Submerged arc furnace data modeling method

Abstract

The invention provides a submerged arc furnace data modeling method. The submerged arc furnace data modeling method comprises the following steps: preparing submerged arc furnace data modeling; modeling data of the submerged arc furnace are input; modeling and preprocessing data of the submerged arc furnace; training and outputting a submerged arc furnace data model; performing model training evaluation on the processed data to obtain a trained model; and performing prediction through the trained model to realize data pushing and storage. Database processing measures are adopted, and application data of local deployment and cloud deployment are considered. Standard cleaning is carried out on submerged arc furnace data, and the accuracy of the submerged arc furnace data is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent manufacturing, and in particular to a data modeling method for an electric arc furnace. Background Art

[0002] The smelting process of an electric arc furnace is a complex high-temperature physical and chemical reaction process involving dozens of reaction variables. It is a typical nonlinear time-varying system with complex structure and unknown mechanism. The production control of electric arc furnace is difficult to solve with conventional detection and control means. It mainly relies on manual experience judgment and subjective analysis. The data is highly discrete and the algorithm model is lacking.

[0003] Currently, several existing mechanistic models and sensor-based expert systems for submerged arc furnace control exist within the industry, but these have proven ineffective in field applications. To date, closed-loop intelligent control of submerged arc furnace smelting remains unattainable. Summary of the Invention

[0004] Given the aforementioned complex, nonlinear, time-varying system nature of the submerged arc furnace smelting process, which presents technical challenges difficult to address using conventional detection and control methods, a method for processing and modeling submerged arc furnace data is provided. This invention leverages existing submerged arc furnace control systems and integrates submerged arc furnace production processes to provide an efficient method for processing submerged arc furnace data, implement practical methods for modeling and analyzing submerged arc furnace data, and improve submerged arc furnace smelting control capabilities.

[0005] The technical means adopted in the present invention are as follows:

[0006] A method for data modeling of an electric arc furnace comprises the following steps:

[0007] S201: Preparation of data modeling for an electric arc furnace; the preparation of data modeling for an electric arc furnace includes: preparation of data source, preparation of parameter configuration, preparation of network detection, and preparation of basic general data function blocks;

[0008] S202: Inputting data for submerged arc furnace modeling; the submerged arc furnace modeling data input includes: online input and offline input;

[0009] S203: Submerged arc furnace data modeling preprocessing; the submerged arc furnace data modeling preprocessing includes: data format preprocessing, data filtering preprocessing and auxiliary data preprocessing;

[0010] S204: training and outputting a submerged arc furnace data model; conducting model training and evaluation on the processed data to obtain a trained model; and performing predictions using the trained model to achieve data push and storage.

[0011] Furthermore, the data source preparation includes: preparation of online data sources and preparation of offline data sources; the parameter configuration preparation includes: online data, files, model configuration parameters; the network detection preparation includes: IP address management, polling ping detection and abnormal behavior alarm module.

[0012] Furthermore, the online input is firstly detected for network connection and then obtained through the raw data online initialization method of the raw data function block; the offline input is obtained through the raw data local initialization method of the raw data function block.

[0013] Furthermore, the data format preprocessing includes: database and file time processing and time type conversion; the data format preprocessing is implemented by a time function block;

[0014] The data filtering preprocessing realizes data filtering by filtering data function block class;

[0015] The auxiliary data preprocessing includes: a spare data function block, an electrode insertion depth function block, a working end length data function block, an average value data function block and a user-defined function block.

[0016] Furthermore, in said S204, the data result set resulting from data cleaning, data dimensionality reduction and data feature extraction is trained through a training data set module, and the target value of the sample data is calibrated by combining manual judgment and function constraints.

[0017] Furthermore, the training data set module includes: a six-dimensional training data set of electrode 1 electric quantity, a six-dimensional training data set of electrode 2 electric quantity, a six-dimensional training data set of electrode 3 electric quantity, and a seven-dimensional training data set of electric furnace non-electric quantity;

[0018] The characteristic vector of the electrode-electricity six-dimensional training data set is expressed as:

[0019]

[0020] The characteristic vector of the six-dimensional training data set of electrode two charges is expressed as:

[0021]

[0022] The characteristic vector of the six-dimensional training data set of the three-electrode electrical quantities is expressed as:

[0023]

[0024] The feature vector of the electrode non-electrical seven-dimensional training data set is expressed as:

[0025]

[0026] For a new sample X test , for each sample in the training set Calculate the Manhattan distance as:

[0027] X test The Manhattan distance of electrode 1 in six dimensions:

[0028] X test The Manhattan distance of electrode 2 in six dimensions:

[0029] X test Manhattan distance of electrodes in three and six dimensions:

[0030] X test The Manhattan distance of the seven-dimensional non-electric quantity of the electric stove:

[0031] Sort the distances between all electrode 1, electrode 2, electrode 3 and the non-electricity training samples of the electric furnace and And take the first K nearest neighbors; for these K neighbors, calculate the weighted average and predict the result as follows:

[0032] X text The health of the electrode in six dimensions is:

[0033] X test The health of the electrodes in two and six dimensions is:

[0034] X test The health of the electrodes in three and six dimensions is:

[0035] X text The health of the seven non-electricity dimensions of an electric furnace is:

[0036] Among them, y i Indicates the health of the i-th neighbor of the original data; d i Represents Manhattan distance; ∈ represents a small positive number to prevent the denominator from being 0; Indicates the predicted health.

[0037] Furthermore, in S204, the K-nearest neighbor algorithm is used to create a prediction model, and Manhattan distance analysis and calculation are cited.

[0038] Compared with the prior art, the present invention has the following advantages:

[0039] 1. Standardization of data processing of submerged arc furnace

[0040] This invention uses database processing measures to take into account application data deployed locally and in the cloud. It uses threshold filtering, low-pass filtering, custom filtering and other methods to perform standard cleaning on the submerged arc furnace data to improve the accuracy of the submerged arc furnace data.

[0041] 2. Visualization of Submerged Arc Furnace Data Flow

[0042] The present invention adopts data block name extraction means to perform syntax analysis, context association and sequence diagram analysis, automatically obtains basic naming information, comprehensively analyzes the functions and uses of data blocks, displays the interaction process between objects, realizes full-process data visualization, and enhances the convenience of system operation and maintenance.

[0043] 3. Generalization of Submerged Arc Furnace Data Modeling

[0044] The present invention adopts the K-nearest neighbor algorithm to create a prediction model, quotes Manhattan distance analysis and measurement, performs hybrid prediction with multiple models, introduces production process theory, promotes model evaluation and optimization, promptly diagnoses fault causes, and delivers production guidance in real time. Each module has strong versatility and is conducive to rapid transplantation.

[0045] 4. Intelligent application of submerged arc furnace data

[0046] This invention utilizes dynamic IP list management and a network detection module switching mechanism to effectively overcome the limitations of single-node detection and achieve more accurate network status perception. It also employs time type conversion to accurately capture time and translate easily readable information. By applying intelligent algorithm models, it assists production control personnel in decision-making and analysis, enhancing the intelligence level of submerged arc furnaces. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0048] Figure 1 This is a flow chart of the data modeling of the ore-fired furnace of the present invention. DETAILED DESCRIPTION

[0049] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0050] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0051] like Figure 1 As shown, the present invention takes submerged arc furnace data processing and data modeling as the core, adopts dynamic IP network detection technology and time conversion technology, creates an intelligent optimization algorithm application suitable for submerged arc furnaces, comprehensively integrates process variables such as electricity, raw materials, and flue gas, comprehensively processes submerged arc furnace data, extracts data-specific features, combines cloud deployment with edge deployment, applies the K-nearest neighbor algorithm to create a prediction model, quotes Manhattan distance analysis and measurement, performs hybrid prediction with multiple models, introduces production process theory, promotes model evaluation and optimization, timely diagnoses fault causes, pushes production guidance opinions in real time, assists production control personnel in decision-making and analysis, effectively reduces product power consumption, and improves the intelligence level of submerged arc furnaces.

[0052] The present invention provides a data modeling method for a submerged arc furnace, comprising the following steps:

[0053] S201: Preparation of data modeling for an electric arc furnace; the preparation of data modeling for an electric arc furnace includes: preparation of data source, preparation of parameter configuration, preparation of network detection, and preparation of basic general data function blocks;

[0054] S202: Inputting data for submerged arc furnace modeling; the submerged arc furnace modeling data input includes: online input and offline input;

[0055] S203: Submerged arc furnace data modeling preprocessing; the submerged arc furnace data modeling preprocessing includes: data format preprocessing, data filtering preprocessing and auxiliary data preprocessing;

[0056] S204: training and outputting a submerged arc furnace data model; conducting model training and evaluation on the processed data to obtain a trained model; and performing predictions using the trained model to achieve data push and storage.

[0057] As a preferred embodiment, considering the differences in network characteristics between local and cloud deployments, and to adapt to dynamically changing network environments, and to address the misjudgment or missed detection issues inherent in traditional static ping detection, the present invention adds a detection module at the network level, employing dynamic IP list management and a network detection module switching mechanism. This effectively overcomes the limitations of single-node detection, enabling more accurate network status perception and, in turn, accurately determining the validity and authenticity of data sources. The data source preparation includes: preparation of online and offline data sources; the parameter configuration preparation includes: online data, files, and model configuration parameters; and the network detection preparation includes: IP address management, polling ping detection, and an abnormal behavior alarm module.

[0058] As a preferred embodiment, in this application, the IP address pool management module is used to establish an IP address database that includes all node addresses that may participate in network communications. The database should be able to be updated in real time to capture new devices that join or leave the network.

[0059] In this application, the polling ping detection module is used to loop through the IP address list at a certain frequency, send ping requests one by one, and collect response time and success rate data to evaluate the network health.

[0060] In this application, the abnormal behavior alarm module is used to immediately trigger an alarm when it detects that a node has failed to respond multiple times or the response time has increased significantly, reminding the user to take countermeasures.

[0061] Preparation of online data sources and initialization of online data for submerged arc furnaces: Based on the data type, data quantity, and data classification required at the submerged arc furnace site, you can customize the database of required data, provide a database initialization configuration method for on-site data collection, and provide a storage method for online data.

[0062] Offline data source: The submerged arc furnace data is exported from the data stored on site to generate offline files, which can be used for modeling.

[0063] Furthermore, the online input is firstly detected for network connection and then obtained through the raw data online initialization method of the raw data function block; the offline input is obtained through the raw data local initialization method of the raw data function block.

[0064] Furthermore, the data format preprocessing includes: database and file time processing and time type conversion; the data format preprocessing is implemented by a time function block;

[0065] The data filtering preprocessing realizes data filtering by filtering data function block class;

[0066] The auxiliary data preprocessing includes: a spare data function block, an electrode insertion depth function block, a working end length data function block, an average value data function block and a user-defined function block.

[0067] Furthermore, in said S204, the data result set resulting from data cleaning, data dimensionality reduction and data feature extraction is trained through a training data set module, and the target value of the sample data is calibrated by combining manual judgment and function constraints.

[0068] Furthermore, the training data set module includes: a six-dimensional training data set of electrode 1 electric quantity, a six-dimensional training data set of electrode 2 electric quantity, a six-dimensional training data set of electrode 3 electric quantity, and a seven-dimensional training data set of electric furnace non-electric quantity;

[0069] The characteristic vector of the electrode-electricity six-dimensional training data set is expressed as:

[0070]

[0071] The characteristic vector of the six-dimensional training data set of electrode two charges is expressed as:

[0072]

[0073] The characteristic vector of the six-dimensional training data set of the three-electrode electrical quantities is expressed as:

[0074]

[0075] The feature vector of the electrode non-electrical seven-dimensional training data set is expressed as:

[0076]

[0077] For a new sample X test , for each sample in the training set Calculate the Manhattan distance as:

[0078] X test The Manhattan distance of electrode 1 in six dimensions:

[0079] X test The Manhattan distance of electrode 2 in six dimensions:

[0080] Xtest Manhattan distance of electrodes in three and six dimensions:

[0081] X test The Manhattan distance of the seven-dimensional non-electric quantity of the electric stove:

[0082] Sort the distances between all the training samples of electrode 1, electrode 2, electrode 3 and the non-electricity of the electric furnace and And take the first K nearest neighbors; for these K neighbors, calculate the weighted average and predict the result as follows:

[0083] X test The health of the electrode in six dimensions is:

[0084] X test The health of the electrodes in two and six dimensions is:

[0085] X test The health of the electrodes in three and six dimensions is:

[0086] X test The health of the seven non-electricity dimensions of an electric furnace is:

[0087] Among them, y i Indicates the health of the i-th neighbor of the original data; d i Represents Manhattan distance; ∈ represents a small positive number to prevent the denominator from being 0; Indicates the predicted health.

[0088] Based on the model prediction values at the above four levels, users can provide multi-angle operation evaluation standards and equipment status assessment for production. By continuously accumulating operation data with excellent production indicators and extracting efficient production operation parameter combinations, users can guide production to operate in a high-efficiency mode, leading to the healthy and stable operation of the blast furnace equipment.

[0089] Furthermore, in S204, the K-nearest neighbor algorithm is used to create a prediction model, and Manhattan distance analysis and calculation are cited.

[0090] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0091] In the above embodiments of the present invention, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0092] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only exemplary. For example, the division of the units can be a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0093] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0094] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0095] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc. Various media that can store program codes.

[0096] 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 above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above 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 method for submerged arc furnace data modeling, characterized in that: The following steps are involved: S201: Preparation of data modeling for submerged arc furnace; The preparation of the submerged arc furnace data modeling includes: data source preparation, parameter configuration preparation, network detection preparation and basic general data function block preparation; S202: Inputting data for submerged arc furnace modeling; the submerged arc furnace modeling data input includes: online input and offline input; S203: Submerged arc furnace data modeling preprocessing; the submerged arc furnace data modeling preprocessing includes: data format preprocessing, data filtering preprocessing and auxiliary data preprocessing; S204: training and outputting a submerged arc furnace data model; conducting model training and evaluation on the processed data to obtain a trained model; and performing predictions using the trained model to achieve data push and storage.

2. The method for data modeling of a submerged arc furnace according to claim 1, characterized in that: The data source preparation includes: preparation of online data source and preparation of offline data source; the parameter configuration preparation includes: online data, files, model configuration parameters; the network detection preparation includes: IP address management, polling ping detection and abnormal behavior alarm module.

3. The method for data modeling of a submerged arc furnace according to claim 1, characterized in that: The online input is firstly detected for network connection and then obtained through the raw data online initialization method of the raw data function block; the offline input is obtained through the raw data local initialization method of the raw data function block.

4. The method for data modeling of a submerged arc furnace according to claim 1, characterized in that: The data format preprocessing includes: database and file time processing and time type conversion; the data format preprocessing is implemented by the time function block; The data filtering preprocessing realizes data filtering by filtering data function block class; The auxiliary data preprocessing includes: a spare data function block, an electrode insertion depth function block, a working end length data function block, an average value data function block and a user-defined function block.

5. The method for data modeling of a submerged arc furnace according to claim 1, characterized in that: In S204, the data result set resulting from data cleaning, data dimensionality reduction, and data feature extraction is trained through a training data set module, and the target value of the sample data is calibrated by combining manual judgment and function constraints.

6. The method for data modeling of a submerged arc furnace according to claim 5, characterized in that: The training data set module includes: a six-dimensional training data set of electrode one electricity, a six-dimensional training data set of electrode two electricity, a six-dimensional training data set of electrode three electricity, and a seven-dimensional training data set of electric furnace non-electricity; The characteristic vector of the electrode-electricity six-dimensional training data set is expressed as: The characteristic vector of the six-dimensional training data set of electrode two charges is expressed as: The characteristic vector of the six-dimensional training data set of the three-electrode electrical quantities is expressed as: The feature vector of the electrode non-electrical seven-dimensional training data set is expressed as: For a new sample X test , for each sample in the training set Calculate the Manhattan distance as: X test The Manhattan distance of electrode 1 in six dimensions: X test The Manhattan distance of electrode 2 in six dimensions: X test Manhattan distance of electrodes in three and six dimensions: X test The Manhattan distance of the seven-dimensional non-electric quantity of the electric stove: Sort the distances between all electrode 1, electrode 2, electrode 3 and the non-electricity training samples of the electric furnace and And take the first K nearest neighbors; for these K neighbors, calculate the weighted average and predict the result as follows: X test The health of the electrode in six dimensions is: X test The health of the electrodes in two and six dimensions is: X test The health of the electrodes in three and six dimensions is: X test The health of the seven non-electricity dimensions of an electric furnace is: Among them, y i Indicates the health of the i-th neighbor of the original data; d i Represents Manhattan distance; ∈ represents a small positive number to prevent the denominator from being 0; Indicates the predicted health.

7. The method for data modeling of a submerged arc furnace according to claim 1, characterized in that: In S204, a K-nearest neighbor algorithm is used to create a prediction model, and Manhattan distance analysis and calculation are cited.

Images