A historical data analysis platform

Abstract

The present invention provides a historical big data analysis platform, comprising: a data access module for acquiring historical data collected by a data monitoring device and transmitting the historical data to a historical database; a data report generation module for structurally organizing the historical data in the historical database to generate a data report and performing data analysis on the data; a data display module for performing 3D visualization conversion on the data visualization report to generate a high-dimensional visualization model; and a data result processing module for establishing a tree distribution model of an abnormal event solution strategy based on the high-dimensional visualization model according to the high-dimensional visualization model. The beneficial effects are as follows: The present invention realizes data visualization and automatic generation of abnormal event strategies, which not only saves costs and improves efficiency for customers, but more importantly, creates value for customers.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly relates to a historical data analysis platform. Background Art

[0002] The development of technology and the update of techniques have provided powerful impetus for the informatization of enterprises. In this process, with the emergence of new business and new demands of enterprises, the data within enterprises is increasing. New data is constantly generated, and after many businesses are completed, the entire processed data of the business becomes historical data. In the historical data, there are problems that occur during the implementation of the business, and of course, there are also corresponding solutions. However, in the prior art, historical data is only simply stored or simply summarized and analyzed, and the results of the analyzed data cannot be intuitively displayed. Therefore, historical data is often overlooked data and is rarely reused.

[0003] Under the existing conditions, when an emergency occurs, people often actively retrieve relevant data from various departments based on their accumulated experience and knowledge reserves, and then clarify the internal connections between various data according to experience. Based on this, a processing decision is formed. In this case, when the processing decision is generated, the development trend of the emergency is likely to reach an unsolvable situation.

[0004] Moreover, in the prior art, the method of processing data by people through experience is difficult to quickly process the increasingly large amount of data, and it is difficult to consider all relevant factors as a whole, thus making it difficult to give a relatively reasonable processing decision in a timely manner. For example, after an attack event occurs, in a network system, when the number of servers or computers attacked by an enterprise reaches a certain level, for example, all the enterprise's office electronic devices, if there happens to be someone with historical experience in solving relevant events, it is easy to solve the event. However, if there is no encounter with a similar attack event, and then looking for similar historical events encountered by others, analyzing the historical events, and then generating a strategy often easily causes delays. Therefore, analyzing historical data to generate a solution strategy that can be retrieved at any time is also a direction that needs to be explored. Summary of the Invention

[0005] The present invention provides a historical data analysis platform to solve the situation where the method of processing data by people through experience is difficult to quickly process the increasingly large amount of data, and it is difficult to consider all relevant factors as a whole, thus making it difficult to give a relatively reasonable processing decision in a timely manner.

[0006] A historical big data analysis platform includes:

[0007] A data access module: used to obtain historical data collected by a data monitoring device and transmit the historical data to a historical database;

[0008] Data report generation module: used to structurally organize the historical data in the historical database, generate a data report, and perform data analysis on the data;

[0009] Data display module: used to perform 3D visualization conversion on the data visualization report to generate a high-dimensional visualization model;

[0010] Data result processing module: used to establish a tree distribution model of an abnormal event solution strategy based on high-dimensional visualization according to the high-dimensional visualization model.

[0011] As an embodiment of the present invention: the data access module:

[0012] Data distribution unit: used to determine different monitoring targets according to the data monitoring device, and generate a historical data distribution map based on the monitoring targets; wherein,

[0013] The historical data distribution map is used to determine the generation source of historical data and collect historical data;

[0014] Collection unit: used to determine the service corresponding to the historical data according to the historical data distribution map, and establish a collection channel based on different historical data;

[0015] Database unit: used to establish a historical database of the corresponding type according to the collection channel; wherein,

[0016] The historical database is composed of four visual layers, and the four visual layers are: data attribute layer, data structure layer, data package layer, and data rendering layer;

[0017] The data attribute layer is used to store the attribute data of historical data;

[0018] The data structure layer is used to store the structure data of historical data;

[0019] The data package layer is used to store the logical data of historical data;

[0020] The rendering layer is used to store preset 3D visualization rendering data.

[0021] As an embodiment of the present invention: the data report generation module includes:

[0022] Statistical report unit: used to structurally analyze the historical data types according to preset standards, and generate a data directory of the corresponding type; wherein,

[0023] The preset factors include business type determination criteria, business cause-and-effect differentiation criteria, business process division criteria, business distribution scope criteria, business relevance determination criteria, and work order type identification criteria;

[0024] The data statistical report is subjected to granular analysis according to a preset period;

[0025] The preset period includes: daily period, monthly period, and annual period;

[0026] The preset period includes

[0027] Custom template report unit: used to create report templates for various purposes, and integrate with the data statistical report to automatically generate reports;

[0028] Data report analysis unit: used to retrieve various reports in the statistical report unit for data analysis; among them,

[0029] The data analysis includes business type analysis, business process analysis, business cause-and-effect analysis, business distribution analysis, business association analysis, and business work order analysis.

[0030] As an embodiment of the present invention: the data report analysis unit includes:

[0031] Data attribute analysis subunit: used to determine the retrieval method of historical data with different attributes according to the historical database, and perform business type analysis and business process analysis according to the data attributes;

[0032] Data structure analysis subunit: used to determine the composition architecture and execution logic of historical data with different structures and different logics according to the historical database, and perform business cause-and-effect analysis and business association analysis;

[0033] Data range analysis subunit: used to determine the business types and execution logics of historical data with different attributes and different logics according to the historical database, and perform business distribution analysis;

[0034] Business work order analysis subunit: used to determine the completion degree evaluation data of different businesses according to the historical database, and perform business work order analysis according to the completion degree evaluation data;

[0035] The business work order analysis subunit further includes the following steps:

[0036] S1: Obtain business work orders in historical data, evaluate the work order completion degree, and determine problem work orders and non-problem work orders;

[0037] S2: Perform a security mark on the non-problem work orders;

[0038] S2: Investigate the problem work order to determine the business information of the problem work order; among them,

[0039] The business information includes: configuration information, supplier information of the product, technical description of the product, and work order exception information;

[0040] S3: Determine the cause of the exception of the problem work order according to the business information, and determine the solution strategy;

[0041] S4: Analyze the solution strategy of the problem work order according to the solution strategy.

[0042] As an embodiment of the present invention: The data display module includes:

[0043] Chart conversion unit: used to convert the visualization report into a 3D visualization image through image segmentation, feature extraction, two-dimensional gradient operation, and image format reorganization;

[0044] 3D display unit: used to perform high-dimensional superposition on the historical data according to the 3D visualization image to generate a high-dimensional visualization model.

[0045] As an embodiment of the present invention: The 3D display unit includes:

[0046] Visual image spatial arrangement sub-unit: used to build an infinite three-dimensional space through the positional relationship and spatial arrangement of visual elements;

[0047] Stereo simulation camera sub-unit: used to import the 3D visualization image into the infinite three-dimensional space, and establish a high-dimensional symmetric model based on the simulation camera in the infinite three-dimensional space to form a high-dimensional visualization model based on depth simulation.

[0048] As an embodiment of the present invention: The data result processing module includes:

[0049] Workbench unit: used to display different historical data analysis results through the high-dimensional visualization model to generate business simulation operation scenarios corresponding to different historical data;

[0050] Abnormal event judgment unit: used to judge whether an abnormal event occurs according to the business simulation operation scenario and the business operation result in the historical data;

[0051] Strategy generation unit: used to extract the solution strategy corresponding to the event in the historical data and optimize the strategy when an abnormal event can occur;

[0052] Tree-shaped distribution unit: used to arrange the solution strategy and the business behavior corresponding to the abnormal event in a parallel tree shape.

[0053] As an embodiment of the present invention: The system further includes:

[0054] Data particle model construction module: used to determine particle distribution information by dividing historical data into grid particles through the WRF model;

[0055] Attribute assignment module: used to assign attributes to data particles of each historical data according to the example distribution information;

[0056] Particle association module: used to determine the association between different examples according to the attribute assignment, and connect the particles according to the association to generate a particle model;

[0057] Verification module: used to comprehensively verify the data of the high-dimensional visualization model according to the particle model and output a verification result.

[0058] As an embodiment of the present invention: The system further includes:

[0059] Data acquisition module: used to analyze the real-time data collected by the data monitoring device and generate a visualization report based on real-time services;

[0060] Data calling module: will call the same type of business data as the real-time service in the tree distribution model according to the visualization report;

[0061] Data comparison module: used to compare the same type of business data and real-time data to determine whether there is data anomaly;

[0062] Anomaly discovery module: used to retrieve anomaly data when there is data anomaly and determine the corresponding anomaly event solution strategy through the tree distribution model.

[0063] As an embodiment of the present invention: The anomaly discovery module includes:

[0064] Visualization information determination unit: used to determine the data attributes and data dimensions of the anomaly data according to the anomaly data;

[0065] Data identification unit: used to determine the distribution position of the anomaly event in the tree distribution model according to the data dimension and data tree;

[0066] Rule comparison module: used to determine the comparison rules of the historical data corresponding to the distribution position according to the distribution position, perform rule determination on the anomaly data according to the comparison rules, and obtain the corresponding solution strategy after all comparison rules are met.

[0067] The beneficial effects of the present invention are as follows: The historical data analysis platform is a data value discovery and utilization platform, which provides customers with professional, agile, and easy-to-use big data analysis, mining, and visualization tools. Aiming at data value addition, the platform provides customers with professional historical data processing and analysis methods to meet the needs of different roles in the organization for data value mining and application. It is oriented to data analysis personnel and data value utilizers at all levels of the enterprise, integrating data visualization exploration, in-depth data analysis, and model application development. First, it can access and process multiple data sources; it can achieve various capabilities such as data access, data processing, data analysis, result application, and issuing of exception event handling strategies; users can conduct intuitive analysis through data visualization and also discover the deep rules hidden in the data through data mining. It can be jointly used by enterprise leaders, business personnel at all levels, and technical personnel. The product not only saves costs and improves efficiency for customers, but more importantly, it creates value for customers.

[0068] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] The accompanying drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention, but do not constitute a limitation to the present invention. In the accompanying drawings:

[0070] Figure 1 It is a flowchart of a historical big data analysis platform in an embodiment of the present invention;

[0071] Figure 2 It is a schematic diagram of a data access module of a historical big data analysis platform in an embodiment of the present invention;

[0072] Figure 3 It is a step diagram of work order management report analysis of a historical big data analysis platform in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0073] The following describes the preferred embodiments of the present invention with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only for the purpose of illustrating and explaining the present invention, and are not used to limit the present invention.

[0074] Embodiment 1:

[0075] The embodiment of the present invention provides a historical big data analysis platform, as Figure 1 shown, including:

[0076] A data access module: used to obtain historical data collected by data monitoring devices and transmit the historical data to a historical database;

[0077] Data report generation module: used to structurally organize the historical data in the historical database, generate a data report, and perform data analysis on the data;

[0078] Data display module: used to perform 3D visualization conversion on the data visualization report to generate a high-dimensional visualization model;

[0079] Data result processing module: used to establish a tree distribution model of an abnormal event solution strategy based on high-dimensional visualization according to the high-dimensional visualization model.

[0080] The working principle of the above technical solution is as follows: In the prior art, for the processing of monitoring data in a data center, the utilization of historical data is very limited, or it is directly deleted without utilization. Most of them are simple summaries or storage as backup information. However, with the development of enterprises, it is very likely to encounter the same business and the same situation as the historical data. At this time, the historical data is the best reference data. Therefore, the present invention collects historical data according to data monitoring devices. Since all business data is collected through data monitoring devices, historical data is collected through data monitoring devices and a dedicated historical database is established for storage. The purpose of generating a data report is to comprehensively analyze the historical data. When the existing event is the same as the corresponding event in the historical data, all comparisons are based on these data reports and the comparison of the statistical analysis content of different data. The 3D visualization conversion is a virtual state conversion in the present invention. For example, when the data report shows the fluctuation state of data traffic, the data report may show the details of data traffic and the traffic state at every moment; after the 3D visualization conversion, the traffic state and the fluctuation state will be converted into the same as the water flow in the pipeline. When the data fluctuates, the water flow fluctuates, and the state of the data is the state of the water flow. Finally, it is converted into a high-dimensional visualization model, which is a technology based on high-dimensional visualization. That is, on the basis of 3D conversion, a high-dimensional visualization graph is generated. From this graph, abnormal events can be analyzed more clearly and the processing strategy of abnormal events can be predicted.

[0081] The beneficial effects of the above technical solution are as follows: The historical data analysis platform is a data value discovery and utilization platform that provides customers with professional, agile, and easy-to-use big data analysis, mining, and visualization tools. The platform aims at data value addition, provides customers with professional historical data processing and analysis methods, and meets the needs of different roles in the organization for data value mining and application. It is targeted at data analysts and data value utilizers at all levels in the enterprise, integrating data visualization exploration, in-depth data analysis, and model application development. First, it can access and process multiple data sources; it realizes multiple capabilities such as data access, data processing, data analysis, result application, and the formulation of exception event handling strategies; users can conduct intuitive analysis through data visualization and also discover the in-depth rules hidden in the data through data mining. It can be jointly used by enterprise leaders, business personnel at all levels, and technical personnel. The product not only saves costs and improves efficiency for customers, but more importantly, creates value for customers.

[0082] Embodiment 2:

[0083] In a specific embodiment, as shown in the appendix Figure 2 The data access module includes:

[0084] Data distribution unit: used to determine different monitoring targets according to the data monitoring device and generate a historical data distribution map based on the monitoring targets; among them,

[0085] The historical data distribution map is used to determine the generation source of historical data and collect historical data;

[0086] Collection unit: used to determine the business corresponding to the historical data according to the historical data distribution map and establish a collection channel based on different historical data;

[0087] Database unit: used to establish a historical database of the corresponding type according to the collection channel; among them,

[0088] The historical database consists of four visual layers, and the four visual layers are: data attribute layer, data structure layer, data wrapper layer, and data rendering layer;

[0089] The data attribute layer is used to store the attribute data of historical data;

[0090] The data structure layer is used to store the structure data of historical data;

[0091] The data wrapper layer is used to store the logical data of historical data;

[0092] The rendering layer is used to store the preset 3D visualization rendering data.

[0093] The working principle of the above technical solution is as follows: Since the present invention mainly processes the data collected by the monitoring devices in the book data center, the present invention determines the types of data to be monitored based on the monitoring purpose, and establishes a historical data distribution map based on the type of this data and the distribution area of the data. Thus, when comparing with the existing abnormal data, it is possible to more quickly discover where abnormal events occur. During data collection, data is collected through different collection channels, which is more convenient and more targeted. The data is more comprehensive and accurate. The database is established according to the collection channels and data types of the collection, facilitating the storage of different types of data. Finally, since the present invention needs to perform 3D visualization conversion and high-dimensional visualization conversion, corresponding visual technologies are required, and different visual layers store different data, facilitating data retrieval, thereby more quickly establishing 3D, high-dimensional, and data charts.

[0094] The beneficial effects of the above technical solution are as follows: Through the above database, a large amount of data can be sorted out and presented in the form of reports, making the data easier to express and understand, and providing an intuitive and clear data source for subsequent data result analysis.

[0095] Example 3:

[0096] In a specific embodiment, the statistical report unit includes:

[0097] The statistical report unit: used to perform a structured analysis on the historical data types according to preset criteria, and generate data charts of corresponding types; wherein,

[0098] The preset criteria include data type determination criteria, data causality discrimination criteria, data process division criteria, data range criteria, data correlation determination criteria, and work order type identification criteria;

[0099] The custom template report unit: used to create various report templates for different purposes, and integrate them into the data report to automatically generate reports;

[0100] The data report analysis unit: used to retrieve various reports in the statistical report unit for data analysis; wherein,

[0101] The data analysis includes business type analysis, business process analysis, business causality analysis, business dispersion analysis, business correlation analysis, and business work order analysis.

[0102] The working principle of the above technical solution is as follows: When generating a data report, the present invention conducts a structural analysis according to the historical data types, can perform professional analysis on different types of data, and adopts dedicated analysis techniques. Then, preset standards are used to generate a data report according to the data standards in the prior art, including data type determination standards, data causality discrimination standards, data process division standards, data range standards, data correlation determination standards, and work order type identification standards. Corresponding type charts are generated according to these standards, such as cause-and-effect charts for data anomalies or data development, icons for data process division rules, icons for data monitoring ranges, icons showing which devices or data are associated with each other, and the historical functions corresponding to each data. In the technical implementation, each data monitoring target is a work order.

[0103] The beneficial effects of the above technical solution are as follows: By generating different types of data reports in the above manner, automated report generation is achieved, and specific analysis of data can be carried out through the reports.

[0104] Embodiment 4:

[0105] In a specific embodiment, as Figure 3 shown,

[0106] The data report analysis unit includes:

[0107] The data attribute analysis subunit is used to determine the retrieval method of historical data with different attributes according to the historical database, and conduct business type analysis and business process analysis according to the data attributes;

[0108] The data structure analysis subunit: is used to determine the composition architecture and execution logic of historical data with different structures and different logics according to the historical database, and conduct business causality analysis and business association analysis;

[0109] The data range analysis subunit: is used to determine the business types and execution logics of historical data with different attributes and different logics according to the historical database, and conduct business dispersion analysis;

[0110] The business work order analysis subunit: is used to determine the completion degree evaluation data of different businesses according to the historical database, and conduct business work order analysis according to the completion degree evaluation data;

[0111] As shown in the appendix Figure 3 shown, the business work order analysis subunit further includes the following steps:

[0112] S1: Obtain the business work orders in the historical data, conduct work order completion degree evaluation, and determine problem work orders and non-problem work orders;

[0113] S2: Mark the non-problem work orders as safe;

[0114] S2: Investigate the problem work order to determine the business information of the problem work order; wherein,

[0115] The business information includes: configuration information, supplier information of the product, technical description of the product, and work order exception information;

[0116] S3: Determine the root cause of the problem work order based on the business information, and determine the solution strategy;

[0117] S4: Analyze the solution strategy of the problem work order according to the solution strategy.

[0118] The working principle of the above technical solution is: When performing data analysis, historical data is analyzed as corresponding services according to the tree, structure, and logic of the data. Different work orders are used to track and process different types of transactions. How to use the work order is determined according to specific requirements. A work order can be the work tasks of a project.

[0119] The goal of work order analysis is to minimize business failures caused by infrastructure failures and prevent the recurrence of events related to these errors.

[0120] Therefore, the focus of the present invention is also to discover the root cause of the event and then take measures to improve or correct this situation. An event is the potential cause of some or multiple accidents. The process of analyzing historical data problems is to minimize the impact on customers caused by defects or negligence such as service infrastructure, human errors, and external events, and prevent them from recurring.

[0121] When it comes to specific implementation:

[0122] When evaluating work orders, the submitted work orders need to be evaluated. There are two evaluation results. One is that the problem actually exists and needs to be solved. The operation and maintenance personnel accept the work order and click the "Start Investigation" hyperlink, and the corresponding work order status changes to "Under Investigation". The other is that it is considered that there is no need to solve or the description of the submitted work order is unclear, and the investigation is refused. Click "Refuse Investigation", and the corresponding work order status changes to "Rejected". "Start Investigation" is an action. After execution, the work order status correspondingly changes to "Under Investigation", indicating that the cause of the problem is currently being investigated and diagnosed. The investigation and diagnosis may be a repeated process and need to be carried out multiple times, and each repetition is closer to the solution we want. No remarks need to be entered when executing the start investigation. It is just an action used to change the work order status from "Evaluating" to "Under Investigation" to indicate that the current operation and maintenance personnel have started to handle the problem. Here, the system will be linked with configuration management and infrastructure components. It can be associated with configuration information, product supplier information, product technical descriptions, error information, etc., and the operating conditions information of the nodes where the problem occurs, such as availability reports, performance reports, etc., to provide a basis for staff to analyze the cause of the problem.

[0123] When the problem analyst finds the cause of the problem and finds a temporary or permanent solution to solve the problem, the investigation can be ended. In the operation and maintenance specification, the problem is converted to the "Known Error" state at this time. When ending the investigation, the cause of the problem found through the investigation must be entered. After the action is executed, the work order status changes to "Known Error".

[0124] The beneficial effects of the above technical solution are as follows: By accurately analyzing data, the work efficiency of data center personnel is improved, and the data report analysis unit promotes enterprises to reduce operating costs and effectively improve customer retention, playing a crucial role in enterprise management. It can play the role of communication between the front and back desks, cross-departmental collaboration, and cross-enterprise collaboration.

[0125] Example 5:

[0126] The data display module includes:

[0127] Chart conversion unit: used to process the visualization report through image segmentation, feature extraction, two-dimensional gradient operation, and image format reorganization, and convert it into a 3D visualization image;

[0128] 3D display unit: used to perform high-dimensional superposition on the historical data according to the 3D visualization image to generate a high-dimensional visualization model.

[0129] The working principle of the above technical solution is as follows: The chart conversion unit converts chart data into 3D visualization images through two-dimensional gradient operation and format reorganization; then, a high-dimensional visualization model is generated through high-dimensional superposition. The high-dimensional visualization model is a technology that converts complex data under the superposition of 3D visualization images into a scatter matrix and reduces the image complexity of spatial coordinates. There is a technology model that blurs or directly filters out irrelevant data fully recorded in some chart data and does not display it as the main feature of the data. The 3D visualization image is displayed through a naked-eye 3D display screen, while the high-dimensional visualization model is a spatial model, a model for comprehensive analysis of historical data, and can also be charted.

[0130] In the process of converting chart data into 3D visualization images through two-dimensional gradient operation and format reorganization, the following steps are included:

[0131] Establish a data set of chart data:

[0132] X = {x1, x2, …, x i}; Y = {y1, y2, …, y i}; Z = {z1, z2, …, z i}

[0133] X and Y are respectively the sample data of the chart data; for example, in tabular data, X represents row data; Y represents column data; Z represents the result data; if it is a coordinate graph, X represents the abscissa; Y represents the ordinate; Z represents the vertical coordinate; i is a positive integer, representing a data; there will be a total of n data;

[0134] After the data set is determined, we perform 3D conversion through the following formula:

[0135]

[0136]

[0137] The above

[0138] F(X, Y, Z) represents the 3D conversion function;

[0139] According to this 3D conversion function, a spatial framework is established, and logical data and 3D visualization rendering data are added to form a 3D visualization image.

[0140] The beneficial effects of the above technical solution are as follows: In the above technology, chart data is converted into three-dimensional data, and the three-dimensional data is then converted into a high-dimensional visualization model, realizing the management function of data centers, cabinets, and various devices based on a three-dimensional environment, and constructing a visualization platform for data center environments, devices, and management information.

[0141] Example 6:

[0142] As a specific embodiment of the present invention: the 3D display unit includes:

[0143] Visual image spatial arrangement subunit: used to build an infinite three-dimensional space through the positional relationship and spatial arrangement of visual elements;

[0144] Stereo simulation camera subunit: used to import the 3D visualization image into the infinite three-dimensional space, and establish a high-dimensional symmetric model based on the simulation camera in the infinite three-dimensional space, constituting a high-dimensional visualization model based on depth simulation.

[0145] The working principle of the above technical solution is as follows: the present invention builds an infinite three-dimensional space, establishes coordinate systems at different positions in the space, imports the 3D visualization image into the infinite three-dimensional space, and then establishes a high-dimensional symmetric space model. When this space model is transformed through the simulation camera, dimensional fitting is performed on the 3D visualization image. During the fitting process, some identical data are fitted together, and then an abstract image in space is generated, such as a scatter matrix space. This is because during the fitting process, after the identical data are cancelled out, only dot-like data remain, presenting a scatter state.

[0146] The beneficial effect of the above technical solution is as follows: the present invention builds an infinite three-dimensional space, converts chart data into 3D images for display, and then converts it into a high-dimensional visualization model, making the data easier to manage. For the abstract state of the data, that is, the abnormal state, the management transparency is more obvious, and further effectively improves the efficiency of asset management and monitoring management, truly realizing a three-dimensional and visual new generation historical data analysis platform.

[0147] Example 7:

[0148] The data result processing module includes:

[0149] Workbench unit: used to display different historical data analysis results through the high-dimensional visualization model, and generate business simulation operation scenarios corresponding to different historical data;

[0150] Abnormal event judgment unit: used to judge whether an abnormal event occurs according to the business simulation operation scenario and the business operation results in the historical data;

[0151] Strategy generation unit: used to extract the solution strategies corresponding to the events in the historical data and optimize the strategies when an abnormal event can occur;

[0152] Tree distribution unit: used to arrange the solution strategies and the business behaviors corresponding to the abnormal events in a parallel tree layout.

[0153] The working principle of the above technical solution is as follows: The workbench of the present invention is for simulating operation scenarios corresponding to different historical data. For example, when a traffic monitoring device monitors, the detection indicators of the traffic monitoring device and the real-time detected traffic graph. When judging abnormal events, we first need to determine whether abnormal events are reflected in the historical data. Therefore, through the simulation of the scenario, it can more vividly show whether there are abnormalities, and then call the corresponding solution series in the historical data to achieve policy optimization. Finally, the business abnormal events and corresponding business behaviors are arranged in a tree-like layout, that is, an abnormal situation occurs in the data center being monitored.

[0154] The beneficial effects of the above technical solution are as follows: When the present invention can perform 3D display, data from different systems can be presented on one interface, integrating the most commonly used functions and the most concerned business data of individual users, enabling a clear view of the operating status of each subsystem, and gathering them as shortcuts for users' daily work. Through the workbench unit, users can quickly locate the work to be carried out, improving the efficiency of system application and facilitating daily maintenance support work. However, there may be abnormal situations and solution strategies for the data. For these most-needed data, they will be arranged in an artistic layout, corresponding the events and strategies, which is convenient for quickly finding the events. During actual implementation: The present invention will also give an alarm based on the existing situation.

[0155] For stand-alone device alarms and cabinet-internal device alarms, alarm icons and device color change prompts will directly appear on the device. Click the alarm icon or the alarm device to display detailed alarm information; for cabinet-internal device alarms, alarm icons and cabinet color change prompts will appear on the corresponding cabinet. Click the alarm icon or the alarm device to display detailed alarm information. In the alarm state, double-click the cabinet to enter the cabinet, and the alarm device will show a device color change prompt. Click the device to display the detailed alarm information interface. The display of device alarms shows the requirements for displaying alarm information of computer room devices. All alarm information in the computer room should be displayed in real time on the 3D display interface of the computer room. The main types of alarm devices to be displayed are two: for stand-alone device alarms, alarm icons and device color change prompts will directly appear on this device. Click the alarm icon or the alarm device to display detailed alarm information. For cabinet-internal device alarms, alarm icons and cabinet color change prompts will appear on the corresponding cabinet. Click the alarm icon or the alarm device to display detailed alarm information. In the alarm state, double-click the cabinet to enter the cabinet, and the alarm device will show a device color change prompt. Click the device to display the detailed alarm information interface. Setting stand-alone device alarms and cabinet-internal device alarms can enable staff to find the faulty device faster, facilitating the staff to repair the device in a timely manner and enabling the platform to enter the normal operation state faster.

[0156] In actual real-time: Based on the business simulation operation scenario and the business operation results in historical data, determine whether an abnormal event occurs;

[0157] In this process, first, based on the real-time data model among business simulation operation scenarios;

[0158]

[0159] H represents the coefficient of data change in the business simulation operation scenario, and this coefficient changes with events; Q j，t represents the data status of the j-th device in the monitoring device at time t; R j，t represents the data characteristics of the data monitored by the j-th device in the monitoring device at time t; W j，t represents the change coefficient of the j-th device in the monitoring device at time t from the initial time; t0 represents the initial time; T represents the final time of monitoring; the time t represents real-time; j is a positive integer, j ∈ m, and m represents the total number of monitoring devices;

[0160] Based on historical data, establish a historical model for similar situations in historical data;

[0161]

[0162] Among them, D g represents the event type characteristics corresponding to the g-th historical data; N g represents the event content characteristics corresponding to the g-th historical data; represents the event result characteristics of the event type corresponding to the g-th historical data; V represents an abnormal event; g is a positive integer, g ∈ G, and G represents the total number of historical data;

[0163] Then, by comparing the two models with each other, determine whether there is an abnormal event;

[0164]

[0165] Among them, when P = 0, the data content of the historical event and the real-time simulation operation scenario are the same. Therefore, there is an abnormal event. In this process, the present invention first constructs a real-time data model diagram for the simulation operation scenario; then constructs a model for all abnormal events in historical data. After determining that this real-time abnormal event belongs to the abnormal events in historical data, it is determined whether it is an abnormal event by the method of fitting one by one.

[0166] Embodiment 8:

[0167] As a specific embodiment: The system further includes:

[0168] Data Particle Model Construction Module: used to determine particle distribution information by performing grid-based particle division on historical data through the WRF model;

[0169] Attribute Assignment Module: used to assign attributes to data particles of each historical data according to the example distribution information;

[0170] Particle Association Module: used to determine the association between different examples according to the attribute assignment, and perform particle connection according to the association to generate a particle model;

[0171] Verification Module: used to perform comprehensive data verification on the high-dimensional visualization model according to the particle model and output the verification result.

[0172] The principle of the above technical solution is as follows: The present invention also verifies the obtained model through grid-based particle division to ensure that the obtained model has high accuracy. In this process, the WRF model represents a forecasting model. For particles showing anomalies in data particles, forecasting extension is required. Then, after these particles are connected to form a particle model, they should have a similar spatial shape to the high-dimensional visualization model. However, because particleization only performs clear division of data, the high-dimensional particle model is more accurate in monitoring abnormal data.

[0173] The beneficial effect of the above technical solution is as follows: Through the particleization of historical data, the comprehensive data verification of the high-dimensional visualization model can be verified, thereby ensuring that the visualized data in the technology of the present invention is comprehensive and can meet various situations.

[0174] Example 9:

[0175] In a specific embodiment: The system further includes:

[0176] Data Acquisition Module: used to perform data analysis on the real-time data collected by the data monitoring device and generate a visualization report based on real-time services;

[0177] Data Invocation Module: will retrieve similar business data identical to the real-time service in the tree-shaped distribution model according to the visualization report;

[0178] Data Comparison Module: used to compare the similar business data and real-time data to determine whether there is data anomaly;

[0179] Anomaly Detection Module: used to retrieve abnormal data when there is data anomaly and determine the corresponding abnormal event solution strategy through the tree-shaped distribution model.

[0180] The working principle and beneficial effects of the above technical solution are as follows: The present invention further includes a data acquisition module, which collects real-time monitoring data. The most important purpose of the present invention is to use historical data as an abnormal event strategy library for existing data, so as to realize a quick solution for abnormalities during actual existing monitoring. Therefore, in this process, the present invention also generates visual reports for the monitoring data of real-time services. Since the tree distribution model is established based on a high-dimensional visualization model and has the basis for visualizing data reports, it is easier to compare existing data with historical data through these to determine whether there are data abnormalities, so as to find corresponding solution strategies.

[0181] Example 10:

[0182] The abnormal discovery module includes:

[0183] Visual information determination unit: used to determine the data attributes and data dimensions of the abnormal data according to the abnormal data;

[0184] Data identification unit: used to determine the distribution position of the abnormal event in the tree distribution model according to the data dimension and data tree;

[0185] Rule comparison module: used to determine the comparison rules for the historical data corresponding to the distribution position according to the distribution position, perform rule determination on the abnormal data according to the comparison rules, and obtain the corresponding solution strategy after all comparison rules are met.

[0186] The principle of the above technical solution is as follows: In the processing of abnormal data, the present invention will first, based on the data attributes and data dimensions, that is, what type of data the data belongs to and what abnormalities the abnormal data shows, be able to clarify in the tree distribution model which contents are corresponding abnormal data and which abnormalities are corresponding solution strategies. However, before extracting the corresponding strategy, a comparison is still needed. This comparison rule is mainly based on historical data for comparison. When comparing, because there is a 3D simulated image scene converted from historical data, the comparison rule can be set according to the simulated scene, and then compared with the real scene to obtain the strategy.

[0187] The beneficial effects of the above technical solution: Through the setting of comparison rules, first, the actual situation can be made more in line with the situation corresponding to historical data, and second, a more accurate solution strategy for the abnormal event corresponding to the abnormal data can be obtained.

[0188] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

Claims

1. A historical big data analysis platform, characterized in that, Including: Data access module: used to obtain historical data collected by data monitoring devices and transmit the historical data to the historical database; Data report generation module: used to structurally organize the historical data in the historical database, generate a data visualization report, and perform data analysis on the data; Data display module: used to perform 3D visualization conversion on the data visualization report to generate a high-dimensional visualization model; Data result processing module: used to establish a tree distribution model of an exception event solution strategy based on high-dimensional visualization according to the high-dimensional visualization model; The data display module includes: Chart conversion unit: used to process the data visualization report through image segmentation, feature extraction, two-dimensional gradient operation, and image format reorganization, and convert it into a 3D visualization image; 3D display unit: used to perform high-dimensional superposition on the historical data according to the 3D visualization image to generate a high-dimensional visualization model; The 3D display unit includes: Visual image space arrangement sub-unit: used to build an infinite three-dimensional space through the positional relationship and spatial arrangement of visual elements; Stereo simulation camera sub-unit: used to import the 3D visualization image into the infinite three-dimensional space, and establish a high-dimensional symmetric model based on the simulation camera in the infinite three-dimensional space to form a high-dimensional visualization model based on depth simulation; The platform further includes: Data collection module: used to perform data analysis on the real-time data collected by the data monitoring device to generate a visualization report based on real-time services; Data call module: will, according to the visualization report, retrieve the same type of business data as the real-time service in the tree distribution model; Data comparison module: used to compare the same type of business data and real-time data to determine whether there is data abnormality; Abnormality discovery module: used to retrieve abnormal data when there is data abnormality, and determine the corresponding abnormal event solution strategy through the tree distribution model.

2. The historical big data analysis platform according to claim 1, wherein The data access module: Data distribution unit: used to determine different monitoring targets according to the data monitoring device, and generate a historical data distribution map based on the monitoring targets; wherein, The historical data distribution map is used to determine the generation source of historical data and collect historical data; Collection unit: used to determine the business corresponding to historical data according to the historical data distribution map, and establish a collection channel based on different historical data; Database unit: used to establish a historical database of the corresponding type according to the collection channel; wherein, The historical database is composed of four visual layers, and the four visual layers are: data attribute layer, data structure layer, data package layer, and data rendering layer; The data attribute layer is used to store the attribute data of historical data; The data structure layer is used to store the structure data of historical data; The data package layer is used to store the logical data of historical data; The rendering layer is used to store preset 3D visualization rendering data.

3. A historical big data analysis platform according to claim 1, characterized in that, The data report generation module includes: Statistical report unit: used to structurally analyze the historical data type according to a preset standard and generate a data chart of the corresponding type; wherein, The preset standards include data type determination standards, data causality differentiation standards, data process division standards, data range standards, data relevance determination standards, and work order type identification standards; Custom template report unit: used to create report templates for various different purposes, integrate them into the data report, and automatically generate reports; Data report analysis unit: used to retrieve various reports in the statistical report unit for data analysis; among them, The data analysis includes business type analysis, business process analysis, business causality analysis, business dispersion analysis, business association analysis, and business work order analysis.

4. The historical big data analysis platform according to claim 3, wherein The data report analysis unit includes: Data attribute analysis subunit: used to determine the retrieval method of historical data with different attributes according to the historical database, and conduct business type analysis and business process analysis according to the data attributes; Data structure analysis subunit: used to determine the composition architecture and execution logic of historical data with different structures and different logics according to the historical database, and conduct business causality analysis and business association analysis; Data range analysis subunit: used to determine the business types and execution logics of historical data with different attributes and different logics according to the historical database, and conduct business dispersion analysis; Business work order analysis subunit: used to determine the completion degree evaluation data of different businesses according to the historical database, and conduct business work order analysis according to the completion degree evaluation data; The business work order analysis subunit further includes the following steps: S1: Obtain the business work orders in the historical data, conduct work order completion degree evaluation, and determine problem work orders and non-problem work orders; S2: Mark the non-problem work orders as safe; S2: Investigate the problem work orders to determine the business information of the problem work orders; among them, The business information includes: configuration information, product supplier information, product technical description, and work order exception information; S3: Determine the abnormal cause of the problem work order according to the business information, and determine the solution strategy; S4: Analyze the solution strategy of the problem work order according to the solution strategy.

5. A historical big data analysis platform according to claim 1, characterized in that, The data result processing module includes: Workbench unit: used to display different historical data analysis results through the high-dimensional visualization model, and generate business simulation operation scenarios corresponding to different historical data; Abnormal event judgment unit: used to judge whether an abnormal event occurs according to the business simulation operation scenario and the business operation results in the historical data; Strategy generation unit: used to extract the solution strategies corresponding to the events in the historical data and optimize the strategies when an abnormal event can occur; Tree-shaped distribution unit: used to arrange the solution strategies and the business behaviors corresponding to the abnormal events in a parallel tree shape.

6. The historical big data analysis platform according to claim 1, wherein The platform further includes: Data particle model construction module: used to divide historical data into grid particles through the WRF model to determine particle distribution information; Attribute assignment module: used to assign attributes to the data particles of each historical data according to the particle distribution information; Particle association module: used to determine the correlation between different particles according to the attribute assignment, and connect the particles according to the correlation to generate a particle model; Verification module: used to comprehensively verify the data of the high-dimensional visualization model according to the particle model, and output a verification result.

7. The historical big data analysis platform according to claim 1, wherein The anomaly discovery module includes: Visualization information determination unit: used to determine the data attributes and data dimensions of the abnormal data according to the abnormal data; Data identification unit: used to determine the distribution position of the abnormal event in the tree-shaped distribution model according to the data dimensions and data attributes; Rule comparison module: used to determine the comparison rule of the historical data corresponding to the distribution position according to the distribution position, perform rule determination on the abnormal data according to the comparison rule, and obtain the corresponding solution strategy after all comparison rules are met.

Images