Automobile size measurement system and method based on multi-dimensional correlation analysis

Abstract

The invention relates to the technical field of data analysis, in particular to an automobile dimension measurement system and method based on multi-dimensional correlation analysis, and the system comprises a code configuration module which displays the dimension type of dimension measurement data and an analysis rule based on a display interface; the data driving module obtains corresponding metadata configuration on the basis of the dimension type and the analysis rule of the size measurement data by the display interface; and the data acquisition module generates a data acquisition task based on the core metadata and the metadata configuration, acquires data integration information according to the data acquisition task and the plurality of groups of adapter modules, and stores the core metadata, the metadata configuration and the data integration information to a system database. According to the invention, the to-be-used customer can quickly complete the configuration of the multi-dimensional correlation analysis automobile dimension measurement system according to the own demand, the use threshold is reduced, and the usability of the system is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data analysis, and in particular to a vehicle size measurement system and method based on multi-dimensional correlation analysis. BACKGROUND

[0002] With the rapid development of the automobile industry, vehicle size measurement plays a crucial role in production, quality control and process optimization. However, existing vehicle size measurement and analysis systems face many technical challenges, especially in data collection, analysis dimension adjustment and intelligent correlation analysis.

[0003] Currently, traditional static measurement systems, rule engine-based measurement systems and machine learning-assisted measurement systems are commonly used. The traditional static measurement system uses fixed measurement rules and data models to predict vehicle part size measurement data. The rule engine-based measurement system processes and analyzes measurement data through prediction rules, while the machine learning-assisted measurement system collects size data of vehicle parts based on automated measurement equipment. Specifically, machine algorithms learn the corresponding rules from the collected size data, and generate a predictive mathematical model from the learned data to predict vehicle part size data measurement.

[0004] However, the acquisition of vehicle size measurement data mainly relies on traditional static acquisition, and in addition, users often need to rely on developers for complex configuration when adjusting analysis dimensions and correlation logic, which has a high technical threshold. SUMMARY

[0005] In order to reduce the use threshold of the system and enable non-technical users to easily complete configuration to achieve dynamic data collection, the present application provides a vehicle size measurement system and method based on multi-dimensional correlation analysis.

[0006] In a first aspect, the present application provides a vehicle size measurement system based on multi-dimensional correlation analysis, which adopts the following technical solution: A vehicle size measurement system based on multi-dimensional correlation analysis, comprising: a code configuration module, which displays the dimension type and analysis rule of the size measurement data based on a display interface; a data driving module, which is configured to configure the core metadata of the size measurement data and dynamically configure the dimension type and analysis rule of the size measurement data based on the display interface to obtain corresponding metadata configuration; A data collection module is connected with the data driven module in network to acquire the core metadata and the metadata configuration, and generates a data collection task based on the core metadata and the metadata configuration, acquires data integration information according to the data collection task and a plurality of adapter modules, and stores the core metadata, the metadata configuration and the data integration information to a system database.

[0007] By adopting the above technical solution, the code configuration module uses the display interface to display the dimension type and analysis rule of the size measurement data, which facilitates the to-be-used customer to dynamically configure the dimension type and analysis rule of the size measurement data based on the display interface, and facilitates the to-be-used customer to quickly complete the configuration of the multi-dimensional correlation analysis automobile size measurement system according to own needs, reduces the use threshold, and improves the ease of use of the system. In addition, the data driven module dynamically configures the core metadata for the size measurement data, the system can flexibly adapt to changing measurement requirements, and the response speed of the system to new data collection requirements is improved. The data collection module configures the modular adapter, the system can quickly access a plurality of measurement devices, reduces the development work of device access, and has stronger expansibility and device compatibility.

[0008] In some embodiments, an analysis processing module is further included, which is connected with the data driven module in network to receive the metadata configuration, and the analysis processing module is connected with the data collection module in network to receive the data integration information; the analysis processing module analyzes the metadata configuration and the data integration information based on a dynamic analysis engine to acquire result display information, and visually displays the result display information.

[0009] By adopting the above technical solution, the dynamic analysis engine can dynamically generate a corresponding analysis task according to the metadata configuration defined by the user, query and aggregate data from the system database, execute the analysis task according to the configured analysis logic to acquire result display information, and visually display the result display information to the to-be-used customer, thereby facilitating the to-be-used customer to make real-time decisions and strengthening the role of automobile size measurement related data.

[0010] In some embodiments, the core metadata includes size indicators, device parameters and process information, the analysis processing module is connected with the data driven module in network to acquire the size indicators, and the analysis processing module is connected with the data collection module in network to acquire size collection data, and acquires size deviation according to the size collection data and the size indicators. The analysis processing module learns the size deviation, the equipment parameter and the process information based on machine learning to obtain a correlation prediction model, obtains a potential correlation mode based on the correlation prediction model, and visually displays the correlation mode. The analysis processing module further obtains a size deviation mode based on the size history data, obtains a deviation reason corresponding to the size deviation mode based on a deviation analysis type, and visually displays the size deviation mode and the corresponding deviation reason.

[0011] By using the above technical solutions, the analysis processing module learns the size deviation, the equipment parameter and the process information based on machine learning to obtain a correlation prediction model, obtains a potential correlation mode based on the correlation prediction model, and visually displays the correlation mode, which can automatically detect the potential relationship between the size deviation of the size index and the equipment parameter and the process information, intelligently analyze the relationship among the size index, the equipment parameter and the process information, and improve the data analysis capability of the system.

[0012] In some embodiments, the data collection module includes a reading unit, an analysis unit and a task generation unit, the reading unit is configured to read the metadata configuration, the analysis unit is configured to extract key information based on the metadata configuration obtained by the reading unit, obtain an analysis result based on the key information, the task generation unit is configured to generate a data collection task based on the analysis result, and distribute the data collection task to a corresponding adapter module to connect a corresponding measurement device.

[0013] By using the above technical solutions, the analysis unit extracts key information based on the metadata configuration read by the reading unit, obtains an analysis result based on the key information, generates a data collection task based on the analysis result, and sends the data collection task to a corresponding adapter, which improves the data collection efficiency, and according to different collection tasks, different adapter modules are distributed to realize automatic collection of size data by corresponding measurement devices, and actively respond to rapidly changing measurement requirements.

[0014] In some embodiments, the data storage module provides a multi-type data storage mode for storing size measurement data collected by different measurement devices, and the algorithm expansion module is configured to store a custom analysis algorithm according to the measurement requirements of a to-be-used customer.

[0015] By adopting the technical scheme, the data storage module provides multiple types of data storage modes to store size measurement data collected by different measurement devices, thereby being capable of providing efficient and reliable data storage and management capabilities to support fast response and complex analysis of the system; in use, the to-be-used customer can upload a custom analysis algorithm to process specific measurement requirements, thereby improving the extensibility of the system and facilitating use of the to-be-used customer.

[0016] In some embodiments, the code configuration module configures the dimension type, analysis rule and correlation logic of the size measurement data in a drag-and-drop manner.

[0017] By adopting the technical scheme, the to-be-used customer can quickly configure collection rules, analysis dimensions and correlation logic in a drag-and-drop manner, which is simple to operate, and the to-be-used customer can preview the configuration result in real time during the configuration process to ensure the correctness of the configuration.

[0018] In a second aspect, the application provides a vehicle size measurement method based on multi-dimensional correlation analysis, which adopts the following technical scheme: A vehicle size measurement method based on multi-dimensional correlation analysis is executed based on the vehicle size measurement system based on multi-dimensional correlation analysis in the first aspect, and includes the following steps: The dimension type and analysis rule of the size measurement data are displayed based on the display interface; The core metadata of the size measurement data is configured, and the dimension type and analysis rule of the size measurement data are dynamically configured based on the display interface to obtain corresponding metadata configuration; Data collection tasks are generated based on the core metadata and the metadata configuration, and data integration information is obtained according to the data collection tasks and a plurality of adapter modules; The core metadata, the metadata configuration and the data integration information are stored in a system database.

[0019] In some embodiments, the analysis rule includes a correlation rule, the metadata configuration includes an analysis rule, a size index, a device parameter and process information, and after the data integration information is obtained according to the data collection tasks and a plurality of adapter modules, the following steps are further included: The analysis rule is parsed based on a dynamic analysis engine, a size deviation is obtained according to the size index and the data integration information, and it is determined whether the analysis rule includes a correlation rule; If the analysis rule includes a correlation rule, the size index, the device parameter and the process information are learned based on machine learning to obtain a correlation prediction model, and a potential correlation pattern is obtained based on the correlation prediction model; if the analysis rule does not exist the association analysis, based on the dynamic analysis engine, the metadata configuration is analyzed to obtain an analysis task; According to the analysis task, the data integration information corresponding to the system database is obtained, the analysis result is obtained based on the analysis logic, and the analysis result is visualized.

[0020] By adopting the above technical scheme, according to the needs of the to-be-used customer, if there is an association rule, it is determined that the to-be-used customer needs to obtain the correlation of the size index, the equipment parameter and the process information, therefore, the size index, the equipment parameter and the process information are learned based on machine learning, an association prediction model is obtained, and a potential association mode is obtained, which facilitates the to-be-used customer to adjust the current equipment parameter and process information in real time according to the size measurement data, and thus the qualified rate of the automobile parts can be improved; when there is no association analysis, the metadata configuration is analyzed based on the dynamic analysis engine to obtain an analysis task; according to the analysis task, the data integration information corresponding to the system database is obtained, the analysis result is obtained based on the analysis logic, and the analysis result is visualized, and then corresponding deviation analysis and trend analysis are automatically generated.

[0021] In some embodiments, before determining whether the analysis rule exists the association rule, the following steps are further included: According to the system database, historical measurement data is obtained, and the historical measurement data, the size index, the equipment parameter and the process information are obtained as to-be-associated data, and the to-be-associated data is judged for missing to obtain processing result data; According to the processing result data, a frequent item set is obtained, and an association rule is generated based on a preset threshold and a preset index; According to the association rule, an association rule flag is obtained, an analysis report is generated based on the association rule, and the analysis report is visualized.

[0022] By adopting the above technical scheme, the deviation analysis type and the deviation reason are analyzed through the support, the confidence and the lift and other corresponding preset indexes, and thus the association rule can be quickly obtained.

[0023] In some embodiments, after the core metadata, the metadata configuration and the data integration information are stored to the system database, the following steps are further included: According to the system database, the size deviation mode corresponding to the historical measurement data is obtained, and the size deviation mode is analyzed in time sequence based on time series analysis to obtain a deviation analysis type and a corresponding deviation reason; Based on the association analysis, the association rule flag of the deviation analysis type and the deviation reason is obtained; Real-time monitoring is performed on the size measurement data based on the cluster analysis and the association rule flag to obtain a monitoring report.

[0024] By adopting the technical solution, the deviation analysis type and the deviation reason of the size measurement data can be monitored in real time based on the time series analysis, and the deviation change alarm rule is set, so that the user can find the problems existing in the current size measurement data in time.

[0025] In summary, the present application has at least one of the following beneficial technical effects: The user can dynamically configure the dimension type and analysis rule of the size measurement data through the display interface, and the data driving module defines the core metadata of the size measurement data, such as data source, size index, and vehicle type classification; the user can quickly adjust the size data acquisition and analysis content to meet the dynamic demand, and the system can flexibly adapt to the changing measurement demand through the dynamic configuration of the metadata; The data acquisition module configures a modular adapter, the system can quickly access multiple measurement devices, reduces the development work of device access, and has stronger expansibility and device compatibility; the adapter module can dynamically generate an acquisition task according to the metadata configuration to realize automatic acquisition and integration of multi-source data; The analysis processing module supports automatic pattern discovery based on machine learning, can automatically detect the potential relationship between the size deviation and the process parameters and the device performance, and the user can also upload a custom analysis algorithm for processing specific measurement requirements; through intelligent correlation analysis, the system can automatically discover the potential relationship between the data; A code configuration module is provided, the user can configure the acquisition rule, analysis dimension, and correlation logic through a drag-and-drop method, which reduces the use threshold of the system, and the user can easily complete complex configuration, so that non-technical users can quickly get started. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 is a schematic diagram of an automobile size measurement system based on multi-dimensional correlation analysis provided by an embodiment of the present application; Figure 2 is a block diagram of an automobile size measurement method based on multi-dimensional correlation analysis provided by an embodiment of the present application; Figure 3 is another method block diagram provided by an embodiment of the present application; Figure 4 is a correlation analysis step block diagram provided by an embodiment of the present application.

[0027] The reference numerals are explained as follows: 10, code configuration module; 20, data driving module; 30, data acquisition module; 40, analysis processing module; 50, data storage module. DETAILED DESCRIPTION

[0028] In order to more clearly understand the purpose, technical solutions and advantages of the present application, the present application is described and explained below in conjunction with the drawings and embodiments. However, it should be understood by those of ordinary skill in the art that the present application can be implemented without these details. In some cases, in order to avoid unnecessary description and make aspects of the present application obscure, well-known methods, processes, systems, components and / or circuits that have been described at a high level will not be described in detail. It is obvious to those of ordinary skill in the art that various changes can be made to the embodiments disclosed in the present application, and the general principles defined in the present application can be applied to other embodiments and application scenarios without deviating from the principles and scope of the present application. Therefore, the present application is not limited to the shown embodiments, but conforms to the broadest scope claimed in the present application.

[0029] The embodiment of the present application discloses a car size measurement system based on multi-dimensional correlation analysis, including metadata driving layer, dynamic data acquisition layer, data storage layer, analysis processing layer, front-end interaction layer and intelligent recommendation layer, hierarchical mode executes corresponding functions, clear responsibilities, and high efficiency. A plurality of modules are set for different division layers to expand the performance of the system, which is convenient for the customer to be used to flexibly respond to changes. The hierarchical design separates data flow, business logic and algorithm decision, through clear boundaries and standardized interfaces, the system can maintain agility and support scale growth.

[0030] As shown in Figure 1 The car size measurement system based on multi-dimensional correlation analysis includes a code configuration module 10, a data driving module 20, and a data acquisition module 30. The code configuration module 10 displays the dimension type and analysis rules of the size measurement data based on the display interface. The data driving module 20 is used to configure the core metadata of the size measurement data, and dynamically configures the dimension type and analysis rules of the size measurement data based on the display interface to obtain the corresponding metadata configuration. The data acquisition module 30 is connected with the data driving module 20 to obtain the core metadata and metadata configuration, and generates a data acquisition task based on the core metadata and metadata configuration. The data acquisition module 30 obtains data integration information according to the data acquisition task and a plurality of adapter modules, and stores the core metadata, metadata configuration and data integration information to a system database.

[0031] The code configuration module 10 is arranged in the front-end interactive layer, and the code configuration module 10 includes a display interface. The display interface is mainly used for reconfiguring the multi-dimensional correlation analysis automobile size measurement system by a to-be-used customer. According to the demand of the to-be-used customer, a configuration unit in the code configuration module 10 can be started, and then the display interface is entered, so that the multi-dimensional correlation analysis automobile size measurement system is reconfigured.

[0032] In one embodiment, the code configuration module 10 is configured in a drag-and-drop manner for the dimension type of the size measurement data, the analysis rule and the correlation logic. Through a low-code configuration unit provided by the code configuration module 10, the system can reduce the technical threshold, so that a non-technical user can quickly start, and the problem that a developer is relied on for configuration in a traditional system is solved.

[0033] The code configuration module 10 is provided with a low-code configuration unit. The configuration unit includes components for realizing drag-and-drop configuration, real-time preview, saving and exporting and the like. The to-be-used customer can quickly complete the configuration of the collection rule, the analysis dimension and the correlation logic by dragging the components. The to-be-used customer can preview the configuration result in real time during the configuration process, so as to ensure the correctness of the configuration. The to-be-used customer can save the configuration result and export it in a standard format, such as JSON, XML and the like.

[0034] It should be noted that XML is a markup language used for marking electronic files to have structural characteristics, and can be used to mark data and define data types. It is a source language for users to define their own markup semantics. JSON is a lightweight data exchange format, which has good readability and the characteristics of being easy to quickly write. In the present embodiment, the use of JSON and XML is not changed, and therefore, the present embodiment will not be described in detail.

[0035] The front-end interactive layer further includes a size visualization module. The visualization module provides dynamic reports, 3D visualization and trend charts, and supports the to-be-used customer to filter and analyze according to vehicle models, parts or time periods. The size data visualization module supports dynamic reports, 3D visualization and trend charts, and the like. The dynamic report is generated according to the dimension type and the analysis rule of the to-be-used customer, and supports filtering and displaying the analysis result according to vehicle models, parts or time periods. The 3D visualization mainly displays the measurement data in combination with a 3D model, and intuitively presents the size deviation and trend change. The trend chart draws the trend of the data over time, and supports deviation trend and process parameter trend analysis.

[0036] It should be noted that how to start the configuration unit depends on the specific system type, including operating system, application software and Web services. Common entry configuration mode can be through the graphical user interface or through the command line interface. In order to ensure the configuration safety of the multi-dimensional correlation analysis automobile size measurement system, corresponding management authority needs to be set for different configuration modes, and the use safety of the multi-dimensional correlation analysis automobile size measurement system is improved.

[0037] Among them, the dimension type mainly refers to the data direction of size measurement, including vehicle type, parts and production stage. The analysis rule refers to the rule of different mode analysis of size measurement data, such as deviation range analysis, correlation factor analysis and trend analysis. However, the dimension type and the analysis rule are not limited to the above-mentioned content. According to the needs of the to-be-used customers, other contents can also be included, as long as they are related to the analysis of automobile size measurement data.

[0038] Among them, the data driving module 20 is arranged in the metadata driving layer. Through the dynamic configuration of metadata driving, the system can flexibly adapt to the changing measurement demand, and solves the problem of slow response of new data acquisition demand in traditional system. The data driving module 20 defines the core metadata related to automobile size measurement, including measurement data source, size index, vehicle type classification, production line parameter, process information, etc. The measurement data source includes laser measurement equipment and CMM equipment, and the size index includes gap, height and width. The core metadata is stored in the system database, which is convenient for real-time updating and querying of data.

[0039] It should be noted that when defining the core metadata, the to-be-used customers can configure the core metadata corresponding to the size measurement data according to their needs, and configure based on the display interface. The to-be-used customers configure the dimension type and analysis rule of the size measurement data through the display interface for dynamic configuration. The dimension type includes vehicle type, parts, production stage, etc., and the analysis rule includes deviation range, correlation factor, etc. The data driving management module can dynamically configure the dimension type and analysis rule of the size measurement data. The to-be-used customers can configure the dimension type according to the vehicle type, parts and production stage, and can define and modify the analysis rule of the deviation range and correlation factor. In addition, the data driving module 20 supports real-time updating of configuration information, which ensures that the system can quickly respond to the needs of users.

[0040] It should be noted that the specific setting value of the deviation range value can be set according to the needs of the to-be-used customer. In order to facilitate the use of the to-be-used customer, when the to-be-used customer sets the deviation range rule, the corresponding standard deviation range can be provided according to the current part type, so as to facilitate the selection of the to-be-used customer. The standard deviation range can be displayed to the to-be-used customer in the form of a list. The display form of the correlation factor definition and the modification analysis rule is the same as or similar to the deviation range rule, which will not be described in detail here. The display form of the dimension type can be displayed in the form of a list, so as to facilitate the selection of the to-be-used customer.

[0041] Among them, the data acquisition module 30 is arranged in the dynamic data acquisition layer, and a modular designed data source adapter is adopted to support connection of various measuring devices such as CMM, laser scanner, vision system and the like, and data platforms such as MES and PLM. Each adapter module is responsible for communication with a specific type of device or platform, supports various communication protocols such as TCP / IP, HTTP, serial communication and the like.

[0042] The adapter module can dynamically generate an acquisition task according to the metadata configuration, and realize automatic acquisition and integration of multi-source data. Through the modular adapter, the system can quickly access various measuring devices, reduce the development work of device access, and solve the problem of difficult device access in traditional systems. The data acquisition task is an acquisition task automatically generated by the data acquisition module 30 according to the metadata configuration, including data format conversion, galaxy rule, data acquisition and standardization processing and the like.

[0043] Based on the core metadata and the metadata configuration, the data acquisition task is generated by reading the metadata configuration of the to-be-used customer through the data acquisition module 30 to obtain the key information of the data to be collected, including collection time, data source, collection frequency, data format and cleaning rule. According to the key information, the collection time is determined to generate the collection task, and the corresponding collection device is obtained based on the collection task. The measurement data collected by the collection device is obtained, and the corresponding processing task is generated according to the data source, collection frequency, data format and cleaning rule. The measurement data is processed according to the processing task to obtain data integration information. The core metadata, metadata configuration and data integration information are stored in the system database.

[0044] In addition, the automobile size measurement system based on multi-dimensional correlation analysis also needs to provide clear commands and command line prompts for the data acquisition module 30 based on the cognitive load theory when the adapter module is designed. According to the different colors, different colors are set based on different collection frequencies, so that the data acquisition module 30 can process the colors accordingly, reduce the amount of information processed at the same time, gradually disclose, hide the advanced settings by default, expand on demand, and avoid interface information overload.

[0045] In one embodiment, the data collection module 30 includes a reading unit, an analysis unit and a task generation unit. The reading unit is configured to read the metadata configuration. The analysis unit is configured to extract key information based on the metadata configuration obtained by the reading unit, obtain analysis results based on the key information, and generate data collection tasks based on the analysis results. The task generation unit is configured to assign the data collection tasks to corresponding adapter modules to connect corresponding measurement devices.

[0046] It should be noted that the specific workflow of the data collection module 30 is as follows: first, the reading unit reads the metadata configuration from the system database, and then the analysis unit analyzes the metadata configuration to extract key information, such as data source, collection frequency, data format and cleaning rules. The task generation unit generates specific collection tasks based on the analysis results, and assigns the tasks to corresponding adapter modules. The adapter modules collect data from devices or platforms according to the task requirements, and perform format conversion and cleaning. The processed data is integrated and stored in the system database.

[0047] In the data collection progress monitoring, the collection progress bar is displayed in real time, such as the tqdm library and the estimated remaining time, to enhance the control feeling of the system. The non-technical language is used to prompt the failure reason, such as server connection timeout, please check the network and retry, to avoid user anxiety.

[0048] In one embodiment, the multi-dimensional correlation analysis-based automobile size measurement system further includes an analysis processing module 40, which is network-connected with the data driving module 20 to receive the metadata configuration, and network-connected with the data collection module 30 to receive the data integration information. The analysis processing module 40 analyzes the metadata configuration and the data integration information based on a dynamic analysis engine to obtain result display information, and visually displays the result display information.

[0049] The analysis processing module 40 is arranged in the analysis processing layer. Through intelligent correlation analysis of the analysis processing module 40, the system can automatically discover the potential relationship between data, solving the problem of lack of intelligent analysis capability in traditional systems. The dynamic analysis engine is mainly used for analyzing the metadata configuration and the data integration information to obtain result display information, which is visually displayed to facilitate the user to quickly obtain the existing analysis results, and support real-time decision-making.

[0050] It should be noted that the dynamic engine can load, parse and execute user-defined rules or models in real time, support hot updates, conditional triggering and adaptive adjustment. The dynamic engine is added to the automobile size measurement system based on multi-dimensional correlation analysis, thereby significantly improving the flexibility, real-time performance and intelligent level of the system. The dynamic engine generates specific analysis tasks according to the metadata configuration, and obtains result display information according to the analysis tasks, so as to facilitate the user to make timely decisions.

[0051] The workflow of the dynamic analysis engine is as follows. First, the user defines and configures the dimension type and analysis rule required for analyzing the current size measurement data on the display interface. Then, the dynamic analysis engine parses the configuration of the user, and generates specific analysis tasks. The dynamic engine queries and aggregates data from the system database to support analysis logic. Finally, the analysis tasks are executed according to the configured analysis logic, such as deviation analysis and trend analysis. The analysis results are visualized and displayed, and dynamic reports, charts and 3D visualization are supported. The analysis results are fed back to the user in real time to support real-time decision-making.

[0052] In one of the embodiments, the core metadata includes size indicators, device parameters and process information. The analysis processing module 40 is network connected with the data driving module 20 to obtain the size indicators, and network connected with the data acquisition module 30 to obtain the size acquisition data, and obtains the size deviation based on the size acquisition data and the size indicators. The analysis processing module 40 learns the size deviation, the device parameters and the process information based on machine learning to obtain a correlation prediction model, obtains a potential correlation pattern based on the correlation prediction model, and visualizes the correlation pattern. The analysis processing module 40 also obtains a size deviation pattern based on size historical data, obtains a deviation reason corresponding to the size deviation pattern based on a deviation analysis type, and visualizes the size deviation pattern and the corresponding deviation reason.

[0053] The analysis processing module 40 supports correlation analysis of the size acquisition data, production process information and device parameters, and provides automatic pattern discovery based on machine learning. The operation of how to obtain the correlation pattern is as follows. First, the size acquisition data is collected, specifically, historical data is obtained from the system database, and the obtained historical data is cleaned, standardized and feature extracted. Statistical analysis, time series analysis, clustering analysis and association rule mining methods are used to detect deviation patterns. Potential correlation patterns are automatically discovered based on machine learning algorithms such as decision tree, random forest and neural network. Finally, the correlation pattern is displayed to the user in a visual manner, and a specific analysis report is provided.

[0054] It should be noted that the algorithm used for machine learning is based on existing technology, which will not be described here. As for the use of statistical analysis, time series analysis, clustering analysis and association rule mining methods to detect deviation patterns, combined with dynamic rule engines and hierarchical architecture, to achieve efficient, flexible and interpretable anomaly detection, to realize the intelligent deviation detection system from "passive monitoring" to "active prediction".

[0055] Among them, the analysis processing module 40 automatically detects common size deviation patterns based on historical data, and recommends possible reasons such as equipment error, abnormal process parameters, etc. The specific workflow of obtaining the deviation reason corresponding to the size deviation pattern based on the deviation analysis type is to collect historical data from the system database, and to clean, standardize and feature extraction. Use statistical analysis, time series analysis, clustering analysis and association rule mining methods to detect deviation patterns. According to the detected deviation pattern, use rule engine, machine learning model and expert system to recommend possible reasons. Through visual display, early warning mechanism and suggestion measures, help users quickly locate problems and take measures. The analysis results are displayed to the user in a visual manner, supporting dynamic reports, charts and 3D visualization. The analysis results are fed back to the user in real time, supporting real-time decision-making.

[0056] Referring to Figure 1 In one of the embodiments, it further includes a data storage module 50 and an algorithm extension module, which is used to store custom analysis algorithms according to the measurement needs of the customers to be used. The data storage module 50 is network connected with the data acquisition module 30 to receive core metadata, metadata configuration and data integration information, and provides multiple types of data storage mode to store core metadata, metadata configuration and data integration information. The data storage module 50 interacts with the analysis processing module 40 to store various analysis data generated by the analysis processing module 40, such as the storage of result display information, association pattern data, size deviation pattern and deviation reason.

[0057] It should be noted that the data storage module 50 is provided in the data storage layer, which can provide efficient and reliable data storage and management capabilities to support the fast response and complex analysis of the system. The data storage layer includes several groups of database components. Specifically, the relational database is used to store structured data such as measurement data, configuration information, etc. It supports efficient query and transaction processing. The non-relational database is used to store unstructured data such as logs, text data, etc. It supports high-concurrency read and write and flexible data models. The data warehouse is used to store historical data and analysis data, which supports complex data analysis and reporting. The data warehouse supports data aggregation and multi-dimensional analysis. The cache system is used to improve the reading speed of the system and store frequently accessed data. It supports fast data retrieval and reduces database pressure. The file storage is used to store static files such as configuration files, log files, etc. It supports persistent storage and backup of files.

[0058] The algorithm expansion module mainly facilitates the to-be-used customer to upload a custom analysis algorithm for processing specific measurement requirements. The expandable algorithm library supports the following functions: The user can upload a custom analysis algorithm through the interface. The system supports management of uploaded algorithms, including version control, permission management, etc. The system supports calling the custom algorithm uploaded by the user in the analysis task.

[0059] It should be noted that the algorithm expansion module can be expanded through parameterization, functionality, and scale. Specifically, parameterization can change fixed values in the algorithm to adjustable parameters, automatically adjust parameters according to input data characteristics, and set a parameter optimization framework through inherited parameter search and optimization methods. For functional expansion, it is specifically processed by expanding from single target to multi-target, processing different types of data, and adding corresponding constraint conditions to facilitate work under specific constraints.

[0060] The algorithm expansion module includes an expansion button, an algorithm upload interface, an automatic recognition unit, and a function matching interface. The to-be-used customer enters the algorithm upload interface through the expansion button. The algorithm upload interface allows the to-be-used customer to upload the corresponding algorithm program. The function matching interface is stored in the algorithm expansion module. The automatic recognition unit identifies the function matching interface and the algorithm program, thereby identifying the to-be-used customer, and automatically uploading the algorithm of the to-be-used customer.

[0061] It should be noted that the algorithm program uploaded by the to-be-used customer is programmed according to the function matching interface, which facilitates the automatic recognition unit to identify.

[0062] The embodiment of the application further discloses a vehicle size measurement method based on multi-dimensional correlation analysis.

[0063] As shown in FIG. 2, the vehicle size measurement method based on multi-dimensional correlation analysis includes the following steps: S100, display the dimension type of the size measurement data and the analysis rule based on the display interface.

[0064] S200, configure the core metadata of the size measurement data, and dynamically configure the dimension type of the size measurement data and the analysis rule based on the display interface to obtain the corresponding metadata configuration.

[0065] S300, generate a data collection task based on the core metadata and the metadata configuration, and obtain data integration information according to the data collection task and a plurality of adapter modules.

[0066] S400, store the core metadata, the metadata configuration, and the data integration information to the system database.

[0067] The other functions described in steps S100 to S400 and the technical details of each function are the same as or similar to the corresponding features of the automobile size measurement system based on multi-dimensional association analysis described above, and therefore will not be described here.

[0068] Referring to Figure 3 In one of the embodiments, the analysis rule includes an association rule, the metadata configuration includes the analysis rule, the size index, the equipment parameter, and the process information, and after obtaining the data integration information according to the data collection task and the plurality of adapter modules, the following steps are further included: S310, based on the dynamic analysis engine, analyze the analysis rule, obtain the size deviation according to the size index and the data integration information, and determine whether the analysis rule exists an association rule.

[0069] S320, if the analysis rule exists an association rule, learn the size index, the equipment parameter, and the process information based on machine learning to obtain an association prediction model, and obtain a potential association pattern based on the association prediction model.

[0070] S330, if the data integration information corresponding to the analysis rule does not exist an association analysis, analyze the metadata configuration based on the dynamic analysis engine to obtain an analysis task.

[0071] S340, obtain the data integration information corresponding to the system database according to the analysis task, obtain an analysis result based on the analysis logic, and visually display the analysis result.

[0072] The analysis rules include association rules, bias analysis rules, change trend rules, bias trend rules, process parameter trend rules, etc. In step S310, it is judged whether the analysis rules of the demand of the to-be-used customer include association rules. If there are association rules, the size index, equipment parameter and process information are learned based on machine learning to obtain an association prediction model, and the potential association mode is obtained based on the association prediction model. However, other rule analysis of the demand of the to-be-used customer is analyzed according to corresponding steps S330-S340 to obtain corresponding analysis results, and the analysis results are displayed correspondingly.

[0073] It should be noted that, for the analysis of the bias mode, the specific workflow of obtaining the bias reason corresponding to the size bias mode based on the bias analysis type is as follows: collecting historical data from the system database, and performing cleaning, standardization and feature extraction. Statistical analysis, time series analysis, clustering analysis and association rule mining methods are used to detect bias patterns. According to the detected bias patterns, rule engines, machine learning models and expert systems are used to recommend possible reasons. Through visual display, early warning mechanism and suggestion measures, users can quickly locate problems and take measures. The analysis results are displayed to the user in a visual manner, supporting dynamic reports, charts and 3D visualization. The analysis results are fed back to the user in real time, supporting real-time decision-making.

[0074] Referring to Figure 4 In one embodiment, before learning the size index, equipment parameter and process information based on machine learning, the following steps are further included: S311, obtaining historical measurement data from the system database, and obtaining the historical measurement data, size index, equipment parameter and process information as to-be-associated data, and performing missing judgment on the to-be-associated data to obtain processing result data.

[0075] S312, obtaining a frequent item set according to the processing result data, and generating association rules based on a preset threshold and a preset index.

[0076] S313, obtaining an association rule flag according to the association rules, generating an analysis report based on the association rules, and visually displaying the analysis report.

[0077] The to-be-associated data represents data that needs to be associated and analyzed, and in this embodiment, refers to parameters corresponding to historical measurement data, size index, equipment parameter and process information. The to-be-associated data is preprocessed to obtain processing result data.

[0078] Specifically, in the data preprocessing, the collected data needs to be analyzed to determine whether there are missing values. If there are, loss aversion prompts need to be given. Missing values may lead to inaccurate detection, automatic type conversion, reduced cognitive load of the to-be-used customer, emotional prompts, and reduced anxiety of the to-be-used customer. For example, there are errors in historical measurement data, which leads to deviation in the obtained processing result data, and then a loss aversion prompt can be generated to facilitate the to-be-used customer to view.

[0079] The processing result data is converted into a Boolean matrix after missing value judgment of the associated data. The frequent item set refers to the data that repeatedly appears in the Boolean matrix. Various threshold values are preset. The preset threshold value is a parameter set in advance. In this embodiment, an automatic recommended threshold value can be used to reduce the decision-making pressure of the to-be-used customer.

[0080] The preset indicators include support, confidence, lift, and certainty. Through the calculation of the preset threshold value and the preset indicators, the association rules are formed. The association rule flag is formed by the combination of various preset indicators. The association rule flag includes high support + low lift, low support + high lift, high confidence + low support, and the like.

[0081] It should be noted that before obtaining the processing result data, the to-be-associated data needs to be subjected to statistical analysis detection, time series analysis, clustering analysis, association rule mining, and bias report generation. Specifically, the statistical analysis detection uses multidimensional descriptive statistics and the peak-end law for optimization, and visualizes the abnormal markers. The time series analysis sets a confidence interval, feedback loop, and interactive visualization. The clustering analysis automatically compresses the dimensions to reduce the user's decision-making pressure. The clustering analysis is used to display social recognition and mark small clusters as abnormal. The association rule mining uses discretization processing and automatic optimization binning. Specifically, the FP-Growth algorithm is used for progress feedback, rule visualization, and loss aversion emphasis. Finally, the bias report is generated using the peak-end law to design the structure.

[0082] It should be noted that this embodiment is based on the FP-Growth algorithm for progress feedback. The specific operation includes FP tree construction, frequent item set mining, rule generation, and finally result output of the to-be-associated data.

[0083] Exemplarily, the periodically acquired size acquisition data is detected, and based on cluster analysis, the size acquisition data is analyzed by corresponding deviation anomaly rules, and then a corresponding deviation report is output. The peak-end law is that at the peak moment, the detected size measurement data is at the peak moment, and at the end moment, the size measurement data is qualified data. Then the system can not only ensure the technical rigor, but also improve the user participation and decision efficiency. All code modules can be encapsulated as a pipeline to support dynamic parameter adjustment and visual interaction. The pipeline is encapsulated in the form of a pipeline, which facilitates dynamic parameter adjustment and visual interaction at each step.

[0084] In one of the embodiments, after storing the core metadata, metadata configuration and data integration information into the system database, the following steps are further included: S410, obtaining a size deviation pattern corresponding to the historical measurement data from the system database, and performing time series analysis on the size deviation pattern based on time series analysis to obtain a deviation analysis type and a corresponding deviation reason.

[0085] S420, obtaining a correlation rule mark of the deviation analysis type and the deviation reason based on correlation analysis.

[0086] S430, real-time monitoring of the size measurement data based on cluster analysis and the correlation rule mark to obtain a monitoring report.

[0087] The deviation analysis type refers to the deviation type of the size measurement data, and the deviation analysis type includes distribution deviation, mean value shift, variance anomaly and time series drift, etc. The deviation reason refers to the reason for the deviation of the size measurement data, such as low measurement equipment precision, unqualified automobile parts, etc. The size deviation pattern is analyzed by time series analysis, and then the deviation analysis type and the deviation reason are obtained. The size measurement data is monitored in real time by cluster analysis and the correlation rule mark, and a monitoring report is generated for the user to view in real time and adjust the equipment parameters and process information of the automobile parts in a timely manner.

[0088] The implementation principle is: The data driving module 20 configures the core metadata of the size measurement data according to the needs of the user, and dynamically configures the dimension type and analysis rule of the size measurement data based on the display interface to obtain the corresponding metadata configuration.

[0089] The data acquisition module 30 generates a data acquisition task based on the core metadata and the metadata configuration, obtains data integration information according to the data acquisition task and a plurality of adapter modules, and stores the core metadata, the metadata configuration and the data integration information into the system database.

[0090] The analysis processing module 40 learns the size deviation, the equipment parameter and the process information based on machine learning to obtain a correlation prediction model, obtains a potential correlation mode based on the correlation prediction model, and visually displays the correlation mode. The analysis processing module 40 obtains a size deviation mode based on size historical data, obtains a deviation reason corresponding to the size deviation mode based on a deviation analysis type, and visually displays the size deviation mode and the corresponding deviation reason.

[0091] It should be understood that although each step in the flowchart of the accompanying drawings is shown in sequence according to the direction of the arrow, these steps are not necessarily executed in sequence according to the direction of the arrow. Unless explicitly stated herein, the execution of these steps is not strictly limited in sequence, and they can be executed in other sequences.

[0092] The above are preferred embodiments of the present application, and are not intended to limit the protection scope of the present application, therefore: any equivalent changes made on the structure, shape, principle of the present application should be covered within the protection scope of the present application.

Claims

1. A multi-dimension correlation analysis based vehicle dimension measurement system, characterized in that, The application relates to a size measurement data analysis system, which comprises the following parts: a code configuration module (10) which is used for displaying the dimension type of size measurement data and analysis rules based on a display interface; a data driving module (20) which is used for configuring core metadata of the size measurement data and dynamically configuring the dimension type of the size measurement data and the analysis rules based on the display interface to obtain corresponding metadata configurations; a data acquisition module (30) which is connected with the data driving module (20) to obtain the core metadata and the metadata configurations, generates a data acquisition task based on the core metadata and the metadata configurations, obtains data integration information according to the data acquisition task and a plurality of adapter modules, and stores the core metadata, the metadata configurations and the data integration information into a system database.

2. The multi-dimension correlation analysis based vehicle dimension measurement system according to claim 1, wherein, The application further comprises an analysis processing module (40) which is connected with the data driving module (20) to receive the metadata configurations and is connected with the data acquisition module (30) to receive the data integration information; the analysis processing module (40) is used for analyzing the metadata configurations and the data integration information based on a dynamic analysis engine to obtain result display information and visually displaying the result display information.

3. The multi-dimension correlation analysis based vehicle dimension measurement system of claim 2, wherein, The core metadata comprises size indexes, device parameters and process information; the analysis processing module (40) is connected with the data driving module (20) to obtain the size indexes, is connected with the data acquisition module (30) to obtain size acquisition data, and obtains size deviation according to the size acquisition data and the size indexes; the analysis processing module (40) is used for learning the size deviation, the device parameters and the process information based on machine learning to obtain a correlation prediction model, obtaining potential correlation modes based on the correlation prediction model, and visually displaying the correlation modes; the analysis processing module (40) is further used for obtaining size deviation modes based on the size historical data, obtaining deviation reasons corresponding to the size deviation modes based on deviation analysis types, and visually displaying the size deviation modes and the corresponding deviation reasons.

4. The multi-dimension correlation analysis based vehicle dimension measurement system of claim 1, wherein, The data acquisition module (30) comprises a reading unit, an analysis unit and a task generation unit; the reading unit is used for reading the metadata configurations; the analysis unit extracts key information based on the metadata configurations obtained by the reading unit, obtains analysis results based on the key information, the task generation unit generates data acquisition tasks based on the analysis results, and distributes the data acquisition tasks to corresponding adapter modules to connect corresponding measuring devices.

5. The multi-dimension correlation analysis based vehicle dimension measurement system of claim 4, wherein, The application further comprises a data storage module (50) and an algorithm expansion module; the data storage module (50) provides a plurality of types of data storage modes and is used for storing size measurement data collected by different measuring devices. The algorithm expansion module is used to store a custom analysis algorithm according to a measurement requirement of a client to be used.

6. The multi-dimension correlation analysis based vehicle dimension measurement system of claim 1, wherein, The code configuration module (10) configures the dimension type, analysis rule and correlation logic of the size measurement data in a drag mode.

7. A method for measuring the size of a vehicle based on multi-dimensional correlation analysis, characterized by, The automobile size measurement system based on multi-dimensional correlation analysis according to any one of claims 1-6 comprises the following steps: The dimension type and analysis rule of the size measurement data are displayed based on a display interface; The core metadata of the size measurement data is configured, and the dimension type and analysis rule of the size measurement data are dynamically configured based on the display interface to obtain corresponding metadata configurations; Data acquisition tasks are generated based on the core metadata and the metadata configurations, and data integration information is obtained according to the data acquisition tasks and a plurality of adapter modules; The core metadata, the metadata configurations and the data integration information are stored in a system database.

8. The method of claim 7, wherein the method is based on a multi-dimensional correlation analysis. The analysis rule includes a correlation rule, the metadata configurations include an analysis rule, a size index, a device parameter and process information, and after the data integration information is obtained according to the data acquisition tasks and a plurality of adapter modules, the following steps are further included: The analysis rule is parsed based on a dynamic analysis engine, a size deviation is obtained according to the size index and the data integration information, and it is determined whether the analysis rule has a correlation rule; If the analysis rule has a correlation rule, the size index, the device parameter and the process information are learned based on machine learning to obtain a correlation prediction model, and a potential correlation mode is obtained based on the correlation prediction model; If the analysis rule corresponding to the data integration information has no correlation analysis, the metadata configurations are analyzed based on a dynamic analysis engine to obtain an analysis task; The data integration information corresponding to the system database is obtained according to the analysis task, an analysis result is obtained based on analysis logic, and the analysis result is visually displayed.

9. The method of claim 8, wherein the method is based on a multi-dimensional correlation analysis. Before it is determined whether the analysis rule has a correlation rule, the following steps are further included: Historical measurement data is obtained from the system database, and the historical measurement data, the size index, the device parameter and the process information are obtained as to-be-associated data, the to-be-associated data is subjected to missing judgment to obtain processing result data; Frequent item sets are obtained according to the processing result data, and a correlation rule is generated based on a preset threshold and a preset index; A correlation rule flag is obtained according to the correlation rule, an analysis report is generated based on the correlation rule, and the analysis report is visually displayed.

10. The method of claim 9, wherein the method is based on a multi-dimensional correlation analysis. After the core metadata, the metadata configurations and the data integration information are stored in the system database, the following steps are further included: A size deviation mode corresponding to the historical measurement data is obtained from the system database, and the size deviation mode is subjected to time series analysis based on time series analysis to obtain a deviation analysis type and a corresponding deviation reason; A correlation rule flag of the deviation analysis type and the deviation reason is obtained based on the correlation analysis; and The size deviation mode is subjected to time series analysis based on time series analysis to obtain a deviation analysis type and a corresponding deviation reason. Real-time monitoring of the size measurement data based on the cluster analysis and the association rule flag to obtain a monitoring report.

Images