message middleware module uses RabbitMQ message middleware to receive and cache device data reported by terminal devices.

[0095] The cloud data processing module includes an SQLite relational database, an Apache Kafka data stream processing platform and a MySQL relational database, is used for consuming device data from the message middleware module, and performs hierarchical processing on the device data, including temporary caching, streaming cleaning, repeated alarm aggregation and effective event record persistent storage.

[0096] The data synchronization and analysis module includes a Logstash data collection engine, an Elasticsearch distributed search engine and a TensorFlow machine learning framework, is used for synchronizing the effective event records in the MySQL to the search engine, establishing a device performance index analysis index, and performing deep analysis and prediction.

[0097] The visualization and report generation module includes a Kibana data visualization platform and a JasperReports report engine, is used for multi-dimensional visual display based on the analysis results, and automatically generates a structured analysis report containing charts and statistical conclusions.

[0098] The terminal device performance data intelligent analysis system provided in the embodiment realizes real-time collection and local processing of device performance data through deployment of a terminal edge computing Agent, completes efficient data transmission by combining an MQTT protocol and a compression algorithm, realizes cleaning and aggregation of massive data by using a cloud multi-level cache architecture and a Kafka stream processing pipeline, performs deep analysis and trend prediction based on an Elasticsearch search engine and a TensorFlow machine learning framework, and finally outputs visual analysis results through an automatic report generation system, forms a complete closed loop from data collection to intelligent early warning, significantly improves the operation and maintenance efficiency of large-scale terminal devices, reduces the cost of manual intervention, and realizes a change in the operation and maintenance mode from passive response to active prevention.

[0099] Figure 3 A hardware structure schematic diagram of an electronic device for implementing various embodiments of the present application.

[0100] The terminal device performance data intelligent analysis method provided by the embodiments of the present application can be applied to an electronic device. Those skilled in the art can understand that the electronic device structure involved in the embodiments of the present application does not constitute a limitation on the electronic device, and the electronic device can include more or fewer components than the illustration, or combine certain components, or different component arrangements. In the embodiments of the present application, the electronic device includes but is not limited to a laptop computer, a desktop computer, a workstation, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of the present application described and / or claimed herein.

[0101] The electronic device can include a processor, an external memory interface, an internal memory, a universal serial bus (USB) interface, a charge management module, a power management module, a battery, a wireless communication module, an audio module, a speaker, a microphone, a sensor module, a key, a camera, a display screen, and a SIM card interface, etc.

[0102] The processor can include one or more processing units, such as: the processor can include a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Among them, different processing units can be independent devices, or can be integrated in one or more processors.

[0103] Among them, the processor can be the nerve center and command center of the electronic device. The controller can generate operation control signals according to instruction operation codes and timing signals to complete the control of fetching and executing instructions.

[0104] The processor can further include a memory that stores instructions and data. In some embodiments, the memory in the processor is a cache memory. The memory can hold instructions or data that the processor has recently used or is likely to use again. If the processor needs to use the instructions or data again, it can call them directly from the memory. This avoids repeated access and reduces the processor's latency, thus improving system efficiency.

[0105] The external memory interface can be used to connect an external memory card, such as a MicroSD card, to extend the storage capacity of the electronic device. The external memory card communicates with the processor through the external memory interface to implement data storage functions. For example, music, video, and other files are saved in the external memory card.

[0106] The internal memory can be used to store computer executable program code, which includes instructions. The processor executes various functional applications and data processing of the electronic device by running the instructions stored in the internal memory. The internal memory can include a program storage area and a data storage area. The internal memory can include a high-speed random access memory and can also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, a universal flash storage (UFS), etc.

[0107] The wireless communication function of the electronic device can be implemented through an antenna, a wireless communication module, a modem processor, and a baseband processor, etc.

[0108] The wireless communication module can provide a wireless communication solution including wireless local area networks (WLAN) (such as a wireless fidelity (Wi-Fi) network), Bluetooth (BT), a global navigation satellite system (GNSS), frequency modulation (FM), near field communication (NFC), infrared technology (IR), etc. applied to the electronic device.

[0109] The electronic device can implement audio functions, etc. through an audio module, a speaker, a receiver, a microphone, an earphone interface, and an application processor, etc.

[0110] The electronic device can implement a shooting function through an ISP, a camera, a video codec, a GPU, a display screen, and an application processor, etc.

[0111] The electronic device can realize display function through GPU, display screen and application processor.

[0112] The GPU is a microprocessor for image processing, connecting the display screen and the application processor. The GPU is used to perform mathematical and geometric calculations for graphics rendering. The processor can include one or more GPUs that execute program instructions to generate or change display information.

[0113] The display screen is used to display images, videos, etc. The display screen includes a display panel.

[0114] The electronic device realizes the terminal device performance data intelligent analysis method, which realizes the local collection and processing of device performance data through the deployment of terminal edge computing Agent, completes efficient data transmission combined with MQTT protocol and compression algorithm, constructs a multi-level cache architecture and Kafka stream processing pipeline in the cloud to realize massive data cleaning and aggregation, and performs deep learning and trend prediction based on Elasticsearch and TensorFlow intelligent analysis platform. Finally, the multi-dimensional analysis results are output through the automatic report generation system, which reduces the cloud load and network transmission cost, improves the data processing efficiency and storage utilization, realizes the precise early warning and predictive maintenance of device failure, and significantly improves the efficiency of large-scale terminal device management and business continuity.

[0115] In the storage medium provided in the application, a program product capable of realizing the terminal device performance data intelligent analysis method is stored.

[0116] The terminal device performance data intelligent analysis method comprises:

[0117] The cloud server manages the data collection rules, and the terminal device obtains the data collection rules associated with the device properties from the cloud server through the HTTP protocol;

[0118] The terminal device collects system performance data in real time according to the data collection rules through the monitoring Agent program, performs calculation and judgment locally to generate standardized event records, processes the device data through the compression algorithm, and reports to the cloud message middleware through the MQTT protocol;

[0119] The cloud data processing service consumes the device data in the message middleware, adopts a hierarchical processing mechanism, stores the device data in the local SQLite database first, and then aggregates and processes the repeated alarms through the data cleaning pipeline to generate unique effective event records and store them in the MySQL database;

[0120] The data synchronization tool synchronizes the effective event records in the MySQL database to the Elasticsearch search engine to establish a device performance index analysis index;

[0121] Based on the device performance index analysis index, data display is performed in combination with the Kibana data visualization platform, and deep analysis is performed by using a TensorFlow machine learning framework to identify device performance rules and generate prediction and early warning information as an analysis result.

[0122] Based on the analysis result, a structured analysis report is generated.

[0123] In some possible implementation manners, the terminal device performance data intelligent analysis method of the present disclosure can be implemented in the form of a program product, which includes program codes for causing a terminal device to perform the steps according to various exemplary embodiments of the present disclosure described in the “Exemplary Method” part of the present specification when the program product is run on the terminal device.

[0124] The storage medium of the present disclosure can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium may, for example, be but is not limited to an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination thereof. More specific examples (non-exhaustive list) of the readable storage medium include an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0125] The above description of disclosed embodiments enables a person skilled in the art to implement or use the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure will not be limited to these embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for intelligent analysis of terminal device performance data, characterized in that, The application comprises the following steps: Data collection rules are managed by a cloud server, and terminal devices obtain data collection rules associated with their own device attributes from the cloud server through HTTP protocol; Terminal devices collect system performance data in real time according to data collection rules through a monitoring Agent program, perform calculation and judgment locally to generate standardized event records, generate device data after compression algorithm processing, and report the device data to a cloud message middleware through MQTT protocol in an asynchronous manner; Device data in the cloud message middleware is consumed by a cloud data processing service, a hierarchical processing mechanism is adopted, device data is first stored in a local SQLite database, repeated alarms are aggregated for processing through a data cleaning pipeline, unique and effective event records are generated and stored in a MySQL database; Effective event records in the MySQL database are synchronized to an Elasticsearch search engine through a data synchronization tool to establish a device performance index analysis index; Based on the device performance index analysis index, data is displayed in combination with a Kibana data visualization platform, deep analysis is performed using a TensorFlow machine learning framework, device performance rules are identified, and prediction and early warning information is generated as an analysis result; Based on the analysis result, a structured analysis report is generated; The terminal device collects system performance data in real time according to data collection rules through a monitoring Agent program, performs calculation and judgment locally to generate standardized event records, and generates device data after compression algorithm processing, and reports the device data to a cloud message middleware through MQTT protocol in an asynchronous manner, which comprises the following steps: On the terminal device, a monitoring Agent program is loaded with data collection rules through local deployment, system performance indicators are monitored in real time, and local calculation and judgment are performed according to corresponding threshold values in the data collection rules, abnormal system performance indicators are identified, the type of abnormality is determined, and standardized event records are generated based on the device identifier of the terminal device; the system performance indicators include CPU usage, memory occupancy, frame rate and decoding time consumption; LZ4 compression algorithm is used to compress the standardized event records to generate device data; A Paho MQTT client is used to publish the device data to a cloud RabbitMQ message middleware, and a heartbeat packet containing device state information is sent periodically.

2. The method of claim 1, wherein, The cloud server manages data collection rules, and terminal devices obtain data collection rules associated with their own device attributes from the cloud server through HTTP protocol, which comprises the following steps: Data collection rules are created on the cloud server, and the data collection rules include CPU usage threshold, memory occupancy threshold, frame rate threshold and decoding time threshold; Data collection rules are associated and bound with device attributes, and the data collection rules are published; the device attributes include device model, device name and device unique identifier; When the terminal device starts, the terminal device obtains the published data collection rules matching its own attributes from a Nginx reverse proxy server through an HTTP request; the terminal device includes but is not limited to a central control screen, an intelligent gateway and an outdoor large screen.

3. The method of claim 1, wherein, The device data in the message middleware is consumed by the cloud data processing service, a hierarchical processing mechanism is adopted, the device data is first stored in a local SQLite database, and then the repeated alarms are aggregated by a data cleaning pipeline to generate a unique effective event record which is stored in a MySQL database, including: Device data is obtained from the RabbitMQ message middleware by the cloud data processing service, and the device data is written into a local SQLite relational database for temporary caching; An Apache Kafka data stream processing pipeline is established, and the device data in the SQLite relational database is cleaned and aggregated based on device identification, abnormal type and time window dimension; The device data of the same device generated in a preset time window and of the same abnormal type is merged into one effective event record; The deduplicated effective event record is persistently stored in a MySQL relational database.

4. The method of claim 3, wherein, The effective event record in the MySQL database is synchronized to an Elasticsearch search engine by a data synchronization tool, an equipment performance index analysis index is established, including: The effective event record is incrementally extracted from the MySQL relational database by a Logstash data collection engine; The extracted effective event record is processed for format conversion and field mapping; The processed effective event record is batch written into a specified index of the Elasticsearch distributed search engine.

5. The method of claim 4, wherein, Based on the equipment performance index analysis index, data display is performed in combination with a Kibana data visualization platform, and deep analysis is performed by using a TensorFlow machine learning framework to identify equipment performance rules and generate prediction and warning information as analysis results, including: Based on the equipment performance index analysis index, the effective event record is extracted in the Elasticsearch distributed search engine, the device data is read, and a multi-dimensional visual board is constructed by using the Kibana data visualization platform to generate a visual chart; The TensorFlow machine learning framework is applied to analyze the device data, the association rule algorithm is used to mine the space-time association between CPU usage and card lag, the LSTM time series prediction algorithm is used to predict the device hardware aging trend based on the historical device data of the past 30 days, and hardware aging risk warning information is generated as the analysis result.

6. The method of claim 5, wherein, Based on the analysis result, a structured analysis report is generated, including: The analysis result is read regularly, the JasperReports report engine service is automatically called by a RESTful API interface, the analysis result and the visual chart generated by the Kibana data visualization platform are combined, a structured analysis report is generated, and the output is in PDF format.

7. A terminal device performance data intelligent analysis system, characterized by, The system adopts the terminal device performance data intelligent analysis method according to any one of claims 1 to 6; The system includes: A cloud rule management module is deployed on a cloud server, used to create, store and manage data collection rules, and send rule configurations associated with device attributes to terminal devices through an HTTP protocol. A device-side data collection module is arranged in the terminal device, used for loading data collection rules, collecting system performance data in real time, calculating and judging locally to generate standardized event records, and reporting the standardized event records to the cloud side through the MQTT communication protocol after compression algorithm processing; A message middleware module is used for receiving and caching device data reported by the terminal device, and adopts RabbitMQ message middleware; A cloud-side data processing module includes a SQLite relational database, an Apache Kafka data stream processing platform, and a MySQL relational database, used for consuming device data from the message middleware module, and performing hierarchical processing on the device data, including temporary caching, stream cleaning, repeated alarm aggregation, and effective event record persistent storage; A data synchronization and analysis module includes a Logstash data collection engine, an Elasticsearch distributed search engine, and a TensorFlow machine learning framework, used for synchronizing effective event records in the MySQL to the search engine, establishing device performance index analysis, and performing deep analysis and prediction; A visualization and report generation module includes a Kibana data visualization platform and a JasperReports report engine, used for multi-dimensional visualization based on analysis results, and automatically generating a structured analysis report including charts and statistical conclusions.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the steps of the terminal device performance data intelligent analysis method according to any one of claims 1 to 6.

9. A storage medium having stored thereon a computer program, characterized in that The computer program is executed by the processor to implement the steps of the terminal device performance data intelligent analysis method according to any one of claims 1 to 6.

Images