A real-time optimization method for electromagnetic field control of two-in-one intermediate frequency furnace

Abstract

The present application belongs to the technical field of intermediate frequency power supply for furnace, and particularly relates to a real-time optimization one-to-two intermediate frequency furnace electromagnetic field control method, which comprises an electromagnetic field detection module, a monitoring processing module, an incremental adjustment module and an intelligent algorithm optimization module. The electromagnetic field detection module identifies the electromagnetic field parameters in the furnace, the monitoring processing module collects electromagnetic field data and communicates with the intelligent algorithm optimization module, and the incremental adjustment module adjusts the electromagnetic field parameters according to the algorithm optimization results. The present application can detect and optimize the electromagnetic field parameters in the furnace in real time, effectively reduces the evaporation loss and the occurrence of problems such as overburning and crystallization, improves the smelting efficiency, and at the same time improves the smelting quality of the intermediate frequency furnace. The present application optimizes the control mode of the intermediate frequency furnace, improves the production efficiency, and has high market, social and economic values.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intermediate frequency power supply for furnaces, and particularly relates to a one-to-two intermediate frequency furnace electromagnetic field control method with real-time optimization. BACKGROUND

[0002] The intermediate frequency furnace is a high-efficiency, rapid and powerful heating process equipment, and is widely used in the metallurgical industry such as steelmaking, ferroalloy, aluminum alloy and non-ferrous metal. The electromagnetic field in the furnace has an important influence on the smelting quality and production efficiency. The traditional intermediate frequency furnace control system is based on manual experience to adjust the electromagnetic field parameters, which is prone to cause problems such as overheating, overburning, crystallization and the like due to improper adjustment, thereby affecting the smelting quality. Therefore, a real-time adjustable electromagnetic field control system is needed to improve the smelting efficiency, reduce the molten iron evaporation amount and improve the smelting quality of the intermediate frequency furnace. SUMMARY

[0003] In view of the technical problem that the traditional intermediate frequency furnace control system is based on manual experience to adjust the electromagnetic field parameters and is prone to cause problems due to improper adjustment, the application provides a one-to-two intermediate frequency furnace electromagnetic field control method with real-time optimization.

[0004] In order to solve the above technical problem, the application adopts the following technical scheme:

[0005] A one-to-two intermediate frequency furnace electromagnetic field control method with real-time optimization comprises the following steps:

[0006] S1. A high-sensitivity and high-reliability electromagnetic field sensor is selected to monitor the electromagnetic field intensity and frequency in the furnace in real time.

[0007] S2. A high-speed and high-precision signal collector is selected to convert the data collected by the electromagnetic field sensor into a digital signal and send it into a monitoring and processing module.

[0008] S3. The electromagnetic field data is analyzed and processed for importance, key data is extracted for monitoring and analysis, and the electromagnetic field data in the furnace is stored in a storage unit for subsequent use.

[0009] S4. The real-time data is analyzed to extract valuable information and learn the data law.

[0010] S5. Based on deep learning, reinforcement learning and other algorithms, the electromagnetic field data is processed and optimized to realize real-time control.

[0011] S6. According to the optimization result given by the intelligent algorithm optimization module, a PID control algorithm is used to realize the incremental adjustment of the electromagnetic field.

[0012] S7. A high-speed electromagnetic field controller is used to send the incremental control signal into the furnace to realize accurate adjustment.

[0013] The method for importance analysis and processing of electromagnetic field data in S3 is:

[0014] S3.1, data preprocessing: filtering and denoising the electromagnetic field data to eliminate interference and noise, improve the reliability and accuracy of the data;

[0015] The collected electromagnetic field data is normalized to a certain range, which is convenient for subsequent processing and analysis;

[0016] S3.2, key data extraction: feature extraction of electromagnetic field data, extraction of key features related to the change of electromagnetic field state in the furnace;

[0017] Apply signal processing technology to convert electromagnetic field data to frequency domain or time domain, extract key frequency, amplitude feature parameters;

[0018] S3.3, importance analysis: using statistical analysis method, variation analysis of key feature parameters, and determination of its importance to the electromagnetic field state in the furnace;

[0019] Using data mining technology, the electromagnetic field data is divided into different categories, and the characteristics and rules of different categories of data are analyzed to determine the importance of each category of data;

[0020] S3.4, anomaly detection: design anomaly detection algorithm, anomaly detection of electromagnetic field data, identify abnormal data inconsistent with normal furnace operation state;

[0021] Based on the results of anomaly detection, the abnormal data is marked and processed for subsequent control and adjustment;

[0022] S3.5, decision analysis: based on the key feature parameters obtained by importance analysis, combined with the results of anomaly detection, comprehensive analysis and decision;

[0023] Formulate appropriate control strategy, according to the actual change of electromagnetic field in the furnace, determine the way and degree of adjusting electromagnetic field parameters, in order to realize the real-time control of electromagnetic field;

[0024] S3.6, data visualization and alarm: visualization of electromagnetic field data after analysis and processing, to intuitively reflect the change trend and state of electromagnetic field in the furnace.

[0025] The signal processing technology in S3.2 uses wavelet transform or spectrum analysis, and the data mining technology in S3.3 uses coefficient of variation or correlation analysis.

[0026] The method for analyzing real-time data, extracting valuable information and learning data rules in S4 is:

[0027] S4.1, data collection and preprocessing

[0028] Obtain electromagnetic field data from the monitoring processing module and perform data cleaning, denoising, and normalization preprocessing operations to ensure data quality and availability;

[0029] S4.2, feature extraction

[0030] After preprocessing the electromagnetic field data, use feature extraction methods to extract representative features from the data;

[0031] Consider the potential rules contained in the data during feature extraction and select appropriate feature extraction methods to effectively represent the electromagnetic field data;

[0032] S4.3, training data model

[0033] Establish appropriate data models based on the data sets obtained through preprocessing and feature extraction;

[0034] Select machine learning methods suitable for electromagnetic field data analysis to train the model and learn the data rules;

[0035] S4.4, model verification and evaluation

[0036] Use the established data model to verify a portion of unused electromagnetic field data to evaluate the prediction or classification accuracy of the model;

[0037] Use cross-validation or confusion matrix methods to verify and evaluate the model;

[0038] S4.5, model application and continuous learning

[0039] In real-time applications, input real-time collected electromagnetic field data into the learned model to perform data analysis and prediction using the model.

[0040] The feature extraction method in S4.2 includes statistical features, frequency domain features, and time domain features; the machine learning method for electromagnetic field data analysis in S4.3 uses support vector machine SVM or neural network NN.

[0041] The method for processing and optimizing electromagnetic field data in S5 is:

[0042] S5.1, data preprocessing

[0043] Perform data cleaning, denoising, and anomaly detection preprocessing operations on the electromagnetic field data collected from the monitoring processing module to improve data quality and accuracy;

[0044] S5.2, feature selection

[0045] Selecting features related to changes in the electromagnetic field state in the furnace from pre-processed electromagnetic field data; using statistical analysis or domain knowledge to select features;

[0046] S5.3, Model training and optimization

[0047] Based on the selected features, a data model is established;

[0048] Use the training set data to train the model, and combine cross-validation and other methods to optimize the model parameters, to improve the fitting ability and generalization ability of the model;

[0049] S5.4, Model application and feedback

[0050] Input the real-time collected electromagnetic field data into the trained and optimized model, and use the model to analyze and predict the electromagnetic field in the furnace;

[0051] Compare the model output results with the actual electromagnetic field data in the furnace, evaluate the prediction accuracy of the model, and use the comparison results as feedback information for model updating and optimization;

[0052] S5.5, Model updating and optimization

[0053] According to the feedback information, periodically update and optimize the model to adapt to the dynamic changes of the electromagnetic field in the furnace;

[0054] Using incremental learning technology, combine new data with existing models to realize iterative updating of the model, continuously improve the accuracy and robustness of the model;

[0055] S5.6, Real-time control and adjustment

[0056] Based on the optimized model, generate incremental control signals for electromagnetic field adjustment, apply the signals to the intermediate frequency furnace electromagnetic field controller through the incremental adjustment module, realize real-time control and adjustment of the electromagnetic field;

[0057] The monitoring and processing module collects electromagnetic field data again and feeds back to the intelligent algorithm optimization module to form a closed loop control, realizing continuous real-time optimization.

[0058] The S5.2 uses statistical analysis or domain knowledge to select features, including electromagnetic field strength, frequency and rate of change; the S5.3 data model uses a deep learning model or a reinforcement learning model.

[0059] The S6 uses a PID control algorithm to realize incremental adjustment of the electromagnetic field.

[0060] S6.1, Parameter initialization

[0061] Initialize three parameters of the PID control algorithm: proportional coefficient Kp, integral coefficient Ki and differential coefficient Kd.

[0062] Preliminary setting according to actual situation and experience, and subsequent adjustment and optimization according to closed-loop control effect;

[0063] S6.2, error calculation

[0064] Calculate the error e between the electromagnetic field set value and the actual value as the input of the PID control algorithm; the error is calculated by the difference between the actual value and the set value, and the electromagnetic field intensity or frequency is selected as the set value;

[0065] S6.3, increment calculation

[0066] According to the increment calculation formula of the PID control algorithm, calculate the increment Δu:

[0067] Δu = Kp * e + Ki * ∫e dt + Kd * de / dt

[0068] The Kp, Ki and Kd are proportional coefficient, integral coefficient and differential coefficient respectively, the e is error, the t is time, and the de / dt is the change rate of error;

[0069] S6.4, increment output

[0070] Output the increment Δu to the electromagnetic field controller of the intermediate frequency furnace to realize the incremental adjustment of the electromagnetic field;

[0071] The increment represents increasing or decreasing the intensity or frequency of the electromagnetic field, and the specific adjustment direction and amplitude are determined according to the positive and negative of the increment Δu;

[0072] S6.5, parameter update

[0073] According to the actual control effect, update and optimize the three parameters of the PID control algorithm; adjust the parameters to improve the stability and performance of the control;

[0074] S6.6, repeat iteration

[0075] Cyclically execute S6.2 to S6.5 to realize the continuous incremental adjustment of the electromagnetic field;

[0076] In each iteration, calculate the increment according to the latest error, and update the parameters to obtain better closed-loop control effect.

[0077] The method of adjusting the parameters in S6.5 adopts adaptive control algorithm and genetic algorithm.

[0078] Compared with the prior art, the present application has the beneficial effects that:

[0079] The application provides a novel one-to-two intermediate frequency furnace electromagnetic field control system, which can detect and optimize electromagnetic field parameters in the furnace in real time, effectively reduces evaporation loss and occurrence of problems such as overburning and crystallization, improves smelting efficiency, and improves smelting quality of the intermediate frequency furnace. The application optimizes the control mode of the intermediate frequency furnace, improves production efficiency, and has high market, social and economic values. BRIEF DESCRIPTION OF DRAWINGS

[0080] In order to more clearly illustrate the embodiments of the application or the technical solutions in the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only exemplary, and for those skilled in the art, other drawings can be obtained from the provided drawings without creative labor.

[0081] Figure 1 The application provides a step flow chart. DETAILED DESCRIPTION

[0082] In order to make the purpose, technical scheme and advantages of the embodiments of the application more clear, the technical scheme in the embodiments of the application will be clearly and completely described below. Obviously, the described embodiments are only some of the embodiments of the application, but not all the embodiments of the application. These descriptions are only for further illustrating the features and advantages of the application, but not for limiting the claims of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the application.

[0083] The specific embodiments of the application will be further described in detail below in combination with the drawings and embodiments. The following embodiments are used to illustrate the application, but not to limit the scope of the application.

[0084] In the embodiment, as shown in the figure, the specific embodiments are as follows: Figure 1

[0085] 1. Electromagnetic field detection module:

[0086] (1) Electromagnetic field sensor: high sensitivity and high reliability electromagnetic field sensor is selected to monitor the electromagnetic field intensity and frequency in the furnace in real time.

[0087] (2) Electromagnetic field signal collector: high-speed and high-precision signal collector is selected to convert the data collected by the electromagnetic field sensor into digital signals and send them into the monitoring and processing module.

[0088] 2. Monitoring and processing module:

[0089] ​(1) Data processing unit: important analysis and processing of electromagnetic field data, extraction of key data for monitoring and analysis.

[0090] The specific method steps of important analysis and processing of electromagnetic field data in the data processing unit are as follows:

[0091] Step 1: Data preprocessing

[0092] Filter and denoise the electromagnetic field data to eliminate interference and noise, improve data reliability and accuracy.

[0093] Normalize the collected electromagnetic field data to a certain range for subsequent processing and analysis.

[0094] Step 2: Key data extraction

[0095] Feature extraction of electromagnetic field data, extraction of key features related to the change of electromagnetic field state in the furnace.

[0096] Apply signal processing techniques such as wavelet transform, frequency spectrum analysis, etc. to convert electromagnetic field data to frequency domain or time domain, extract key frequency, amplitude and other characteristic parameters.

[0097] Step 3: Importance analysis

[0098] Use statistical analysis methods such as coefficient of variation, correlation analysis, etc. to analyze the variability of key characteristic parameters and determine their importance to the electromagnetic field state in the furnace.

[0099] Use data mining techniques such as clustering analysis, classification analysis, etc. to divide electromagnetic field data into different categories, analyze the characteristics and rules of different categories of data, and determine the importance of each category of data.

[0100] Step 4: Abnormal detection

[0101] Design appropriate anomaly detection algorithm to detect anomalies in electromagnetic field data and identify abnormal data that does not conform to normal furnace operation state.

[0102] Based on the results of anomaly detection, mark and process the abnormal data for subsequent control and adjustment.

[0103] Step 5: Decision analysis

[0104] Based on the key characteristic parameters obtained by importance analysis, combined with the results of anomaly detection, comprehensive analysis and decision-making are carried out.

[0105] Develop appropriate control strategies, determine the way and degree of adjusting electromagnetic field parameters according to the actual changes of electromagnetic field in the furnace, to realize real-time control of electromagnetic field.

[0106] Step 6: Data visualization and alarm

[0107] The processed electromagnetic field data is visualized to intuitively reflect the trend and state of the electromagnetic field in the furnace.

[0108] Set reasonable thresholds, and generate alarm information when the electromagnetic field data exceeds the preset range or abnormal situations occur, so that the operator can take timely measures.

[0109] Through the implementation of the above specific method steps, the data processing unit can analyze and process the electromagnetic field data, extract key information and make decisions to achieve precise control and optimization of the electromagnetic field in the furnace.

[0110] (2) Storage unit: Store the electromagnetic field data in the furnace for subsequent use.

[0111] (3) Network unit: Provide electromagnetic field data communication function, intelligent algorithm optimization, and remote adjustment.

[0112] 3. Intelligent algorithm optimization module:

[0113] (1) Data mining module: Analyze real-time data, extract valuable information, and learn data rules.

[0114] The specific method steps of the data mining module in the real-time data analysis, extraction of valuable information, and learning of data rules are as follows:

[0115] Step 1: Data collection and preprocessing

[0116] Get electromagnetic field data from the monitoring and processing module, and perform data cleaning, denoising, normalization and other preprocessing operations to ensure data quality and usability.

[0117] Step 2: Feature extraction

[0118] Extract representative features from the preprocessed electromagnetic field data using feature extraction methods such as statistical features, frequency domain features, and time domain features.

[0119] Consider the potential rules contained in the data during feature extraction, and select appropriate feature extraction methods to effectively represent the electromagnetic field data.

[0120] Step 3: Train data model

[0121] Establish appropriate data models based on the data sets obtained through preprocessing and feature extraction.

[0122] Machine learning methods suitable for electromagnetic field data analysis, such as support vector machines (SVM), neural networks (NN), etc., can be selected for model training to learn data patterns.

[0123] Step 4: Model verification and evaluation

[0124] Using the established data model, a portion of unused electromagnetic field data is verified to evaluate the prediction or classification accuracy of the model.

[0125] Methods such as cross-validation, confusion matrix, etc. can be used for model verification and evaluation.

[0126] Step 5: Model application and continuous learning

[0127] In real-time applications, real-time collected electromagnetic field data is input into the learned model for data analysis and prediction.

[0128] The historical patterns, current state and future trends of the data should be considered in the model application process to optimize electromagnetic field control.

[0129] Step 6: Model updating and optimization

[0130] With the accumulation of data and continuous application, the model is updated and optimized based on new data feedback to adapt to the changes and complexity of electromagnetic field data.

[0131] Incremental learning methods can be used to combine new electromagnetic field data with existing models to achieve dynamic updating and improvement of the model.

[0132] Through the implementation of the above specific method steps, the data mining module can analyze real-time electromagnetic field data, extract valuable information, learn data patterns, and use learned models to predict and optimize control of electromagnetic field data.

[0133] (2) Intelligent optimization module: based on deep learning, reinforcement learning, etc. to process and optimize electromagnetic field data for real-time control.

[0134] In the intelligent optimization module, the specific method steps for processing and optimizing electromagnetic field data are as follows:

[0135] Step 1: Data preprocessing

[0136] Data cleaning, denoising, anomaly detection, etc. are performed on the electromagnetic field data collected from the monitoring and processing module to improve data quality and accuracy.

[0137] Step 2: Feature selection

[0138] Select features related to changes in the electromagnetic field state from the pre-processed electromagnetic field data for modeling.

[0139] Statistical analysis or domain knowledge can be used to select appropriate features, such as electromagnetic field strength, frequency, rate of change, etc.

[0140] Step 3: Model training and optimization

[0141] Based on the selected features, establish appropriate data models, such as deep learning models, reinforcement learning models, etc.

[0142] Use the training set data to train the model, and combine cross-validation and other methods to optimize the model parameters, to improve the fitting ability and generalization ability of the model.

[0143] Step 4: Model application and feedback

[0144] Input real-time collected electromagnetic field data into the trained and optimized model, and use the model to analyze and predict the electromagnetic field in the furnace.

[0145] Compare the model output results with the actual electromagnetic field data in the furnace, evaluate the prediction accuracy of the model, and use the comparison results as feedback information for model updating and optimization.

[0146] Step 5: Model updating and optimization

[0147] According to the feedback information, periodically update and optimize the model to adapt to the dynamic changes of the electromagnetic field in the furnace.

[0148] Incremental learning and other technologies can be used to combine new data with existing models to achieve iterative updates of the model, continuously improving the accuracy and robustness of the model.

[0149] Step 6: Real-time control and adjustment

[0150] Based on the optimized model, generate incremental control signals for electromagnetic field adjustment, and apply the signals to the intermediate frequency furnace electromagnetic field controller through the incremental adjustment module to realize real-time control and adjustment of the electromagnetic field.

[0151] The monitoring and processing module collects electromagnetic field data again and feeds back to the intelligent algorithm optimization module, forming a closed loop control to achieve continuous real-time optimization.

[0152] Through the implementation of the above specific method steps, the intelligent optimization module can effectively process and optimize electromagnetic field data, and realize accurate control and optimization of the electromagnetic field in the furnace according to the learning ability and optimization ability of the model, improving the production efficiency and quality in the furnace.

[0153] 4. Incremental adjustment module:

[0154] (1) Adjustment algorithm: According to the optimization results given by the intelligent algorithm optimization module, the PID control algorithm is used to realize the incremental adjustment of the electromagnetic field.

[0155] The specific method steps of using PID control algorithm to realize the incremental adjustment of the electromagnetic field in the incremental adjustment module are as follows:

[0156] Step 1: Parameter initialization

[0157] Initialize the three parameters of the PID control algorithm: proportional coefficient (Kp), integral coefficient (Ki) and differential coefficient (Kd).

[0158] The initial settings can be made according to actual conditions and experience, and subsequent adjustments and optimizations can be made according to the closed-loop control effect.

[0159] Step 2: Error calculation

[0160] Calculate the error (e) between the set value and the actual value of the electromagnetic field as the input of the PID control algorithm.

[0161] The error can be calculated by the difference between the actual value and the set value, and the electromagnetic field intensity or frequency can be selected as the set value.

[0162] Step 3: Increment calculation

[0163] According to the incremental calculation formula of the PID control algorithm, calculate the increment (Δu):

[0164] Δu = Kp * e + Ki * ∫e dt + Kd * de / dt

[0165] Where Kp, Ki, Kd are the proportional coefficient, integral coefficient and differential coefficient respectively, e is the error, t is the time, and de / dt is the rate of change of error.

[0166] Step 4: Increment output

[0167] Output the increment (Δu) to the electromagnetic field controller of the intermediate frequency furnace to realize the incremental adjustment of the electromagnetic field.

[0168] The increment can be expressed as increasing or decreasing the intensity or frequency of the electromagnetic field, and the specific adjustment direction and amplitude can be determined according to the positive and negative of the increment (Δu).

[0169] Step 5: Parameter update

[0170] According to the actual control effect, update and optimize the three parameters of the PID control algorithm.

[0171] The parameters can be adjusted by adaptive control algorithm, genetic algorithm and other methods to improve the stability and performance of the control.

[0172] Step 6: Repeat iteration

[0173] Cyclically execute steps 2 to 5, achieve the continuous incremental adjustment of the electromagnetic field.

[0174] In each iteration, the increment is calculated according to the latest error, and the parameters are updated to obtain better closed-loop control effect.

[0175] Through the implementation of the above specific method steps, the incremental adjustment module can adjust the increment of the electromagnetic field according to the PID control algorithm, and realize the precise control of the electromagnetic field of the intermediate frequency furnace.

[0176] (2) Increment input unit: use high-speed electromagnetic field controller to send incremental control signal into the furnace to realize precise adjustment.

[0177] The above only details the preferred embodiments of the present application, but the present application is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the purpose of the present application, and all the changes shall be included in the protection scope of the present application.

Claims

1. A method for real-time optimization of electromagnetic field control in a two- furnace intermediate frequency (2IF) furnace, the method comprising: It comprises the following steps: ​ S1, select high sensitivity, high reliability of electromagnetic field sensor, real-time monitoring of electromagnetic field intensity and frequency in the furnace; S2, select high speed, high precision signal acquisition, the data collected by electromagnetic field sensor into digital signal, sent to the monitoring processing module; S3, the importance of electromagnetic field data analysis and processing, extraction of key data for monitoring and analysis, storage unit stores the electromagnetic field data in the furnace, for subsequent use; S4, according to the real-time data analysis, extraction of valuable information, learning data rule; S5, based on deep learning or reinforcement learning algorithm, the electromagnetic field data processing and optimization, realize real-time control; S6, according to the intelligent algorithm optimization module given the optimization results, using PID control algorithm to realize the incremental adjustment of electromagnetic field; S7, using high-speed electromagnetic field controller, the incremental control signal into the furnace, realize accurate regulation.

2. The method according to claim 1, wherein: The method for analyzing and processing the electromagnetic field data in S3 is: S3.1, data preprocessing: filtering and denoising the electromagnetic field data, eliminating interference and noise, improving the reliability and accuracy of the data; The collected electromagnetic field data is normalized, which is standardized to a certain range, so as to facilitate the subsequent processing and analysis; S3.2, key data extraction: feature extraction of electromagnetic field data, extraction of key features related to the change of electromagnetic field state in the furnace; Using signal processing technology, the electromagnetic field data is converted to frequency domain or time domain, and the key frequency and amplitude characteristic parameters are extracted; S3.3, importance analysis: using statistical analysis method, the variation of key characteristic parameters is analyzed, and the importance of electromagnetic field state in the furnace is determined; Using data mining technology, the electromagnetic field data is divided into different categories, the characteristics and rules of different categories of data are analyzed, and the importance of each category of data is determined; S3.4, anomaly detection: design anomaly detection algorithm, detect the anomaly of electromagnetic field data, identify the abnormal data which is not consistent with the normal furnace operation state; Based on the results of anomaly detection, the abnormal data is marked and processed for subsequent control and adjustment; S3.5, decision analysis: based on the key characteristic parameters obtained by importance analysis, combined with the results of anomaly detection, comprehensive analysis and decision are made; Formulate the corresponding control strategy, according to the actual change of electromagnetic field in the furnace, determine the way and degree of adjusting electromagnetic field parameters, in order to realize the real-time control of electromagnetic field; S3.6, data visualization and alarm: the electromagnetic field data after analysis and processing is visualized to reflect the change trend and state of electromagnetic field in the furnace.

3. The method according to claim 2, wherein: The signal processing technology in S3.2 adopts wavelet transform or spectrum analysis, and the data mining technology in S3.3 adopts coefficient of variation or correlation analysis.

4. The method of claim 1, wherein the method is a real-time optimization method for controlling the electromagnetic field of a two-in-one intermediate frequency furnace. The method for analyzing real-time data and extracting valuable information in S4 is: S4.1, data collection and preprocessing Get electromagnetic field data from monitoring processing module, and perform data cleaning, denoising and normalization preprocessing operation to ensure data quality and availability; S4.2, feature extraction The pre-processed electromagnetic field data is extracted by a feature extraction method to extract representative features in the data; In the feature extraction process, the potential rules contained in the data are considered, and a suitable feature extraction method is selected to effectively represent the electromagnetic field data; S4.3, training data model According to the data set obtained by pre-processing and feature extraction, a suitable data model is established; Select a machine learning method suitable for electromagnetic field data analysis to train the model and learn the data rules; S4.4, model verification and evaluation A part of electromagnetic field data not used is verified by using the established data model to evaluate the prediction or classification accuracy of the model; Cross-validation or confusion matrix method is used to verify and evaluate the model; S4.5, model application and continuous learning In real-time application, real-time collected electromagnetic field data is input into the learned model to analyze and predict the data by using the model.

5. The method of claim 4, wherein the method is a real-time optimization method for controlling the electromagnetic field of a two-in-one intermediate frequency furnace. The feature extraction method in S4.2 includes statistical features, frequency domain features, and time domain features; The machine learning method for electromagnetic field data analysis in S4.3 uses support vector machine (SVM) or neural network (NN).

6. The method of claim 1, wherein the method is a real-time optimization method for controlling the electromagnetic field of a two-in-one intermediate frequency furnace. The method for processing and optimizing electromagnetic field data in S5 is: S5.1, data preprocessing Data cleaning, denoising, and anomaly detection preprocessing operations are performed on the electromagnetic field data collected from the monitoring processing module to improve the quality and accuracy of the data; S5.2, feature selection Features related to the change of the electromagnetic field state in the furnace are selected from the pre-processed electromagnetic field data for modeling; Statistical analysis or domain knowledge is used to select features; S5.3, model training and optimization Based on the selected features, a data model is established; The model is trained using the training set data, and the cross-validation method is used to optimize the model parameters to improve the fitting ability and generalization ability of the model; S5.4, model application and feedback Real-time collected electromagnetic field data is input into the trained and optimized model to analyze and predict the electromagnetic field in the furnace; The model output is compared with the actual electromagnetic field data in the furnace to evaluate the prediction accuracy of the model, and the comparison result is used as feedback information for model updating and optimization; S5.5, model updating and optimization According to the feedback information, the model is periodically updated and optimized to adapt to the dynamic changes of the electromagnetic field in the furnace; Incremental learning technology is used to combine new data with the existing model to realize iterative updating of the model and continuously improve the accuracy and robustness of the model; S5.6, real-time control and adjustment Based on the optimized model, an incremental control signal for electromagnetic field adjustment is generated, which is applied to the intermediate frequency furnace electromagnetic field controller through the incremental adjustment module to realize real-time control and adjustment of the electromagnetic field; The monitoring processing module collects electromagnetic field data again and feeds back to the intelligent algorithm optimization module to form a closed-loop control and realize continuous real-time optimization.

7. The method according to claim 6, wherein: In S5.2, statistical analysis or domain knowledge is used to select features including electromagnetic field strength, frequency, and rate of change; in S5.3, the data model uses a deep learning model or a reinforcement learning model.

8. The method of claim 1, wherein the method is a real-time optimization method for controlling the electromagnetic field of a two-in-one intermediate frequency furnace. The method for realizing the incremental adjustment of the electromagnetic field in S6 adopts a PID control algorithm: S6.1, Parameter initialization Initialize the three parameters of the PID control algorithm: the proportional coefficient Kp, the integral coefficient Ki, and the differential coefficient Kd; Preliminary settings are made according to actual conditions and experience, and subsequent adjustments and optimizations are made according to the closed-loop control effect; S6.2, Error calculation Calculate the error e between the set value and the actual value of the electromagnetic field as the input of the PID control algorithm; the error is calculated by the difference between the actual value and the set value, and the electromagnetic field intensity or frequency is selected as the set value; S6.3, Increment calculation According to the incremental calculation formula of the PID control algorithm, calculate the increment ∆u: ∆u = Kp·e + Ki·∫e dt + Kd·de / dt The Kp, Ki, and Kd are the proportional coefficient, integral coefficient, and differential coefficient, respectively, the e is the error, the t is the time, and the de / dt is the error rate; S6.4, Increment output Output the increment ∆u to the electromagnetic field controller of the intermediate frequency furnace to realize the incremental adjustment of the electromagnetic field; The increment represents the increase or decrease of the intensity or frequency of the electromagnetic field, and the specific adjustment direction and amplitude are determined according to the positive and negative of the increment ∆u; S6.5, Parameter update According to the actual control effect, update and optimize the three parameters of the PID control algorithm; adjust the parameters to improve the stability and performance of the control; S6.6, Repeat iteration Cyclically execute S6.2 to S6.5 to realize the continuous incremental adjustment of the electromagnetic field; In each iteration, calculate the increment according to the latest error and update the parameters to obtain better closed-loop control effect.

9. The method according to claim 8, wherein: The method for adjusting the parameters in S6.5 adopts an adaptive control algorithm and a genetic algorithm.

Images