An industrial process data processing and analysis method, terminal, medium and product

Abstract

This invention discloses a method, terminal, medium, and product for processing and analyzing industrial process data, relating to the field of data analysis and processing technology. The method includes: exporting multi-source industrial process data into a common text format file; performing standardization adaptation, cleaning, completion, and time-series alignment and merging on the text data to obtain standardized data in a unified and compatible format; using professional analysis tools to conduct in-depth offline analysis of the standardized data, including periodic statistical analysis, customized signal processing, and mathematical operations oriented towards the process scenario; generating customized analysis reports according to preset templates based on different user groups and completing automated publishing; and exporting the standardized data into the native proprietary format of the analysis tools, archiving and storing it together with the analysis reports. This invention achieves unified processing and in-depth analysis of multi-source heterogeneous industrial data, improves the automation level and reusability of data analysis results, and constructs a complete closed loop from data to knowledge.

Description

Technical Field

[0001] This invention relates to the field of data analysis and processing technology, and in particular to a method, terminal, medium and product for processing and analyzing industrial process data. Background Technology

[0002] In the field of industrial automation control, especially in metallurgical industries such as steel rolling, various monitoring software (such as ABB DriveWindow and Siemens WinCC) are widely used for real-time acquisition and visualization of equipment operating parameters. Furthermore, data may also be archived and managed through third-party tools such as databases and Excel, providing necessary support for production process monitoring and basic traceability.

[0003] However, existing industrial data acquisition and analysis solutions still have significant limitations. First, inconsistent data formats and closed interfaces are common problems. Different monitoring software uses their own proprietary formats to store data, leading to difficulties in cross-platform and cross-system data integration and creating data silos. Second, most existing software only provides basic browsing and simple statistical functions, lacking advanced processing capabilities such as frequency domain analysis, filtering and noise reduction, and feature extraction, making it difficult to meet advanced analytical needs such as equipment fault prediction, process parameter optimization, and product quality traceability. Third, the analysis process relies on manual operation; engineers need to manually complete data export, format conversion, analysis processing, and report compilation, which is inefficient and makes it difficult to ensure the consistency and repeatability of the analysis process. Fourth, data retrieval and traceability mechanisms are imperfect; historical data in different formats is difficult to access quickly, and a large amount of valuable industrial data remains untapped.

[0004] Therefore, there is an urgent need for an integrated solution that can be compatible with multi-source heterogeneous data, has in-depth offline analysis capabilities, and can achieve automated report generation, in order to break down data barriers and improve the utilization efficiency and analysis depth of industrial data. Summary of the Invention

[0005] To address the above issues, this invention provides a method, terminal, medium, and product for processing and analyzing industrial process data. The aim is to construct a standardized data preprocessing workflow to uniformly convert multi-source heterogeneous industrial data into a format compatible with professional analysis tools. Utilizing the powerful offline analysis capabilities of these tools, customized analysis and automated report generation for industrial production scenarios can be achieved, thereby realizing efficient and in-depth utilization of industrial process data and forming a complete processing loop from data to knowledge.

[0006] In a first aspect, the present invention provides a method for processing and analyzing industrial process data, comprising: S1, export multi-source industrial process data into a general text format file, the general text format file containing metadata definitions and process monitoring data; S2 performs standardization, cleaning, and integrity processing on text data, and performs time-series alignment and merging of multi-source text data based on time, converting multi-source text data into standardized data compatible with professional analysis tools. S3 uses professional analysis tools to perform in-depth offline analysis on the processed standardized data, including periodic statistical analysis, customized signal processing, and mathematical operations tailored to specific process scenarios. S4 generates customized reports based on preset templates from the in-depth analysis results and publishes them, according to different user groups; S5 exports the processed standardized data into a proprietary format file for professional analysis tools, and archives and stores it together with the generated report.

[0007] Furthermore, the general text format is TXT or CSV, and a preset symbol is used as the field separator; the metadata definition includes acquisition configuration information, signal variable name, unit, timestamp or length reference, defined in the file header, and clearly defines the one-to-one mapping relationship between each data column and industrial physical parameters.

[0008] Furthermore, the standardization process for the text data includes: removing preset symbols, unifying capitalization and unit suffixes, generating a standard header structure compatible with professional analysis tools, and adjusting the time column to the first column of the data; for raw data without a time column, generating a standard timestamp sequence according to a preset sampling rate to achieve native adaptation of the data to the analysis environment.

[0009] Furthermore, the text data cleaning and integrity processing includes: Based on the industrial process characteristics, a time-series change rate threshold is preset, and normal process mutations are identified by the data time-series change rate, avoiding misjudging normal process fluctuations as outliers. For true outliers and missing values, an interpolation method combining time series and data change trends is used to complete the data, ensuring the temporal integrity and continuity of the data.

[0010] Furthermore, the temporal alignment and merging of multi-source text data based on time includes: Using a unified absolute time as a benchmark, multi-source text data is time-series aligned, and synchronous interpolation is performed according to the sampling frequency of each data source to merge multi-source text data into unified text data with a single time-series dimension, thereby achieving the fusion of cross-system related data.

[0011] Furthermore, the periodic statistical analysis includes: Based on process feature identification and segmentation of independent production cycles, the maximum, minimum, average and standard deviation of core parameters are calculated for each independent production cycle, enabling refined comparison and quantitative evaluation of the operating status of different cycles.

[0012] Furthermore, the customized signal processing includes: The signals in the standardized data are subjected to noise suppression using a low-pass filter with a preset cutoff frequency to retain effective characteristic signals. The filtered signals are then converted into frequency domain spectra by a fast Fourier transform. The frequency domain spectra are analyzed in stages in conjunction with the equipment operating conditions to identify the frequency characteristics of potential equipment faults.

[0013] Furthermore, the mathematical operations oriented towards the process scenario specifically include: calculating the equipment's moving position by performing integral operations on the speed signal through predefined process formulas and virtual functions, thereby realizing coordinate transformation between the time and length dimensions; generating operating condition identification signals based on the threshold characteristics of process parameters, and completing the statistics of the number of times the equipment's operating capacity exceeds the limit; and calculating the mill stiffness based on the changes in rolling force and roll gap, providing quantitative data support for the analysis of equipment deterioration trends.

[0014] Furthermore, the preset template adopts a three-layer structure, including a basic information layer for displaying basic parameters such as equipment information, acquisition configuration, and production batch; an analysis result layer for presenting analysis results such as signal trend graphs, statistical tables, and frequency domain spectrum cores; and a fault suggestion layer for providing fault diagnosis conclusions, equipment maintenance suggestions, and maintenance cycle reminders. The customized report embeds links to original data files and analysis parameter configuration instructions, allowing users to directly call professional analysis tools from the report to open the original data for secondary analysis and result verification, achieving full-link traceability of the analysis process.

[0015] Furthermore, the archiving and storage specifically involves: binding standardized data and analysis reports with unique identifiers to construct a searchable and reusable industrial data asset library, providing unified and complete data support for subsequent equipment maintenance review, historical data tracing, degradation trend analysis, process parameter optimization, and analysis model iteration.

[0016] Secondly, the present invention also provides a computer terminal, comprising: Memory, which stores executable programs; A processor is used to run the program, wherein the program executes the method for processing and analyzing industrial process data during runtime.

[0017] Thirdly, the present invention also provides a computer-readable storage medium comprising a stored executable program, wherein the executable program, when running, controls the device where the computer-readable storage medium is located to execute the method for processing and analyzing industrial process data.

[0018] Fourthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the method for processing and analyzing industrial process data.

[0019] Compared with existing technologies, the advantages of this invention are as follows: By using a common text format as a transit medium, coupled with a standardized preprocessing mechanism, seamless integration and adaptation of cross-platform industrial data is achieved, breaking down format barriers between multi-source data. Furthermore, considering the process characteristics of industries such as steel rolling, the powerful analytical capabilities of professional analysis tools are leveraged to achieve in-depth mining and advanced processing of the fused data. Finally, through a templated report component, the analysis results are automatically synthesized into standardized reports that meet the needs of different departments, forming a complete output loop from data to knowledge. This method significantly improves the depth, efficiency, and automation level of industrial data analysis, demonstrating excellent technical practicality. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this drawing or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this drawing. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the overall process flow of the industrial process data processing and analysis method in an embodiment of the present invention; Figure 2 This is a detailed schematic diagram illustrating the steps of the data preprocessing and standardization process developed in this embodiment of the invention; Figure 3 This is a schematic diagram of the functional modules and execution flow of deep offline analysis in an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. All other embodiments obtained by those skilled in the art based on the embodiments provided by this invention without inventive effort are within the scope of protection of this invention.

[0023] This embodiment uses industrial process data analysis of a steel rolling system as a specific application scenario to describe in detail the method provided by this invention. Based on the data monitoring needs of industrial production sites, multiple platforms such as DriveWindow, WinCC, and ibaPDA are used for curve monitoring and data archiving. These platforms record the operating data of the transmission system and the process data of process control in industrial production, such as key parameters like rolling force, operating speed, torque, operating current, and equipment vibration amplitude, ensuring the comprehensiveness and real-time nature of data acquisition. Since the various data storage formats are incompatible, data integration, in-depth analysis, and report output are required.

[0024] The method for processing and analyzing industrial process data in this embodiment, such as Figure 1 The specific execution steps are as follows: S1, export the multi-source industrial process data into a general text format file, the general text format file containing metadata definitions and process monitoring data.

[0025] The collected real-time curves and archived data are uniformly exported as general text format files, specifically TXT or CSV files. The files use inherent preset symbols as field separators, and define metadata such as acquisition configuration information, signal variable names, units, timestamps or length references in several lines at the beginning of the file, thereby clarifying the mapping relationship between each data column and actual industrial parameters (such as product ID, speed, torque, rolling force, thickness, etc.).

[0026] S2 performs standardization, cleaning, and integrity processing on text data, and performs time-series alignment and merging of multi-source text data based on time, transforming multi-source text data into standardized data compatible with professional analysis tools.

[0027] The customized preprocessing tool developed in this embodiment, such as Figure 2 As shown, the solution to problems such as inconsistent formatting, missing data, and lack of standardized structure in raw text data is divided into three sub-steps: S21, Field Standardization and Adaptation: The original field names are processed to remove preset characters, unify the case of field names and the suffix of physical units, and eliminate naming differences between different systems; a standard header structure compatible with the professional analysis tool ibaAnalyzer is automatically generated, including module channel definitions, signal names, and unit rows, and the absolute time column is forced to be adjusted to the first column of each data column; for vibration data without an absolute time column, a standard timestamp sequence with equal intervals is automatically generated according to its sampling frequency, and aligned with the time reference of other data sources to complete the adaptation of data and analysis environment.

[0028] S22, Customized Data Cleaning and Integrity Processing: For the characteristics of the steel rolling process, a preset threshold for the rolling force temporal change rate is 50% / ms. When the rolling force change rate exceeds this threshold, it is automatically identified as a normal process mutation such as steel biting or steel throwing, and no outlier processing is performed to avoid misjudgment. For real outliers and data breakpoints caused by sensor interference and transmission packet loss, anomaly marking is performed first, and then linear interpolation combined with the trend of data changes before and after is used to complete the data, ensuring the temporal integrity and continuity of the data.

[0029] S23, Multi-source data time-series alignment and merging: Using a unified absolute time as a benchmark, time-series alignment is performed on various preprocessed text data, synchronous interpolation is performed on data with different sampling frequencies, and transmission data, process data, and vibration data are merged into unified text data of the same time-series dimension, realizing the fusion of cross-system related data and obtaining standardized data that can be directly recognized by ibaAnalyzer.

[0030] S3 performs in-depth offline analysis on the processed standardized data using professional analysis tools. For industrial process scenarios, it calls predefined analysis macro templates for the rolling mill system to perform customized in-depth analysis, such as... Figure 3 As shown, it specifically includes four types of core analysis operations: (1) Periodic statistical analysis: Based on the threshold characteristics of the core signal rolling force, the single-pass rolling cycle is automatically identified and segmented. For each rolling pass, the maximum, minimum, average, and standard deviation of rolling force, motor torque, and current are calculated respectively to accurately locate the process abnormality of a single pass and avoid interference from the full-section data statistics.

[0031] (2) Customized signal processing analysis: For the high-frequency acquired transmission current signal and rolling force signal, a Butterworth low-pass filter with a cutoff frequency of 50Hz is used to suppress noise and retain effective characteristic signals; the filtered time domain signal is subjected to fast Fourier transform and converted into a frequency domain spectrum. Combined with the load change of the rolling pass, a phased analysis is performed. By identifying the amplitude change of characteristic frequency, it is possible to accurately determine whether there is an early fault in the equipment.

[0032] (3) Process-oriented parameter calculation: By using predefined process formulas, the rolling speed signal is integrated to calculate the relative displacement between the roll and the steel plate, realizing the coordinate transformation from the time dimension to the steel plate length dimension; the steel-containing signal is automatically generated based on the rolling force threshold, and the number of times the motor torque exceeds the limit during the rolling process is counted; the longitudinal stiffness of the mill is calculated by the rolling force and roll gap change in the calibration section, providing quantitative data support for the analysis of the deterioration trend of the mill body.

[0033] (4) Macro template automated analysis: In response to the analysis needs of similar equipment, this embodiment solidifies all the filter parameters, fast Fourier transform parameters, statistical indicators, process calculation formulas and coordinate transformation rules in the above analysis process into a dedicated analysis macro template. When importing data from similar rolling mills, only one click is needed to call the template to automatically complete the whole process analysis without manual parameter adjustment, ensuring the consistency of analysis standards.

[0034] S4 generates customized reports based on preset templates from the in-depth analysis results and publishes them, depending on the different user groups.

[0035] Report Content and Layout: The report adopts a three-layer layout. The "Basic Information Layer" displays basic data such as product ID, transmission parameters, acquisition configuration, and product batch number. The "Analysis Results Layer" presents core analysis results such as signal trend charts, statistical tables, and frequency domain spectra. The "Fault Recommendation Layer" adds equipment maintenance guidance text and suggested maintenance cycles based on the analysis conclusions.

[0036] End-to-end data traceability: The report embeds links to the original standardized data files and parameter descriptions for the analysis macro templates. Users can directly call ibaAnalyzer from the report to open the original data for secondary analysis or result verification, achieving end-to-end traceability of the analysis process.

[0037] Report output and distribution: Export reports to various common formats such as PDF, HTML, and JPG to meet the needs of different scenarios; at the same time, automatically send reports to designated email addresses of relevant departments via command line for relevant personnel to view.

[0038] S5 exports the processed standardized data into a proprietary format file for professional analysis tools. This file is then linked to the generated analysis report using a unique steel plate ID and equipment number, and archived together in the industrial data asset repository. The archived data can be accessed at any time for subsequent equipment maintenance reviews, historical data comparisons, process parameter optimization, etc., enabling long-term value mining of production process data.

[0039] It should be noted that the present invention is not limited to the above-described embodiments. The above embodiments are merely examples, and any embodiments that have the same structure and perform the same effects as the technical concept within the scope of the present invention are included within the scope of the present invention. Furthermore, various modifications that can be conceived by those skilled in the art to the embodiments, and other ways of constructing by combining some of the constituent elements of the embodiments, without departing from the spirit of the present invention, are also included within the scope of the present invention.

Claims

1. A method for processing and analyzing industrial process data, characterized in that, include: S1, export multi-source industrial process data into a general text format file, the general text format file containing metadata definitions and process monitoring data; S2 performs standardization, cleaning, and integrity processing on text data, and performs time-series alignment and merging of multi-source text data based on time, converting multi-source text data into standardized data compatible with professional analysis tools. S3 uses professional analysis tools to perform in-depth offline analysis on the processed standardized data, including periodic statistical analysis, customized signal processing, and mathematical operations tailored to specific process scenarios. S4 generates customized reports based on preset templates from the in-depth analysis results and publishes them, according to different user groups; S5 exports the processed standardized data into a proprietary format file for professional analysis tools, and archives and stores it together with the generated report.

2. The method for processing and analyzing industrial process data as described in claim 1, characterized in that, The general text format is TXT or CSV, and a preset symbol is used as the field separator; the metadata definition includes acquisition configuration information, signal variable name, unit, timestamp or length reference, and is defined in the file header.

3. The method for processing and analyzing industrial process data as described in claim 1, characterized in that, The standardization process for text data includes: removing preset symbols, unifying capitalization and unit suffixes, generating a standard header structure compatible with professional analysis tools, and adjusting the time column to the first column of the data; for raw data without a time column, generating a standard timestamp sequence based on a preset sampling rate.

4. The method for processing and analyzing industrial process data as described in claim 1, characterized in that, The text data cleaning and integrity processing includes: Based on the industrial process characteristics, a time-series change rate threshold is preset, and normal process mutations are identified by the data time-series change rate. For true outliers and missing values, an interpolation method combining time series and data change trends is used to complete the data.

5. The method for processing and analyzing industrial process data as described in claim 1, characterized in that, The time-based alignment and merging of multi-source text data includes: Using a unified absolute time as a benchmark, multi-source text data is time-series aligned, and synchronous interpolation is performed according to the sampling frequency of each data source to merge multi-source text data into unified text data with a single time-series dimension.

6. The method for processing and analyzing industrial process data as described in claim 1, characterized in that, The periodic statistical analysis includes: Based on process feature identification and segmentation of independent production cycles, the maximum, minimum, average and standard deviation of core parameters are calculated for each independent production cycle, enabling refined comparison and quantitative evaluation of the operating status of different cycles.

7. The method for processing and analyzing industrial process data as described in claim 1, characterized in that, The customized signal processing includes: The signals in the standardized data are subjected to noise suppression using a low-pass filter with a preset cutoff frequency. The filtered signals are then converted into frequency domain spectra using a fast Fourier transform. The frequency domain spectra are analyzed in stages in conjunction with the equipment operating conditions to identify the frequency characteristics of potential equipment faults.

8. A computer terminal, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the method according to any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the computer-readable storage medium is located to perform the method of any one of claims 1 to 7.

10. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 7.

Images