A smelting method based on a large-tonnage double-frequency induction smelting electric furnace

Abstract

The present application relates to a kind of based on large tonnage double-frequency induction melting electric furnace melting method, belong to metal melting technical field, including step one: carry out new time's smelting production;Step two: detection module detects temperature data of temperature time variation and uploads to control module, control module will the temperature data of same production batch as historical data;Step three: control module is predicted with machine learning algorithm according to historical data temperature with time variation prediction value when this time smelting production;Step four: control module judges the difference between detection value of detection module and prediction value under same time of carrying out reaches deviation threshold value and executes step five, otherwise executes step six;Step five: control module is switched to low-frequency mode according to the time of reaching switching threshold value in prediction value, induction furnace;Step six: control module judges detection value and switches to low-frequency mode when reaching switching threshold value, induction furnace;It has the characteristics of switching accurately, give consideration to smelting effect and efficiency.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of metal smelting, and particularly relates to a smelting method based on a large-tonnage double-frequency induction smelting electric furnace. BACKGROUND

[0002] An induction furnace is a common device for smelting metal. During the smelting process, a certain smelting temperature and stirring intensity are required to ensure that the metal is completely melted into a liquid and that the components in the molten metal are uniform.

[0003] Generally, the method for smelting metal by using an induction smelting electric furnace, such as the method for extracting vanadium and removing phosphorus from molten iron disclosed in Chinese Patent Application CN118957308A, comprises the following steps: heating the molten iron by using an induction furnace, adding a vanadium extraction and phosphorus removal treatment agent, adjusting the power supply parameters of the induction furnace to the heat preservation gear, using the current of the heat preservation gear to provide electromagnetic stirring force for the slag-gold reaction, removing the vanadium slag after the reaction is completed, obtaining low-phosphorus semi-steel, and transporting the semi-steel to a steelmaking converter for steelmaking. By setting the molten iron to be heated and adding a vanadium extraction and phosphorus removal treatment agent, adjusting the power supply parameters of the induction furnace to the heat preservation gear, and stirring after smelting, the quality of the semi-steel is improved, which makes it more suitable for use, and the adaptability of the product is increased.

[0004] In the above structure, although the electromagnetic stirring force is provided by the current of the heat preservation gear to perform the slag-gold reaction, the switching of the two gears is completed, and the characteristics of the induction furnace under the two frequencies are utilized. However, when the switching time is too early, the molten steel may not be completely melted, and the molten steel is switched from the medium-frequency mode with strong heating capacity to the low-frequency mode with strong stirring capacity, which may result in incomplete melting. When the switching time is too late, the molten steel may still not be stirred after being completely melted, and the switching time is largely dependent on the result of the temperature detection of the metal to be smelted in the furnace, which shows whether the temperature of the metal to be smelted reaches a threshold value. When the switching is performed after the threshold value is reached, the molten steel is switched to the low-frequency mode when it is completely melted to a certain extent. However, for high-temperature molten metal, the temperature measurement may be inaccurate, which may result in inaccurate switching time based on the temperature measurement result. Therefore, a smelting method based on a large-tonnage double-frequency induction smelting electric furnace is needed, which has accurate switching time and takes into account the smelting effect and efficiency. SUMMARY

[0005] To solve the above problems in the prior art, the application provides a smelting method based on a large-tonnage double-frequency induction smelting electric furnace, which has accurate switching time when switching between medium-frequency and low-frequency and takes into account the smelting effect and efficiency.

[0006] The purpose of the application can be achieved by the following technical solutions:

[0007] A smelting method based on a large-tonnage double-frequency induction smelting electric furnace, comprising the following steps:

[0008] Step one: carry out a new smelting production;

[0009] Step two: the detection module detects the temperature data of the temperature change over time and uploads it to the control module, and the control module archives the temperature data of the same production batch as historical data;

[0010] Step three: the control module uses machine learning algorithm to predict the predicted value of the temperature change over time in this smelting production according to the historical data, and calculates the time when the predicted value reaches the switching threshold;

[0011] Step four: the control module judges whether the difference between the detection value of the detection module and the predicted value at the same time reaches the deviation threshold, and when the judgment result is yes, step five is executed, otherwise step six is executed;

[0012] Step five: the control module switches the induction furnace from the medium frequency mode to the low frequency mode according to the time when the predicted value reaches the switching threshold, and executes step seven;

[0013] Step six: the control module judges whether the detection value reaches the switching threshold, and when the judgment result is yes, the induction furnace is switched from the medium frequency mode to the low frequency mode, and step seven is executed;

[0014] Step seven: repeat steps one to six.

[0015] As a preferred technical solution of the present application, the step three further comprises: the control module determines the algorithm parameters (p, d, q) by using ARIMA algorithm according to the time-archived power consumption data, and uses ARIMA algorithm to determine the predicted value of the temperature change over time in this smelting production.

[0016] As a preferred technical solution of the present application, the step five further comprises: the control module judges whether the mean value of the predicted value and the detection value at the same time in the predicted value reaches the switching threshold, and when the judgment result is yes, the induction furnace is switched from the medium frequency mode to the low frequency mode, and step seven is executed.

[0017] As a preferred technical solution of the present application, the step two further comprises: the detection module detects the temperature data of the temperature change over time by using a plurality of temperature sensors and uploads the temperature data group to the control module, and the control module archives the temperature data of the same production batch as historical data; the step three further comprises: the control module judges whether the temperature variance in the temperature data group uploaded for five times in succession exceeds the variance threshold, and when the judgment result is yes, the control module corrects the deviation threshold upward, and the step four further comprises: the control module judges whether the difference between the detection value of the detection module and the predicted value at the same time reaches the upwardly corrected deviation threshold, and when the judgment result is yes, step five is executed, otherwise step six is executed.

[0018] As a preferred embodiment of the present invention, step three further includes: the control module pre-inputs a variance threshold F0, the control module determines whether the temperature variance in five consecutive uploaded temperature data sets exceeds the variance threshold, and if the determination result is yes, the control module increases the deviation threshold by a factor of A1, where A1 = F / F0 × c; step four further includes: the control module determines whether the difference between the detected value of the detection module and the predicted value at the same time reaches the deviation threshold after increasing by a factor of A1, if the determination result is yes, step five is executed, otherwise step six is ​​executed.

[0019] As a preferred embodiment of the present invention, step five further includes: the control module switches the induction furnace from medium frequency mode to low frequency mode after a delay based on the time when the switching threshold is reached in the predicted value, and then executes step seven; step six further includes: the control module determines whether the detection value has reached the switching threshold, and if the determination result is yes, the induction furnace switches from medium frequency mode to low frequency mode after a delay, and then executes step seven.

[0020] As a preferred technical solution of the present invention, step one further includes: performing a new smelting production, inputting the melting point of this smelting, and the control module increasing the delay when the melting point exceeds the melting point threshold and decreasing the delay when the melting point is lower than the melting point threshold.

[0021] As a preferred embodiment of the present invention, step three further includes: after predicting the predicted value of temperature change over time during this smelting production, the control module determines the heating rate, and when the heating rate exceeds the rate threshold, the delay is corrected downwards, and when the heating rate is lower than the rate threshold, the delay is corrected upwards.

[0022] The beneficial effects of this invention are as follows:

[0023] (1) By having the control module calculate the predicted value of temperature change over time based on historical data, and when the difference between the predicted values ​​at the time corresponding to the detected value reaches the deviation threshold, there is a high probability of inaccurate detection. In this case, the detected value is stopped and the predicted value is used to guide the switching time, which reduces the impact of inaccurate temperature measurement on the switching time. At the same time, when the detection result is relatively accurate, the detected value is used to guide the switching time. Compared with the predicted value, which represents a typical situation but has a gap with the actual temperature, the detected value is closer to the actual temperature, which improves the accuracy of the switching time.

[0024] (2) By having the control module determine whether the average of the predicted value and the detected value at the same time reaches the switching threshold, the average of the predicted value and the detected value is used. Compared with a single detected value or predicted value, the average of the detected value and the predicted value can make the obtained value closer to the actual temperature value to a greater extent or with a greater probability. This makes the switching time point based on the average value closer to the time point when the metal to be melted is completely melted, thus improving the accuracy of the switching timing.

[0025] (3) By having the control module determine whether the temperature variance in the temperature data set uploaded five times in a row exceeds the variance threshold, and when the determination result is yes, the control module corrects the deviation threshold upwards, so that when the detection value is normal, the control module is likely to be misjudged as exceeding the deviation threshold, the judgment standard is relaxed and the detection value that is closer to the actual situation is used to determine the switching time. When the probability of the detection value being misjudged is low, the judgment standard is tightened to avoid the impact of the inaccuracy of the detection value on the switching time.

[0026] (4) By making the control module adjust the delay downward when the heating rate exceeds the threshold, when the heating rate is lower than the threshold, it means that the heating efficiency of the metal after entering the alternating electric field is slow and the probability of the metal being fully melted is low. It is necessary to extend the delay and adjust the delay upward to ensure the melting effect. When the metal generates a higher temperature when entering the same alternating electric field, it means that the heating rate is fast and the metal is more likely to be fully melted. At this time, the threshold is adjusted downward to ensure the operation efficiency. Attached Figure Description

[0027] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0028] Figure 1 This is a block diagram of the control loop of the present invention; Detailed Implementation

[0029] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0030] Please see Figure 1 A smelting method based on a large-tonnage dual-frequency induction melting electric furnace includes the following steps:

[0031] Step 1: Proceed to a new smelting production;

[0032] Specifically, production is carried out in batches. In each batch of production, the composition, melting point and target product of the metal to be melted are the same. When any one or more of the composition, melting point or target product changes, it is recorded as a new batch of production.

[0033] Meanwhile, the induction melting furnace includes at least a furnace body and a coil arranged around the furnace body. When an alternating current is passed through the coil, an alternating magnetic field is periodically generated in the furnace body, causing the metal block in the furnace body to generate an eddy current electric field due to the alternating magnetic field, thereby using the electrothermal effect to melt the metal. The coil is electrically connected to the control module, and the control module controls the frequency and operating power of the coil.

[0034] When a smelting production begins, the metal to be smelted is placed into the furnace, and then a medium-frequency current is passed through the coil to smelt the metal. After smelting is completed, the process is switched to passing a low-frequency current for stirring.

[0035] After step one is completed, proceed to step two;

[0036] Step 2: The detection module detects temperature data that changes over time and uploads it to the control module. The control module archives and organizes the temperature data of the same production batch as historical data.

[0037] Specifically, since it is impossible to collect enough data as historical data when a batch starts production, in this embodiment, when the temperature data of temperature change over time is detected by the detection module in the first four production cycles of a certain batch and uploaded to the control module, the control module archives and organizes the temperature data of the same production batch as historical data. When the fifth production cycle is reached, step one is executed. At this time, in step two, the temperature data of the past four production cycles are archived and organized as historical data. In the subsequent nth production cycle, in step two, the temperature data of the past n-1 production cycles are archived and organized as historical data.

[0038] After step two is completed, proceed to step three.

[0039] Step 3: The control module uses machine learning algorithms based on historical data to predict the temperature change over time during this smelting production, and calculates the time when the switching threshold is reached in the predicted value.

[0040] Specifically, the control module determines the temperature value that changes over time in the current production cycle based on historical data, and plots a graph of temperature changes over time as a prediction value. The starting point of the graph is the time point when production begins.

[0041] After step three is completed, step four is executed: the control module determines whether the difference between the detection value of the detection module and the predicted value under the same time reaches the deviation threshold. If the determination result is yes, step five is executed; otherwise, step six is ​​executed.

[0042] Specifically, the control module starts timing synchronously during each production run. Then, the control module obtains the number of seconds since the start of production every second and finds the temperature value corresponding to the current number of seconds in the prediction value chart as the current prediction value. At the same time, the detection module counts the temperature of the metal to be melted every second and uploads it to the control module as the detection value. After receiving the detection value each time, the control module determines whether the difference between the detection value and the current prediction value exceeds the deviation threshold. When the difference exceeds the deviation threshold for five consecutive times, the control module determines that the difference between the detection value and the prediction value at the same production time has reached the threshold.

[0043] Although the current predicted value cannot fully represent the actual temperature value, it can represent the typical temperature value at the current time point. When there is a large deviation between the detected value and the predicted value, it means that there is a high probability that the detection is inaccurate. At this time, it is necessary to stop accepting the detected value in this production and use the predicted value to guide the switching time. At this time, proceed to step five.

[0044] Step 5: Based on the predicted time of reaching the switching threshold, the control module switches the induction furnace from medium-frequency mode to low-frequency mode, and then executes Step 7;

[0045] At this point, the control module stops accepting the detection value during this production. The control module obtains the number of seconds corresponding to the point when the temperature value reaches the switching threshold from the prediction value chart. Then, the control module displays the time when this number of seconds is reached, which means that the metal in the furnace has a high probability of reaching the target temperature and completely melting. The control module instructs the induction furnace to switch from medium frequency mode to low frequency mode, completes the cessation of accepting the detection value, and uses the prediction value to guide the switching timing.

[0046] During the above process, if the deviation threshold is not exceeded for five consecutive judgments, the control module continues to receive the detection values ​​uploaded by the detection module. When the detection value reaches the switching threshold, it means that the detection module has detected that the metal in the furnace has reached the target temperature and is completely melted, and the detection result is reliable. At this time, the control module instructs the induction furnace to switch from medium frequency mode to low frequency mode, thus completing the switching timing guided by the detection value.

[0047] After completing step five or six, proceed to step seven, which is to return to step one and start a new production process.

[0048] By having the control module calculate predicted temperature changes over time based on historical data, and stopping the adoption of the detected value when the difference between the detected value and the predicted value at the same time reaches a deviation threshold, indicating a high probability of inaccurate detection, the predicted value is used to guide the switching timing. This reduces the impact of inaccurate temperature measurement on the switching timing. At the same time, when the detection results are relatively accurate, using the detected value to guide the switching timing is more accurate than the predicted value, which represents a typical situation but differs from the actual temperature. The detected value is closer to the actual temperature, thus improving the accuracy of the switching timing.

[0049] In step three above, the control module uses a machine learning algorithm based on historical data to predict the temperature change over time during this smelting production, and calculates the time when the switching threshold is reached in the predicted value. Specifically, the control module uses the ARIMA algorithm to determine the algorithm parameters (p,d,q) based on the power consumption data archived by time, and uses the ARIMA algorithm to predict the temperature change over time during this smelting production.

[0050] Specifically, it includes the following steps:

[0051] The control module receives historical data and preprocesses it. First, it concatenates the data on temperature changes over time from several past production runs in sequence. Then, it uses a unit root test to check the stationarity of the data. In this embodiment, the test method is the ADF test.

[0052] The control module then determines whether the historical data is stable. If the result is negative, it performs a difference operation on the historical data to achieve stability. First, it performs a difference operation on the original data to determine whether it is stable. If the result is negative, it repeats the difference operation until the data reaches a stable state and records the number of differences d required.

[0053] Subsequently, the control module calculates the autocorrelation function (ACF) and partial autocorrelation function (PACF) based on different p and q values, analyzes the ACF and PACF plots under different p and q values, and selects the combination of minimum information criteria (such as AIC or BIC), autoregression order p, and moving average order q.

[0054] At this point, the control module determines the algorithm parameters (p,d,q), establishes the ARIMA(p,d,q) model, and then uses the established ARIMA model to obtain historical data and the temperature trend over time. It then extracts the temperature trend over time from the start of production until the temperature reaches the switching threshold as the predicted value.

[0055] When a large deviation occurs between the detected value and the predicted value in the above process, although the predicted value, which represents a typical case, can be closer to the actual value with a higher probability, the detected value also represents the actual temperature. In this case, it is necessary to use both the detected value and the predicted value and calculate the mean of the two data.

[0056] By having the control module determine whether the average of the predicted value and the detected value at the corresponding time reaches the switching threshold, the average of the predicted value and the detected value is used. Compared with a single detected value or predicted value, the average of the detected value and the predicted value can make the obtained value closer to the actual temperature value to a greater extent or with a greater probability. As a result, the switching time point based on the average value is closer to the time point when the metal to be melted is completely melted, thus improving the accuracy of the switching timing.

[0057] Since a single temperature sensor may detect temperature inaccurately, step two also includes: several temperature sensors in the detection module detect temperature data that changes over time and upload the temperature data set to the control module. The control module archives and organizes the temperature data of the same production batch as historical data.

[0058] Specifically, several temperature sensors synchronously upload temperature data once per second. After receiving the temperature data uploaded by the several temperature sensors, the control module takes the average of the temperature data as the detection value.

[0059] In the above process, when the variance of several temperature data received by the detection module is large, it means that the temperature of several parts of this batch of production varies greatly under the influence of random factors. At this time, even if there is no case of inaccurate temperature detection, the deviation between the detected value and the predicted value will exceed the deviation threshold. In this case, the control module will stop accepting the detected value that is more representative of the actual situation and use the predicted value instead, even if there is no case of inaccurate temperature detection. Therefore, when the variance of several temperature data received by the detection module is large, it is necessary to relax the judgment criteria for whether the temperature detection is inaccurate.

[0060] Therefore, step three also includes: the control module determines whether the temperature variance in the temperature data group uploaded five times in a row exceeds the variance threshold, and if the determination result is yes, the control module corrects the deviation threshold upward. Step four also includes: the control module determines whether the difference between the detection value of the detection module and the predicted value under the same time reaches the deviation threshold after upward correction. If the determination result is yes, step five is executed; otherwise, step six is ​​executed.

[0061] Specifically, step three also includes: the control module pre-inputs a variance threshold F0, the control module determines whether the temperature variance in five consecutive uploaded temperature data sets exceeds the variance threshold, and if the determination result is yes, the control module increases the deviation threshold by a factor of A1, where A1 = F / F0 × c; step four also includes: the control module determines whether the difference between the detected value of the detection module and the predicted value at the same time reaches the deviation threshold after increasing by a factor of A1, if the determination result is yes, step five is executed, otherwise step six is ​​executed;

[0062] When the variance F is large, it means that there is a high probability that the temperature detection is inaccurate and the deviation between the detected value and the predicted value still exceeds the deviation threshold. It is necessary to relax the judgment criteria for whether the detection is inaccurate. At this time, the value of A1 = F / F0 × c is large, and the control module increases the deviation threshold by A1 times. Under the same conditions, it is more difficult for the difference between the detected value and the predicted value to be judged as exceeding the deviation threshold. This completes the relaxation of the judgment criteria when the difference between the detected value and the predicted value is easily misjudged as exceeding the deviation threshold under normal circumstances.

[0063] When the variance F is small, it means that the deviation between the detected value and the predicted value still exceeds the deviation threshold. This indicates that the temperature detection is inaccurate and the judgment standard for whether the temperature detection is inaccurate needs to be narrowed. At this time, the value of A1 = F / F0 × c is small, and the control module increases the deviation threshold by A1 times. Under the same conditions, the difference between the detected value and the predicted value is more likely to be judged as exceeding the deviation threshold. This tightens the judgment standard when the difference between the detected value and the predicted value is easily misjudged as exceeding the deviation threshold, and when the detected value is misjudged as exceeding the deviation threshold under normal conditions.

[0064] By having the control module determine whether the temperature variance in five consecutive uploaded temperature data sets exceeds a variance threshold, and if the determination result is yes, the control module adjusts the deviation threshold upwards. This allows the control module to relax the judgment criteria when the detected value is likely to be misjudged as exceeding the deviation threshold under normal conditions, and to use a detected value that is closer to the actual situation to determine the switching timing. When the probability of the detected value being misjudged is low, the judgment criteria are tightened to avoid the impact of inaccurate detected values ​​on the switching timing.

[0065] In the above process, when the detected or predicted temperature reaches the threshold, there is still a probability that there are unmelted steel blocks in some parts of the furnace. At this time, it is necessary to delay for a certain period of time until the steel is completely melted before switching to low frequency mode. Therefore, step five also includes: the control module switches the induction furnace from medium frequency mode to low frequency mode after a delay based on the time when the predicted value reaches the switching threshold, and then executes step seven; step six also includes: the control module determines whether the detected value has reached the switching threshold. If the determination result is yes, the induction furnace switches from medium frequency mode to low frequency mode after a delay, and then executes step seven.

[0066] Setting a delay allows the equipment to remain in medium-frequency mode for a longer period of time to melt the steel block, reducing the probability of incomplete melting and improving the melting effect and the accuracy of switching timing.

[0067] The required delay varies depending on the situation. For example, when melting metals with high melting points, there is a greater probability of unmelted metal when the detected or predicted temperature reaches the threshold. In this case, the delay needs to be extended so that the equipment can remain in medium-frequency mode for a longer period of time to melt the steel block and ensure the melting effect. When the metal has a low melting point, there is a lower probability or degree of unmelted metal when the detected or predicted temperature reaches the threshold. In this case, there is no need to extend the delay, and the equipment should be switched to low-frequency mode as soon as possible to improve efficiency.

[0068] Therefore, step one also includes: conducting a new smelting production, inputting the melting point of this smelting, and the control module increasing the delay when the melting point exceeds the threshold and decreasing the delay when the melting point is below the threshold;

[0069] By setting the control module, the delay is increased when the melting point exceeds the threshold and decreased when the melting point is below the threshold. This ensures sufficient melting when smelting metals with higher melting points and improves work efficiency when smelting metals with lower melting points.

[0070] In use, different metals have different effects when smelted in an induction furnace. Some metals generate higher temperatures and faster heating rates when the same alternating electric field is applied, which means that the metal is more likely to be fully melted. In this case, the delay needs to be corrected downward and the low-frequency mode should be switched as soon as possible. When the metal heating rate is slow, it means that the probability of the metal being fully melted is low. In this case, the delay needs to be extended to ensure the melting effect. Therefore, step three also includes: after predicting the predicted value of temperature change over time during this melting production, the control module judges the heating rate. When the heating rate exceeds the threshold, the delay is corrected downward; when the heating rate is below the threshold, the delay is corrected upward. Then, in steps five and six, the upward or downward correction delay is adopted.

[0071] By adjusting the delay downwards when the heating rate exceeds a threshold, and adjusting the delay upwards when the heating rate is below the threshold (meaning the metal is heating slowly after being introduced into the alternating electric field, and the probability of the metal being fully melted is low), the control module can ensure the melting effect by extending the delay. Conversely, when the metal reaches a higher temperature under the same alternating electric field, it indicates a faster heating rate, and the metal is more likely to be fully melted. In this case, the threshold can be adjusted downwards to ensure operational efficiency.

[0072] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A smelting method based on a large-tonnage dual-frequency induction melting electric furnace, characterized in that: Includes the following steps Step 1: Proceed to a new smelting production; Step 2: The detection module detects temperature data that changes over time and uploads it to the control module. The control module archives and organizes the temperature data of the same production batch as historical data. Step 3: The control module uses machine learning algorithms based on historical data to predict the temperature change over time during this smelting production, and calculates the time when the switching threshold is reached in the predicted value. Step 4: The control module determines whether the difference between the detected value and the predicted value at the same time reaches the deviation threshold. If the determination result is yes, proceed to step 5; otherwise, proceed to step 6. Step 5: Based on the predicted time of reaching the switching threshold, the control module switches the induction furnace from medium-frequency mode to low-frequency mode, and then executes Step 7; Step Six: The control module determines whether the detection value has reached the switching threshold. If the determination result is yes, the induction furnace is switched from medium frequency mode to low frequency mode, and Step Seven is executed. Step 7: Repeat steps 1 through 6; Step three also includes: the control module uses the ARIMA algorithm to determine the algorithm parameters (p, d, q) based on the power consumption data archived by time, and uses the ARIMA algorithm to calculate the predicted value of temperature change over time during this melting production. Step five further includes: the control module determines whether the average of the predicted value and the detected value at the same time reaches the switching threshold. If the determination result is yes, the induction furnace is switched from medium frequency mode to low frequency mode, and step seven is executed.

2. The smelting method based on a large-tonnage dual-frequency induction melting furnace according to claim 1, characterized in that: Step two further includes: several temperature sensors in the detection module detect temperature data that changes over time and upload the temperature data sets to the control module; the control module archives and organizes the temperature data of the same production batch as historical data; Step three further includes: the control module determines whether the temperature variance in five consecutive uploaded temperature data sets exceeds the variance threshold, and if the determination result is yes, the control module corrects the deviation threshold upwards; Step four further includes: the control module determines whether the difference between the detected value of the detection module and the predicted value at the same time reaches the upwardly corrected deviation threshold; if the determination result is yes, Step five is executed, otherwise Step six is ​​executed.

3. The smelting method based on a large-tonnage dual-frequency induction melting furnace according to claim 1, characterized in that: Step five further includes: the control module switches the induction furnace from medium frequency mode to low frequency mode after a delay based on the time when the switching threshold is reached in the predicted value, and then executes step seven; Step six further includes: the control module determines whether the detection value has reached the switching threshold, and if the determination result is yes, the induction furnace switches from medium frequency mode to low frequency mode after a delay, and then executes step seven.

4. The smelting method based on a large-tonnage dual-frequency induction melting furnace according to claim 3, characterized in that: Step one also includes: conducting a new smelting production, inputting the melting point of this smelting, and the control module increasing the delay when the melting point exceeds the melting point threshold and decreasing the delay when the melting point is below the melting point threshold.

5. A smelting method based on a large-tonnage dual-frequency induction melting furnace according to claim 4, characterized in that: Step three also includes: after predicting the temperature change over time during this smelting production, the control module determines the heating rate, and if the heating rate exceeds the rate threshold, the delay is corrected downwards, and if the heating rate is below the rate threshold, the delay is corrected upwards.

Images