Feature extraction method and device for complex process technology

Abstract

The embodiments of this specification provide a method and device for extracting features of a complex process technology, wherein the method includes: preprocessing and evaluating process data generated by the process industry based on a time series to obtain stand-by data; based on the stand-by data, screening out key parameters that have a significant impact on the operating risks of complex process technologies; using a data-driven algorithm to mine historical data to autonomously identify the coupling relationship between the key parameters of key complex process technologies; based on the key parameters and the coupling relationship, using a feature extraction method to extract feature indicators that can reflect changes in system status, and determining the inherent safety risk characteristics of the complex process technology based on the feature indicators.

Description

Technical Field

[0001] This document relates to the field of computer technology, and in particular to a method and device for extracting features from a complex process. Background Art

[0002] In the existing technologies, the research on dynamic risk intelligent control technology and the establishment of tool sets for process industries are in their infancy. my country's process industry is at a critical stage of intelligent upgrading, facing challenges such as improving the scientific nature of production and operation management, enhancing the accuracy of risk analysis to cope with complex process mechanisms, strong variable coupling and process dynamics. The existing risk control systems at home and abroad often show problems such as weak stability of risk control systems and low levels of collaborative innovation in dealing with the complexity and dynamics of process industries, especially in terms of multi-security collaborative protection that integrates functional safety, information security and physical security, real-time risk monitoring and early warning, and intelligent decision support. There are obvious deficiencies. It is urgent to establish a full-factor risk perception and intelligent analysis system, a multi-security collaborative protection technology architecture and an intelligent control technology platform to support the collaborative control of dynamic risks in the process industry.

[0003] In terms of risk perception, analysis, assessment and decision-making of all factors in production and operation of process industries, current research focuses on multimodal data alignment and fusion technology, complex data feature extraction, and dynamic decision support for risk assessment models. Domestic and foreign scholars have conducted in-depth research on multimodal data fusion technology and proposed a variety of fusion algorithms and models, such as multimodal fusion methods based on deep learning and multi-source data fusion technology based on graph neural networks. In view of the complex characteristics of industrial data, Gaussian mixture models (GMM), support vector regression (SVR), cointegration analysis (CA), etc. have also been applied to feature extraction of industrial data.

[0004] With the continuous development of technologies such as big data, artificial intelligence and the Internet of Things, research on risk perception, analysis, assessment and decision-making of all factors in production and operation of process industries will show a trend that data-driven decision support will become more popular, the level of intelligence and automation will continue to improve, and cross-domain integration will become the norm.

[0005] In summary, the dynamic risk intelligent control technology in the process industry is developing rapidly with the help of cutting-edge technologies such as industrial Internet, big data, and artificial intelligence. Significant progress has been made in data fusion, risk assessment models, risk control technology, real-time monitoring and early warning at home and abroad, but domestic research and application are relatively lagging behind, and the platform's versatility and intelligence level need to be improved. Therefore, there is an urgent need for a feature extraction method for complex process technology. Summary of the invention

[0006] The purpose of the present invention is to provide a method and device for extracting features of a complex process, aiming to solve the above-mentioned problems in the prior art.

[0007] The present invention provides a method for extracting features of a complex process, comprising:

[0008] Preprocess and evaluate the process data generated by the process industry based on time series to obtain the data to be used;

[0009] Based on the stand-by data, key parameters that have a significant impact on the operational risks of complex process technology are screened out;

[0010] Use data-driven algorithms to mine historical data and autonomously identify the coupling relationship between key parameters of key complex process steps;

[0011] Based on the key parameters and the coupling relationship, a feature extraction method is used to extract characteristic indicators that can reflect changes in system status, and based on the characteristic indicators, the intrinsic safety risk characteristics of a complex process are determined.

[0012] The present invention provides a feature extraction device for a complex process, comprising:

[0013] The preprocessing module is used to preprocess and evaluate the process data generated by the process industry based on time series to obtain the data to be used;

[0014] A screening module, for screening out key parameters that have a significant impact on the operation risk of complex process technology based on the stand-by data;

[0015] An identification module, which is used to mine historical data using a data-driven algorithm and autonomously identify the coupling relationship between the key parameters of key complex process steps;

[0016] The extraction module is used to extract characteristic indicators that can reflect changes in system status based on the key parameters and the coupling relationship, and determine the intrinsic safety risk characteristics of the complex process technology based on the characteristic indicators.

[0017] An embodiment of the present invention also provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the steps of the feature extraction method for the above-mentioned complex process when executed by the processor.

[0018] An embodiment of the present invention further provides a computer-readable storage medium, on which a program for implementing information transmission is stored. When the program is executed by a processor, the steps of the feature extraction method for the above-mentioned complex process technology are implemented.

[0019] By adopting the embodiment of the present invention, through the feature extraction technology based on the fusion of data and knowledge in complex process technology, it is possible to characterize the complex features of process data such as strong coupling, nonlinearity, non-stationarity, multi-modality, etc. The intrinsic safety risk features are extracted, and by integrating advanced data processing and analysis technologies, comprehensive monitoring and evaluation of potential risks in production operations are achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate one or more embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0021] Figure 1 is a flow chart of a method for extracting features from a complex process technology according to an embodiment of the present invention;

[0022] Figure 2 is a detailed schematic diagram of a feature extraction method for a complex process technology process according to an embodiment of the present invention;

[0023] Figure 3 is a schematic diagram of a feature extraction device for a complex process technology according to an embodiment of the present invention;

[0024] Figure 4 is a schematic diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0025] In order to enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the following will be combined with the drawings in one or more embodiments of this specification to clearly and completely describe the technical solutions in one or more embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this document.

[0026] Method Embodiment

[0027] According to an embodiment of the present invention, a method for extracting features of a complex process is provided. Figure 1 is a flow chart of a method for extracting features from a complex process according to an embodiment of the present invention. Figure 1 As shown, the feature extraction method of a complex process according to an embodiment of the present invention specifically includes:

[0028] Step S101, preprocessing and data evaluation are performed on process data generated by the process industry based on time series to obtain stand-by data;

[0029] Among them, the process data generated by the process industry include functional safety related data, information security related data, and physical safety related data. Among them, the functional safety related data specifically include: HSE management system, safety risk analysis, equipment and safety redundancy design, safe operation safety control, fault response safety control, production personnel safety management, safety target responsibilities, safety system management, safety skills training, safe site management, safety application management, accident management, and continuous safety improvement. The information security related data specifically include: network structure planning and design, network security hardware configuration, OS application and data security, professional network security personnel. The physical safety related content specifically includes: storage area tank area, production area, and material transportation.

[0030] Among them, preprocessing specifically includes: data cleaning, standardization, data correction, redundancy removal, small sample problem, data balancing, data labeling, data multi-source alignment, data multi-modal fusion and other processing.

[0031] Step S102, based on the standby data, screen out key parameters that have a significant impact on the operation risk of complex process technology; specifically includes:

[0032] Based on the stand-by data, through multidimensional variable analysis techniques such as data correlation analysis, principal component analysis, and cluster analysis, combined with process risk theory and expert experience information, key parameters that have a significant impact on the operating risks of complex process technologies are screened out.

[0033] Step S103, using a data-driven algorithm to mine historical data and autonomously identify the coupling relationship between the key parameters of the key complex process; specifically includes:

[0034] Data-driven algorithms such as neural networks, support vector machines, decision trees and random forests are used to mine historical data and autonomously identify the coupling relationships between the key parameters of key complex process steps.

[0035] Step S104, based on the key parameters and the coupling relationship, a feature extraction method is used to extract feature indicators that can reflect changes in system status, and the intrinsic safety risk characteristics of the complex process are determined based on the feature indicators. Specifically, it includes:

[0036] Based on the key parameters and the coupling relationship, a feature extraction method is used to extract characteristic indicators that can reflect changes in the system state, and based on the characteristic indicators, with feature sparsity as the goal, the essential characteristics of the safety risk are extracted to obtain the intrinsic safety risk characteristics of the complex process technology, wherein the intrinsic safety risk characteristics have the characteristics of strong coupling, nonlinearity, non-stationarity and multimodality, and the feature extraction method specifically includes: a method based on statistical modeling, a method based on machine learning, a method based on deep learning, a method based on cointegration theory and a method based on generalized information theory.

[0037] The embodiments of the present invention form a method logical system for extracting features from multi-source data, which fully integrates statistical modeling, machine learning, and cointegration theory to transform and compress multi-source heterogeneous, multi-lag, few-label, and high-interference data, realize the extraction of essential intrinsic features, and provide theoretical support for early warning and decision-making.

[0038] like Figure 2 As shown in the figure, in order to solve the problem that all-factor risks in complex process technology are difficult to accurately perceive and effectively manage, a study on all-factor risk perception based on multi-source heterogeneous data is carried out. This study integrates heterogeneous data resources from personnel, processes, equipment, environment, network and other aspects, and realizes the deep integration and efficient use of data through advanced data processing technologies and methods. Innovatively apply advanced data analysis technologies such as multi-source data fusion and process data flow modeling, as well as adversarial analysis methods such as human-machine intelligent identification and machine learning algorithms, deeply explore the risk characteristics and laws behind the data, and study and construct a risk perception model with high precision and high robustness, so as to realize the real-time and accurate perception and early warning of all-factor risks in the production and operation of the process industry.

[0039] There is a strong coupling and nonlinear relationship between the production parameters of complex industrial processes, and potential risks are difficult to accurately predict. Through multidimensional variable analysis techniques such as data correlation analysis, principal component analysis, and cluster analysis, combined with process risk theory and expert experience, key parameters that have a significant impact on process operation risks are screened out to reduce data dimensions and highlight important variables. Data-driven algorithms such as neural networks, support vector machines, decision trees, and random forests are used to mine historical data, autonomously identify the coupling relationship between key process parameters, and extract characteristic indicators that can reflect changes in system states. The method of extracting the essential features of safety risks is studied with feature sparsity as the goal, and the characteristics of essential safety risks are mined.

[0040] In summary, the embodiment of the present invention proposes a feature extraction technology based on the fusion of data and knowledge in a complex process process, which characterizes the complex features of process data such as strong coupling, nonlinearity, non-stationarity, and multimodality; there are strong coupling and nonlinear relationships between industrial production parameters of complex processes, and potential risks are difficult to accurately predict. Through multidimensional variable analysis technologies such as data correlation analysis, principal component analysis, and cluster analysis, combined with process risk theory and expert experience, the key parameters that have a significant impact on process operation risks are screened out, so as to achieve the purpose of reducing data dimensions and highlighting important variables. Data-driven algorithms such as neural networks, support vector machines, decision trees, and random forests are used to mine historical data, autonomously identify the coupling relationship between key process parameters, extract characteristic indicators that can reflect changes in system states, and study the method of extracting the essential features of safety risks with feature sparsity as the goal, and mine the characteristics of intrinsic safety risks. The technical solution of the embodiment of the present invention integrates efficient feature extraction, accurate risk perception, and intelligent analysis technology, and combines advanced algorithms of signal processing, machine learning, and artificial intelligence to deeply mine and refine the key feature set in all-factor data, effectively dealing with complex characteristics such as nonlinearity, dynamics, multimodality, and non-stationarity of data.

[0041] Device Example 1

[0042] According to an embodiment of the present invention, a device for extracting features from a complex process is provided. Figure 3 Schematic diagram of a feature extraction device for a complex process according to an embodiment of the present invention. Figure 3 As shown, the feature extraction device for a complex process according to an embodiment of the present invention specifically includes:

[0043] The preprocessing module 30 is used to preprocess and evaluate the data based on the time series to obtain the stand-by data; wherein, the process data generated by the process industry includes functional safety related data, information security related data, and physical safety related data, wherein the functional safety related data specifically includes: HSE management system, safety risk analysis, equipment and safety redundancy design, safe operation safety control, fault response safety control, production personnel safety management, safety target responsibilities, safety system management, safety skills training, safety site management, safety application management, accident management, and continuous safety improvement; the information security related data specifically includes: network structure planning and design, network security hardware configuration, OS application and data security, professional network security personnel; the physical safety related content specifically includes: storage tank area, production area, and material transportation.

[0044] Among them, preprocessing specifically includes: data cleaning, standardization, data correction, redundancy removal, small sample problem, data balancing, data labeling, data multi-source alignment, data multi-modal fusion and other processing.

[0045] The screening module 32 is used to screen out key parameters that have a significant impact on the operation risk of complex process technology based on the standby data; specifically, it is used to:

[0046] Based on the stand-by data, through multidimensional variable analysis techniques such as data correlation analysis, principal component analysis, and cluster analysis, combined with process risk theory and expert experience information, key parameters that have a significant impact on the operating risks of complex process technologies are screened out.

[0047] The identification module 34 is used to mine historical data using a data-driven algorithm and autonomously identify the coupling relationship between the key parameters of the key complex process; specifically, it is used to:

[0048] Data-driven algorithms such as neural networks, support vector machines, decision trees and random forests are used to mine historical data and autonomously identify the coupling relationships between the key parameters of key complex process steps.

[0049] The extraction module 36 is used to extract characteristic indicators that can reflect the change of system status based on the key parameters and the coupling relationship by using a characteristic extraction method, and determine the intrinsic safety risk characteristics of the complex process based on the characteristic indicators. Specifically used for:

[0050] Based on the key parameters and the coupling relationship, a feature extraction method is used to extract characteristic indicators that can reflect changes in the system state, and based on the characteristic indicators, with feature sparsity as the goal, the essential characteristics of the safety risk are extracted to obtain the intrinsic safety risk characteristics of the complex process technology, wherein the intrinsic safety risk characteristics have the characteristics of strong coupling, nonlinearity, non-stationarity and multimodality, and the feature extraction method specifically includes: a method based on statistical modeling, a method based on machine learning, a method based on deep learning, a method based on cointegration theory and a method based on generalized information theory.

[0051] The embodiment of the present invention is a device embodiment corresponding to the above method embodiment. The specific operations of each module can be understood by referring to the description of the method embodiment, which will not be repeated here.

[0052] Device Example 2

[0053] An embodiment of the present invention provides an electronic device, such as Figure 4 As shown, it includes: a memory 40, a processor 42, and a computer program stored in the memory 40 and executable on the processor 42. When the computer program is executed by the processor 42, the steps described in the method embodiment are implemented.

[0054] Device Example 3

[0055] An embodiment of the present invention provides a computer-readable storage medium, on which a program for implementing information transmission is stored. When the program is executed by the processor 42, the steps described in the method embodiment are implemented.

[0056] The computer-readable storage medium in this embodiment includes, but is not limited to, ROM, RAM, magnetic disk or optical disk, etc.

[0057] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A feature extraction method for a complex process, characterized in that: include: Preprocess and evaluate the process data generated by the process industry based on time series to obtain the data to be used; Based on the stand-by data, key parameters that have a significant impact on the operational risks of complex process technology are screened out; Use data-driven algorithms to mine historical data and autonomously identify the coupling relationship between key parameters of key complex process steps; Based on the key parameters and the coupling relationship, a feature extraction method is used to extract characteristic indicators that can reflect changes in system status, and based on the characteristic indicators, the intrinsic safety risk characteristics of a complex process are determined.

2. The method according to claim 1, characterized in that: Based on the standby data, the key parameters that have a significant impact on the operation risk of complex process technology are screened out, including: Based on the stand-by data, through multidimensional variable analysis techniques such as data correlation analysis, principal component analysis, and cluster analysis, combined with process risk theory and expert experience information, key parameters that have a significant impact on the operating risks of complex process technologies are screened out.

3. The method according to claim 1, characterized in that The data-driven algorithm is used to mine historical data and autonomously identify the coupling relationship between the key parameters of key complex process processes, including: Data-driven algorithms such as neural networks, support vector machines, decision trees and random forests are used to mine historical data and autonomously identify the coupling relationships between the key parameters of key complex process steps.

4. The method according to claim 1, characterized in that: Based on the key parameters and the coupling relationship, characteristic indicators that can reflect changes in system status are extracted, and based on the characteristic indicators, the intrinsic safety risk characteristics of the complex process are determined, specifically including: Based on the key parameters and the coupling relationship, a feature extraction method is used to extract characteristic indicators that can reflect changes in the system state, and based on the characteristic indicators, with feature sparsity as the goal, the essential characteristics of the safety risk are extracted to obtain the intrinsic safety risk characteristics of the complex process technology, wherein the intrinsic safety risk characteristics have the characteristics of strong coupling, nonlinearity, non-stationarity and multimodality, and the feature extraction method specifically includes: a method based on statistical modeling, a method based on machine learning, a method based on deep learning, a method based on cointegration theory and a method based on generalized information theory.

5. A feature extraction device for a complex process, characterized in that: include: The preprocessing module is used to preprocess and evaluate the process data generated by the process industry based on time series to obtain the data to be used; A screening module, for screening out key parameters that have a significant impact on the operation risk of complex process technology based on the stand-by data; An identification module, which is used to mine historical data using a data-driven algorithm and autonomously identify the coupling relationship between the key parameters of key complex process steps; The extraction module is used to extract characteristic indicators that can reflect changes in system status based on the key parameters and the coupling relationship, and determine the intrinsic safety risk characteristics of the complex process technology based on the characteristic indicators.

6. The device according to claim 5, characterized in that The screening module is specifically used for: Based on the stand-by data, through multidimensional variable analysis techniques such as data correlation analysis, principal component analysis, and cluster analysis, combined with process risk theory and expert experience information, key parameters that have a significant impact on the operating risks of complex process technologies are screened out.

7. The device according to claim 5, characterized in that The identification module is specifically used for: Data-driven algorithms such as neural networks, support vector machines, decision trees and random forests are used to mine historical data and autonomously identify the coupling relationships between the key parameters of key complex process steps.

8. The device according to claim 5, characterized in that The extraction module is specifically used for: Based on the key parameters and the coupling relationship, a feature extraction method is used to extract characteristic indicators that can reflect changes in the system state, and based on the characteristic indicators, with feature sparsity as the goal, the essential characteristics of the safety risk are extracted to obtain the intrinsic safety risk characteristics of the complex process technology, wherein the intrinsic safety risk characteristics have the characteristics of strong coupling, nonlinearity, non-stationarity and multimodality, and the feature extraction method specifically includes: a method based on statistical modeling, a method based on machine learning, a method based on deep learning, a method based on cointegration theory and a method based on generalized information theory.

9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the steps of the feature extraction method for a complex process as described in any one of claims 1 to 4 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores an implementation program for information transmission, and when the program is executed by a processor, the steps of the feature extraction method for a complex process as described in any one of claims 1 to 4 are implemented.

Images