Data-driven method for determining quenching parameters of medium and thick steel plates

Abstract

The invention relates to the field of metallurgical technology, and in particular to a method for determining quenching parameters of medium and thick steel plates based on data-driven, comprising the following steps: step one, arranging a plurality of temperature sensors on a steel plate sample to be quenched according to preset measuring point positions; step two, collecting data recorded by a plurality of temperature sensors through a data acquisition device, and drawing a temperature change curve; step three, extracting a plurality of key characteristic values ​​from the drawn temperature change curve; step four, inputting the plurality of key characteristic values ​​as input variables into a prediction model, and evaluating the prediction results; step five, adjusting the measuring point position and depth, and storing them as a new measurement scheme, assigning a unique identification label to each new measurement scheme, and reinstalling the temperature sensor according to the new measurement scheme; step six, repeatedly executing steps two to five to determine the optimal measurement scheme, and collecting quenching parameters through the optimal measurement scheme.

Description

Technical Field

[0001] The invention relates to the technical field of metallurgy, and in particular to a method for determining quenching parameters of medium and thick steel plates based on data driving. Background Art

[0002] Medium and thick plates refer to steel plates with a thickness of 4.5-25.0mm, 25.0-100.0mm is called thick plate, and those with a thickness of more than 100.0mm are called extra thick plates. Medium and thick plates are mainly used in construction engineering, machinery manufacturing, container manufacturing, shipbuilding, bridge construction, etc. In the current fierce competition in the steel industry, various companies are gradually turning to a development track that emphasizes quality and cost. The new generation of high-precision medium and thick plate rolling control system based on big data analysis developed by Northeastern University is a model. The system not only improves the economic benefits of the company, but also enhances the company's ability to resist market risks.

[0003] Based on the concept and technology of big data, a massive data collection and management system covering the production process of medium and thick steel plates can be established. By real-time monitoring or recording various parameters in the production process, including temperature, cooling rate, internal stress of steel plates, etc., the important information contained therein can be mined and studied. After preprocessing, these data can be used for further analysis and model building.

[0004] In summary, the data-driven method for determining the quenching parameters of medium and thick steel plates has important practical significance and significant social benefits. Through the comprehensive application of big data analysis, machine learning algorithms and adaptive modeling technology, the quenching process of medium and thick steel plates can be precisely controlled, thereby improving product quality, reducing costs and enhancing the competitiveness of enterprises. In the future, with the continuous advancement and innovation of technology, this field will usher in more developments and breakthroughs. Summary of the invention

[0005] The object of the present invention is to provide a technical solution to analyze the collected data and establish prediction models. These models can predict the steel plate properties after quenching based on historical data, so as to adjust the measurement plan and use the optimal measurement plan to measure the quenching parameters to achieve the best effect.

[0006] The technical solution of the present invention is: a method for determining quenching parameters of medium and thick steel plates based on data drive, the method comprising the following steps:

[0007] Step 1: Arrange multiple temperature sensors on the steel plate sample to be quenched according to the preset measurement point positions;

[0008] Step 2: Collect data recorded by multiple temperature sensors through a data acquisition device and draw a temperature change curve;

[0009] Step 3: extract multiple key characteristic values ​​from the drawn temperature change curve;

[0010] Step 4: Input multiple key feature values ​​as input variables into the prediction model, record the predicted values ​​output by the model, and evaluate the prediction results;

[0011] Step 5: Adjust the position and depth of the measurement point and store it as a new measurement plan, assign a unique identification tag to each new measurement plan, and reinstall the temperature sensor according to the new measurement plan;

[0012] Step 6: Repeat steps 2 to 5 to determine the optimal measurement scheme, and collect quenching parameters through the optimal measurement scheme.

[0013] Preferably, the step 1 comprises the following steps:

[0014] Select multiple measuring points according to the size and shape of the selected steel plate sample, and the number of measuring points shall not be less than 6;

[0015] Drill holes at the measuring point locations that match the number of measuring points, and install temperature sensors in the holes according to the measuring process requirements;

[0016] The new measurement scheme includes the position information of the measuring point on the steel plate and the depth information inside the steel plate.

[0017] Preferably, the step 2 comprises the following steps:

[0018] The temperature sensor is connected to the data collector, and the data collector establishes a data connection with the cloud. The data collector sends the acquired temperature data to the cloud for preprocessing, and adds a time tag corresponding to the acquisition time to the preprocessed temperature data. The temperature data with the time tag constitutes a temperature-time data set;

[0019] The drawing of the temperature variation curve comprises:

[0020] Get the temperature-time dataset from the cloud and download it to the local storage device;

[0021] Use the matplotlib drawing library to generate a time series temperature change curve graph based on the temperature-time dataset.

[0022] Preferably, the step three comprises the following steps:

[0023] Use Python or MATLAB to load the temperature-time dataset;

[0024] By finding the maximum and minimum values, we can obtain the highest temperature value U max and the minimum temperature U min ;

[0025] Calculate the time length t from the highest temperature moment to the lowest temperature moment, t is the cooling time;

[0026] By calculating the difference ΔU between the highest temperature and the lowest temperature, we get the cooling rate V = ΔU / t;

[0027] The temperature-time data is constructed into a time series of temperature changes over time, and a numerical differentiation method is used to obtain multiple gradient values ​​p and form a gradient data set;

[0028] Set the gradient change threshold ΔY and traverse the data set. When |p|≤ΔY, the gradient does not change significantly. When |p|>ΔY, the gradient changes significantly. Record the gradient value and time point at this time.

[0029] Preferably, constructing the temperature-time data into a time series of temperature changes over time comprises the following steps:

[0030] Constructing the temperature-time data into a time series of temperature changes over time includes the following steps:

[0031] Matching the temperature data acquired by the temperature sensor with the corresponding time point;

[0032] Convert data to CSV or Pandas DataFrame format;

[0033] Use the time series tool in Python to create a time series object, using time as the index and temperature as the value.

[0034] Preferably, the step 4 comprises the following steps:

[0035] Building a predictive model: Among them, X1, X2, X3, X4, X5 are input variables; α, β, γ, δ, ρ0, λ are regression coefficients; X0ρ0 is the intercept; is the output variable;

[0036] Train the prediction model:

[0037] Retrieve historical temperature change curves and corresponding steel plate performance data from a database;

[0038] Clean the steel plate performance data, including removing outliers, processing missing values, and possible data conversion to ensure data quality;

[0039] Extract key feature values ​​from the historical temperature curve, and use these extracted key feature values ​​as input variables to construct a feature set for training a prediction model;

[0040] The feature set is divided into a training set and a test set, and the model is trained or fitted by the least squares method using the feature data (input variables) in the training set and the corresponding steel plate performance data (output variables);

[0041] Model Evaluation:

[0042] Set the model evaluation threshold Δy, when When , the model is valid, record Y c ; Where Y is the actual measured value, is the output value of the prediction model.

[0043] Preferably, each new measurement scheme corresponds to a new temperature change curve, new key feature values ​​are extracted from the new temperature change curve, and these new feature values ​​are used as input variables to construct a new feature set;

[0044] Using the data in the new feature set as the input variable of the prediction model can obtain the corresponding output variable

[0045] if Then record the Y at this time ic ;

[0046] where Y i is the i-th measurement value. After repeating multiple measurement schemes, multiple Y ic Values, forming a threshold set {Y ic}, where i = 1, 2, 3, ..., n, and n is the number of solutions.

[0047] Preferably, the elements in the threshold set are arranged in ascending order, and the smallest Y ic The measurement scheme corresponding to the value is the optimal measurement scheme, and the optimal measurement scheme is used to collect quenching parameters.

[0048] The present invention also provides an electronic device, which includes a processor, a memory connected to the processor, and a communication module, and the electronic device executes the data-driven method for determining quenching parameters of medium and thick steel plates.

[0049] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement the data-driven method for determining quenching parameters of medium and thick steel plates.

[0050] Beneficial effects of the present invention:

[0051] 1. In the present invention, according to the preset measurement point positions, appropriate positions will be selected on the steel plate sample to arrange temperature sensors. These measurement points are distributed on the entire steel plate sample to ensure that the temperature changes at different positions can be fully monitored. During the quenching process, the steel plate sample will be heated to a high temperature and then rapidly cooled to change its microstructure and mechanical properties. During this process, the temperature sensor will monitor and record the data of the temperature sensor, and detailed information about the temperature changes at different positions of the steel plate sample can be obtained; these data will help us understand the temperature distribution during the quenching process and the impact of temperature changes at different positions on the performance of the steel plate; according to the collected temperature data, the quenching process can be optimized and adjusted to improve the quality and performance of the steel plate; by analyzing the temperature data, we can determine the best quenching parameters, such as heating temperature, cooling rate, etc., to achieve the desired microstructure and mechanical properties. By arranging multiple temperature sensors on the steel plate sample to be quenched, and monitoring and recording the temperature data, we can better understand the temperature changes during the quenching process, thereby optimizing the quenching process and improving the quality and performance of the steel plate.

[0052] 2. In the present invention, by connecting temperature sensors, data collectors and using the cloud to preprocess data and draw temperature change curves; each temperature sensor is connected to a data collector, which usually involves connecting the output interface of the sensor to the input port of the data collector, and setting the data collector to establish a stable data connection with the cloud through wired (such as Ethernet) or wireless (such as Wi-Fi, 4G / 5G) methods. The data collector sends the acquired temperature data to the cloud regularly or in real time, which includes setting the frequency of data transmission, the target cloud server address, etc. In the cloud, a data receiving point is established, such as a cloud server or a data center, to receive temperature data from the data collector. The collected temperature data is preprocessed in the cloud, which may include filtering noise, data smoothing, format conversion and other operations to improve data quality. After the preprocessing is completed, a time label corresponding to the acquisition time is added to each data point to ensure the timing of the data. This time label should be consistent with the timestamp of the data recorded by the data collector. The temperature data with time tags are organized into a temperature-time dataset in a format that is easy to store and transmit, such as CSV or JSON. The organized temperature-time dataset is stored in a cloud database or storage system for subsequent access and use. The temperature-time dataset is obtained from the cloud and downloaded to a local storage device, such as a computer or mobile device. The Python matplotlib library is used to generate a time series temperature change curve chart based on the temperature-time dataset, realizing a complete process from field data collection to cloud preprocessing, and then to local data analysis and visualization, providing a scientific basis for monitoring and control of the quenching process.

[0053] 3. In the present invention, a prediction model is constructed, and the optimal measurement scheme is found through iterative optimization; in this embodiment, the prediction model adopts a linear regression model, wherein the input variables represent key feature values ​​extracted from the temperature change curve, such as the maximum temperature, cooling rate, etc.; the output variable is the predicted steel plate performance data, such as hardness or strength, and the historical temperature change curve and the corresponding steel plate performance data are retrieved from the database, and the key feature values ​​are extracted from the historical measured temperature curves. These features are used as input variables to construct a feature set, and the feature set is divided into a training set and a test set. The model is trained using the feature data (input variables) in the training set and the corresponding steel plate performance data (output variables). The least squares method is used to train or fit the model. This process involves finding the best regression coefficient and intercept so that the residual sum of squares between the predicted value and the actual value is minimized. Small, in order to evaluate the effectiveness of the model, set an evaluation threshold, above this threshold, the model is considered valid, when the model is valid, record the model parameters (regression coefficient and intercept) at this time, for each new measurement scheme, extract new key feature values ​​and build a new feature set, use these new features as input variables, and obtain the corresponding output variables through the prediction model, if the prediction result of the new measurement scheme is better than the previous one, record the model parameters at this time, after repeating multiple measurement schemes, arrange the elements in the threshold set in ascending order, and select the measurement scheme corresponding to the smallest value as the optimal scheme, use the optimal measurement scheme to collect quenching parameters to ensure the most accurate prediction results, you can systematically build, train and optimize the prediction model, and finally find the most suitable measurement scheme to improve the monitoring and control level of the quenching process. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 Shown is a schematic flow chart of the data-driven method for determining quenching parameters of medium and thick steel plates of the present invention. DETAILED DESCRIPTION

[0055] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art can think of other obvious variations. The basic principles of the present invention defined in the following description can be applied to other embodiments, variations, improvements, equivalents, and other technical solutions that do not deviate from the spirit and scope of the present invention.

[0056] It is to be understood that the term "one" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of an element may be one, while in another embodiment, the number of the element may be multiple, and the term "one" should not be understood as a limitation on the quantity.

[0057] refer to Figure 1The present invention provides a technical solution: a method for determining quenching parameters of medium and thick steel plates based on data-driven, the method comprising the following steps:

[0058] Step 1: Arrange multiple temperature sensors on the steel plate sample to be quenched according to the preset measurement point positions;

[0059] Step 2: Collect data recorded by multiple temperature sensors through a data acquisition device and draw a temperature change curve;

[0060] Step 3: extract multiple key characteristic values ​​from the drawn temperature change curve;

[0061] Step 4: Input multiple key feature values ​​as input variables into the prediction model, record the predicted values ​​output by the model, and evaluate the prediction results;

[0062] Step 5: Adjust the position and depth of the measurement point and store it as a new measurement plan, assign a unique identification tag to each new measurement plan, and reinstall the temperature sensor according to the new measurement plan;

[0063] Step 6: Repeat steps 2 to 5 to determine the optimal measurement scheme, and collect quenching parameters through the optimal measurement scheme.

[0064] The step 1 includes the following detailed steps:

[0065] First, multiple measurement points are determined according to the size and shape of the selected steel plate sample. The number of these measurement points should be no less than 6 to ensure that the temperature data of the steel plate can be fully and accurately obtained. The selection of these measurement points should be based on actual needs and the characteristics of the steel plate to facilitate subsequent temperature monitoring and analysis. Next, holes consistent with the number of measurement points are drilled at the determined measurement point locations. The size and depth of these holes should match the size of the temperature sensor to ensure that the sensor can be stably installed in the hole. Finally, the temperature sensor (thermocouple, etc.) is installed in the hole according to the measurement process requirements. The new measurement scheme includes the position information of the measurement point on the steel plate and the depth information inside the steel plate. The temperature sensor is installed on the steel plate for subsequent temperature monitoring and analysis. By selecting appropriate measurement points, accurately drilling holes and correctly installing sensors, accurate and reliable temperature data can be ensured, providing strong support for subsequent work.

[0066] The second step is to effectively draw the temperature change curve. The detailed steps are as follows: ensure that each temperature sensor is correctly connected to the data collector, check whether the connection is stable to ensure that data can be continuously collected during the quenching process, configure the data collector so that it can establish a connection with the cloud server through a wired or wireless network, ensure the stability of the network connection to avoid data loss, the data collector sends the temperature data to the cloud in real time, and performs data preprocessing in the cloud, including filtering out noise, outlier processing, etc., adding accurate time tags to each data point to ensure the timing of the data, and downloading the preprocessed temperature-time data set from the cloud to the local storage device.

[0067] Check the data integrity to ensure there are no missing values ​​or time label errors. Import the matplotlib plotting library and read the temperature-time data.

[0068] Use the plotting function of matplotlib to generate a temperature change curve graph with temperature data as the vertical axis and time label as the horizontal axis. Configure the chart settings, such as adding titles, axis labels, legends, etc., to improve the readability of the chart. Based on experience, you can select a temperature change curve of a representative measurement point as a representative curve and extract feature values ​​from the representative curve.

[0069] The step three includes the following steps: using Python or MATLAB to load the temperature-time data set; obtaining the highest temperature value U by finding the maximum and minimum values; max and the minimum temperature U min ; Calculate the time length t from the highest temperature moment to the lowest temperature moment, where t is the cooling time; By calculating the difference ΔU between the highest temperature and the lowest temperature, the cooling rate V = ΔU / t is obtained; The temperature-time data is constructed into a time series of temperature changes over time, and a plurality of gradient values ​​p are obtained by numerical differentiation method to form a gradient data set; Set the gradient change threshold ΔY, traverse the data set, when |p|≤ΔY, the gradient does not change significantly, when |p|>ΔY, the gradient changes significantly, and the gradient value and time point at this time are recorded.

[0070] In order to construct the temperature-time data into a time series of temperature changes over time: First, make sure that each temperature data is matched with the corresponding time point. Check the consistency and accuracy of the data to ensure that there are no missing values ​​or incorrect timestamps. Second, convert the temperature data into CSV or Pandas DataFrame format for subsequent processing. CSV file is a common text file format that can easily store and read data. Pandas DataFrame is a commonly used data structure in Python, which provides rich data processing functions. Import the required libraries in Python, such as pandas and datetime.

[0071] Then, use the Series or DataFrame class of pandas to create a time series object. Set time as the index and temperature as the value. You can use the read_csv function of pandas to read data from the CSV file and convert it into a time series object. The temperature-time data is successfully constructed into a time series of temperature changes over time, providing basic data for further analysis and application.

[0072] To build and evaluate a predictive model, the following steps can be followed:

[0073] Building a predictive model: Among them, X1, X2, X3, X4, X5 are input variables; α, β, γ, δ, ρ0, λ are regression coefficients; X0ρ0 is the intercept; is the output variable;

[0074] Train the prediction model:

[0075] Retrieve historical temperature change curves and corresponding steel plate performance data from a database;

[0076] Clean the steel plate performance data, including removing outliers, processing missing values, and possible data conversion to ensure data quality;

[0077] Extract key feature values ​​from the historical measured temperature curves, and use these extracted key feature values ​​as input variables to construct a feature set for training a prediction model;

[0078] The feature set is divided into a training set and a test set, and the model is trained or fitted by the least squares method using the feature data (input variables) in the training set and the corresponding steel plate performance data (output variables);

[0079] Model evaluation:

[0080] Set the model evaluation threshold Δy, when When , the model is valid, record Y c ; Where Y is the actual measured value, is the output value of the prediction model.

[0081] Each new measurement scheme corresponds to a new temperature change curve, new key feature values ​​are extracted from the new temperature change curve, and these new feature values ​​are used as input variables to construct a new feature set;

[0082] Using the data in the new feature set as the input variable of the prediction model can obtain the corresponding output variable

[0083] if Then record the Y at this time ic ;

[0084] where Y i is the i-th measurement value. After repeating multiple measurement schemes, multiple Y ic Values, forming a threshold set {Y ic}, where i = 1, 2, 3, ..., n, and n is the number of solutions;

[0085] In order to determine the optimal measurement solution, we sort the elements in the threshold set in ascending order. ic The measurement scheme corresponding to the value is considered to be the optimal measurement scheme. This is because the smallest value can indicate that the measurement scheme has the best performance in the prediction model. Once the optimal measurement scheme is determined, we can use it to collect quenching parameters. This means that we will use this measurement scheme to obtain parameter data related to the quenching process. These data can be used to further analyze and optimize the quenching process to improve product quality and production efficiency.

[0086] In summary, by extracting new key feature values ​​and constructing new feature sets, we can use predictive models to evaluate the performance of different measurement schemes. By repeating multiple measurement schemes and recording the output variable values, we can determine the optimal measurement scheme and use it to collect quenching parameters. This approach can help us better understand and optimize the quenching process, thereby improving product quality and production efficiency.

[0087] The embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. The embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part, and / or installed from a removable medium. When the computer program is executed by the central processing unit (CPU), the above-mentioned functions defined in the method of the present application are executed. It should be noted that the above-mentioned computer-readable medium of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium can be, for example, but not limited to, a system, device or device of an electrical, magnetic, optical, electromagnetic, infrared segment, or semiconductor, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection with one or more wire segments, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, electrical wire, optical cable, RF, etc., or any suitable combination of the foregoing.

[0088] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0089] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are only examples and do not limit the present invention. The objectives of the present invention have been fully and effectively achieved. The functional and structural principles of the present invention have been demonstrated and explained in the embodiments. Without departing from the principles, the implementation methods of the present invention may be deformed or modified in any way.

Claims

1. A data-driven method for determining quenching parameters of medium and thick steel plates, characterized in that: The method comprises the following steps: Step 1: Arrange multiple temperature sensors on the steel plate sample to be quenched according to the preset measurement point positions; Step 2: Collect data recorded by multiple temperature sensors through a data acquisition device and draw a temperature change curve; Step 3: extract multiple key characteristic values ​​from the drawn temperature change curve; Step 4: Input multiple key feature values ​​as input variables into the prediction model, record the predicted values ​​output by the model, and evaluate the prediction results; Step 5: Adjust the position and depth of the measurement point and store it as a new measurement plan, assign a unique identification tag to each new measurement plan, and reinstall the temperature sensor according to the new measurement plan; Step 6: Repeat steps 2 to 5 to determine the optimal measurement scheme, and collect quenching parameters through the optimal measurement scheme.

2. The method for determining quenching parameters of medium and thick steel plates based on data drive according to claim 1, characterized in that: The step 1 comprises the following steps: Select multiple measuring points according to the size and shape of the selected steel plate sample, and the number of measuring points shall not be less than 6; Drill holes at the measuring point locations that match the number of measuring points, and install temperature sensors in the holes according to the measuring process requirements; The new measurement scheme includes the position information of the measuring point on the steel plate and the depth information inside the steel plate.

3. The method for determining quenching parameters of medium and thick steel plates based on data drive according to claim 1, characterized in that: The step 2 comprises the following steps: The temperature sensor is connected to the data collector, and the data collector establishes a data connection with the cloud. The data collector sends the acquired temperature data to the cloud for preprocessing, and adds a time tag corresponding to the acquisition time to the preprocessed temperature data. The temperature data with the time tag constitutes a temperature-time data set; The drawing of the temperature variation curve comprises: Get the temperature-time dataset from the cloud and download it to the local storage device; Use the matplotlib drawing library to generate a time series temperature change curve graph based on the temperature-time dataset.

4. The method for determining quenching parameters of medium and thick steel plates based on data drive according to claim 1, characterized in that: The step three comprises the following steps: Use Python or MATLAB to load the temperature-time dataset; By finding the maximum and minimum values, we can obtain the highest temperature value U max and the minimum temperature U min ; Calculate the time length t from the highest temperature moment to the lowest temperature moment, t is the cooling time; By calculating the difference ΔU between the highest temperature and the lowest temperature, we get the cooling rate V = ΔU / t; The temperature-time data is constructed into a time series of temperature changes over time, and a numerical differentiation method is used to obtain multiple gradient values ​​p and form a gradient data set; Set the gradient change threshold ΔY and traverse the data set. When |p|≤ΔY, the gradient does not change significantly. When |p|>ΔY, the gradient changes significantly. Record the gradient value and time point at this time.

5. The method for determining quenching parameters of medium and thick steel plates based on data drive according to claim 3, characterized in that: The temperature-time data is constructed into a time series of temperature changes over time, comprising the following steps: Constructing the temperature-time data into a time series of temperature changes over time includes the following steps: Matching the temperature data acquired by the temperature sensor with the corresponding time point; Convert data to CSV or Pandas DataFrame format; Use the time series tool in Python to create a time series object, using time as the index and temperature as the value.

6. The method for determining quenching parameters of medium and thick steel plates based on data drive according to claim 1, characterized in that: The step 4 comprises the following steps: Building a predictive model: Among them, X1, X2, X3, X4, X5 are input variables; α, β, γ, δ, ρ0, λ are regression coefficients; X0ρ0 is the intercept; is the output variable; Train the prediction model: Retrieve historical temperature change curves and corresponding steel plate performance data from a database; Clean the steel plate performance data, including removing outliers and processing missing values ​​to ensure data quality; Extract key feature values ​​from the historical temperature curve, and use these extracted key feature values ​​as input variables to construct a feature set for training a prediction model; The feature set is divided into a training set and a test set, and the model is trained or fitted by the least squares method through the feature data in the training set - input variables and the corresponding steel plate performance data - output variables; Model evaluation: Set the model evaluation threshold Δy, when When , the model is valid, record Y c ; Where Y is the actual measured value, is the output value of the prediction model.

7. The method for determining quenching parameters of medium and thick steel plates based on data drive according to claim 6, characterized in that: Each new measurement scheme corresponds to a new temperature change curve, new key feature values ​​are extracted from the new temperature change curve, and these new feature values ​​are used as input variables to construct a new feature set; Using the data in the new feature set as the input variable of the prediction model can obtain the corresponding output variable if Then record the Y at this time ic ; where Y i is the i-th measurement value. After repeating multiple measurement schemes, multiple Y ic Values, forming a threshold set {Y ic }, where i = 1, 2, 3, ..., n, and n is the number of solutions.

8. The method for determining quenching parameters of medium and thick steel plates based on data drive according to claim 7, characterized in that: Arrange the elements in the threshold set in ascending order, and the smallest Y ic The measurement scheme corresponding to the value is the optimal measurement scheme, and the optimal measurement scheme is used to collect quenching parameters.

9. An electronic device, comprising a processor, a memory connected to the processor, and a communication module, wherein: The electronic device executes the data-driven method for determining quenching parameters of medium and thick steel plates as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement the data-driven method for determining quenching parameters of medium and thick steel plates as described in any one of claims 1 to 8.

Images