Temperature change monitoring system and method in metal raw material melting process

Abstract

The invention relates to the field of temperature measurement, in particular to a temperature change monitoring system and method in a metal raw material melting process, and the temperature change trend is captured by collecting temperature data in a melting furnace in real time and combining a temperature change rate calculation formula. A DBSCAN clustering algorithm is adopted to carry out segmentation processing on temperature data to obtain different temperature change stages, curve fitting is carried out through a least square method, and error factors at all moments are calculated. And finally correcting temperature data based on the error factor to ensure that the measured value is closer to the actual temperature so as to judge whether the smelting furnace reaches the target temperature and avoid the influence of temperature error on the melting effect. The high-precision sensor, the clustering algorithm and error correction are adopted, the temperature change trend is dynamically captured, precise monitoring of the temperature in the metal melting process is achieved, and the melting uniformity and the product quality are ensured.

Description

Technical Field

[0001] The invention relates to the field of temperature measurement, and in particular to a system and method for monitoring temperature changes in a metal raw material melting process. Background Art

[0002] In the metal 3D printing and high-precision metal processing industries, the quality of metal powder directly determines the performance and quality of the final product. The manufacturing process of metal powder requires heating the metal raw materials to a molten state and then using atomization technology to spray them into particles. Therefore, temperature control during the melting process is crucial to the morphology and purity of the metal particles. During the melting of metal raw materials, temperature changes are complex and diverse. The gradual rise in temperature includes multiple stages, and the temperature change rate in each stage is different. Especially when approaching the melting point, the temperature fluctuation is small, but it has a significant impact on the melting effect of the metal. However, the measurement accuracy and response speed of traditional temperature sensors are limited, and they cannot accurately capture subtle changes in temperature in complex environments. In addition, the temperature data is affected by environmental factors such as equipment vibration and electromagnetic interference, and may deviate from the actual temperature, resulting in incomplete or excessive melting of the metal, affecting product quality.

[0003] Some existing temperature monitoring methods, for example, the Chinese patent application document with publication number CN116817603A discloses a high-temperature melting furnace molten pool temperature monitoring and inversion method based on the inverse problem of heat conduction, which constructs a non-steady-state heat conduction forward problem model along the thickness direction of the furnace wall, presets the parameters of the non-steady-state heat conduction forward problem model and the initial molten pool temperature, and solves the temperature of the outer wall of the high thermal conductivity lining, so as to monitor the melting condition in the furnace and avoid abnormal operation affecting product quality in advance. However, this method relies on complex model calculations, and it is difficult to quickly and accurately capture the temperature changes in the melting process, especially when facing external interference, it is easy to have large errors. Summary of the invention

[0004] In response to the above-mentioned problem that large errors are prone to occur when facing external interference, in the first aspect, the present invention proposes a method for monitoring temperature changes in the melting process of metal raw materials, including: collecting all temperatures from the start of furnace heating to the current collection time within a set time interval; taking the melting point of the metal raw material as the preset target temperature; obtaining the change rate of each temperature to obtain a change rate sequence, clustering the change rate sequence to obtain multiple clustering clusters; obtaining the temperature set corresponding to each clustering cluster, and performing curve fitting on each temperature set; calculating the error factor of each temperature, and the error factor is positively correlated with the difference between the temperature and the fitting temperature value; in response to the error factor of the current temperature being greater than or less than a threshold, taking the fitting temperature value corresponding to the current temperature as the actual temperature; in response to the error factor of the current temperature being equal to a threshold, taking the current temperature as the actual temperature; the threshold is the average of the error factors of all temperatures in the temperature set except the current temperature; when the actual temperature reaches the preset target temperature, warning the staff to perform subsequent processing operations.

[0005] The present invention collects temperature data by setting time intervals and analyzes the temperature change rate, thereby realizing accurate monitoring of the metal melting process. In this process, cluster analysis is performed using the change rate sequence, and curve fitting is performed on the temperature data, which greatly improves the sensitivity and accuracy of temperature changes. The error factor calculation method further corrects the real-time temperature data, and uses the fitted temperature value as the actual temperature to ensure monitoring accuracy. When the actual temperature reaches the preset target temperature, the system will automatically issue an early warning to remind the staff to perform subsequent operations. Compared with the prior art, this automated and intelligent temperature monitoring mechanism significantly improves the accuracy of abnormal detection and response, and improves the quality control level of the metal melting process.

[0006] Furthermore, before performing curve fitting on each temperature set, the method further includes: obtaining the comprehensive similarity between each temperature and the temperature corresponding to each cluster center, and merging each temperature into the cluster where the cluster center with the largest comprehensive similarity is located; the calculation method of the comprehensive similarity is specifically:

[0007] in Indicates temperature With The temperature corresponding to the cluster center The comprehensive similarity of Indicates temperature The rate of change of Indicates temperature The rate of change of Indicates temperature The corresponding collection time; Indicates temperature The collection time; is the normalization function.

[0008] Cluster analysis is performed before curve fitting. By calculating the comprehensive similarity between the temperature and the cluster center at each acquisition time, accurate classification of temperature data is achieved. The temperature data is merged into the cluster where the cluster center with the highest similarity is located, which helps to improve the effect of curve fitting. This method dynamically adjusts the clustering results, enhances the flexibility and adaptability of data analysis, and enables the monitoring system to respond to temperature changes in real time. Compared with traditional static clustering methods, the present invention provides higher accuracy and flexibility, making data processing more intelligent and efficient.

[0009] Furthermore, the error factor acquisition method is:

[0010] in Indicates The error factor of each temperature; Indicates Temperature; Indicates The fitted temperature value for a temperature.

[0011] Furthermore, the error factor acquisition method is:

[0012] in Indicates The error factor of each temperature; Indicates Temperature; Indicates The fitting temperature value of each temperature; Indicates the temperature concentration except the The rest of the temperature Temperature; Indicates the temperature concentration except the Except the temperature The fitting temperature value corresponding to each temperature; Indicates the temperature concentration except the The total number of temperatures other than the temperature; Represents the tuning factor.

[0013] The present invention provides a method for accurately measuring the error of temperature data through an error factor, wherein the error factor is related to the difference between the temperature and the fitted temperature value, thereby ensuring accurate correction of the temperature data. By calculating the mean of the error factor at other times except the current acquisition time, the accuracy of the current temperature data can be dynamically adjusted. This method effectively reduces the impact of inaccurate temperature data, ensures that the fitted temperature value is closer to the actual situation, significantly improves the reliability and data accuracy of the monitoring system, and has a stronger error correction capability than traditional methods.

[0014] Furthermore, the method for obtaining the change rate is:

[0015] in Indicates The rate of change of temperature; Indicates Temperature; Indicates Temperature; Indicates the preset target temperature; Represents the tuning factor.

[0016] The method for obtaining the temperature change rate improves the ability to capture temperature change trends by optimizing the calculation formula. This method takes into account the difference between the current temperature and the target temperature, as well as the temperature changes at previous and subsequent time points, and enhances the system's ability to respond to rapid temperature changes. By accurately calculating the temperature change rate at each acquisition moment, the system can more effectively predict temperature trends, providing higher sensitivity and accuracy than traditional linear prediction methods, helping to detect abnormal temperature changes in a timely manner.

[0017] Furthermore, clustering the change rate sequence to obtain a clustering result also includes: clustering using a DBSCAN clustering algorithm based on the acquisition time and temperature in the change rate sequence to obtain a clustering result.

[0018] The DBSCAN algorithm is used to cluster the change rate sequence, and the automated clustering process is realized by analyzing the acquisition time and temperature data. This method can effectively identify patterns and anomalies in the data set and improve the accuracy of data analysis. Compared with the traditional K-means-based clustering algorithm, DBSCAN does not require a preset number of clusters, can better adapt to the natural distribution of data, improves the intelligence and adaptability of clustering, and makes the monitoring system more sensitive to abnormal changes.

[0019] Furthermore, performing curve fitting on each temperature set also includes: performing curve fitting on the temperatures included in each temperature set using a least square method.

[0020] In a second aspect, the present invention provides a system for monitoring temperature changes during the melting process of metal raw materials, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the method for monitoring temperature changes during the melting process of metal raw materials of the present invention is implemented.

[0021] The technical effects of the present invention are: The present invention refines the temperature change stages during the melting process of metal raw materials by introducing the DBSCAN clustering algorithm and the least squares method. Compared with the traditional exponential smoothing algorithm, the new system can dynamically capture the complex characteristics of temperature changes, especially when the temperature is close to the melting point, which improves the accuracy of the measurement data. In addition, by calculating the error factor and correcting the real-time data, the system can effectively avoid the uneven melting caused by temperature deviation, ensure the higher production quality of metal powder, and is suitable for the field of high-precision metal processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein: Figure 1 is a flow chart schematically showing a method for monitoring temperature changes during the melting of a metal raw material in an embodiment of the present invention; Figure 2 is a schematic diagram schematically showing the grayscale of a metal melting furnace in an embodiment of the present invention; Figure 3 It is a block diagram schematically showing the structure of a system for monitoring temperature changes during the melting process of metal raw materials in an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0024] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0025] Example of a method for monitoring temperature changes during the melting of metal raw materials: like Figure 1 As shown, the temperature change monitoring method of the metal raw material melting process of the present invention comprises: S1. Collect furnace temperature data in real time and set a preset target temperature.

[0026] In the high-temperature environment where metal raw materials are melted, accurate temperature monitoring is the key to ensuring the quality of melting. In order to capture the temperature changes throughout the process, in this embodiment, high-precision temperature sensors and data acquisition equipment are configured to track the temperature dynamics in the furnace in real time. Figure 2 As shown. The tungsten-rhenium KW-602 metal melting temperature test instrument equipped with a thermocouple is hung on the inner wall of the furnace. Its temperature measurement range is 0~2000℃, which can meet the temperature requirements of most furnaces. In addition, the accuracy of the test instrument is 0.01. When the metal raw material begins to melt, the temperature data of the furnace is obtained using a fixed acquisition frequency. In this embodiment, the acquisition frequency can be set to 10s. The implementer can also set the acquisition frequency to 20s, 30s or 40s according to actual conditions.

[0027] The melting point temperature of the metal raw material used for melting in the embodiment in the furnace is obtained, and it is set as the preset target temperature of the furnace. For example, there is a 3D printing metal raw material which is titanium alloy and has a melting point of 1604°C. When melting this raw material, the preset target temperature of the furnace is set to 1604°C.

[0028] S2. Obtaining a temperature change rate based on changes in adjacent temperature data and a difference between a current temperature and a preset target temperature.

[0029] In the process of heating the metal raw material from solid to liquid, the temperature change in the furnace will go through multiple stages. In the melting process of this embodiment, it can be divided into four stages: initial heating, rapid heating, approaching the melting point, and reaching the melting point. The rate of temperature change in each stage is different. When the furnace is just starting to heat up or is about to reach the melting point, the rate of temperature change may be relatively slow. If only the difference calculation is used, the real-time temperature change may not be accurately captured. Therefore, the concept of relative change is introduced in this embodiment, and the rate of change of temperature data is analyzed by combining the difference between the current temperature data and the target temperature. The following examples illustrate, specifically: For the current The specific calculation formula for the temperature data at the moment and its change rate is as follows: ; in Indicates The rate of change of temperature at a given moment; Indicates The temperature of the moment; Indicates The temperature of the moment; Indicates the preset target temperature; Represents the parameter adjustment factor. Since the accuracy of the temperature acquisition device in this embodiment is 0.01, it can be set to an empirical value of 0.001 here.

[0030] when The larger the value, the Time has come The more the temperature rises, the The rate of change of temperature at a moment The bigger; when The smaller the time, the The closer the temperature at the moment is to the preset target temperature, the greater the impact of the temperature change on the metal melting process during the period of time closer to the preset target temperature. The rate of change of temperature at a moment The bigger.

[0031] The method for obtaining the temperature change rate improves the ability to capture temperature change trends by optimizing the calculation formula. This method takes into account the difference between the current temperature and the target temperature, as well as the temperature changes at previous and subsequent time points, and enhances the system's ability to respond to rapid temperature changes. By accurately calculating the temperature change rate at each acquisition moment, the system can more effectively predict temperature trends, providing higher sensitivity and accuracy than traditional linear prediction methods, helping to detect abnormal temperature changes in a timely manner.

[0032] S3. Obtaining the comprehensive similarity between the currently collected temperature data and the cluster center; and completing secondary clustering of the temperature data based on the comprehensive similarity.

[0033] Although the temperature change rate is obtained in step S2, which can reflect the temperature change trend, since the temperature change in the furnace is not linear during the metal melting process, but there are different gradient change stages, the temperature data can be processed in segments in this embodiment to better reflect the change trend of the furnace temperature.

[0034] In one embodiment, temperature data is collected from the beginning of metal melting. Since the initial stage of the furnace operation is a rapid heating stage, the temperature is far from the preset target temperature, and the amount of temperature data collected in the initial stage is small, which is not enough for stage-by-stage analysis. Therefore, in this embodiment, 60 temperature data can be collected for 10 minutes as the initial temperature sequence. ; Then the initial temperature series can be The change rate of each temperature data in the initial temperature sequence is clustered. In this embodiment, the DBSCAN clustering algorithm can be used. The input of the algorithm includes the collection time and change rate of each temperature data in the initial temperature sequence, and also includes the neighborhood radius. And the minimum number of points In this embodiment, the empirical value can be taken , The output of the algorithm includes cluster labels, data point indexes, and the total number of clusters; finally, the center point of each cluster is obtained and the The temperature data corresponding to the center point of each cluster is recorded as ,Will The corresponding collection time is recorded as , and put The corresponding rate of change is recorded as The above DBSCAN clustering algorithm is a well-known technology and will not be described in detail here.

[0035] The DBSCAN algorithm is used to cluster the change rate sequence, and the automated clustering process is realized by analyzing the acquisition time and temperature data. This method can effectively identify patterns and anomalies in the data set and improve the accuracy of data analysis. Compared with the traditional K-means-based clustering algorithm, DBSCAN does not require a preset number of clusters, can better adapt to the natural distribution of data, improves the intelligence and adaptability of clustering, and makes the monitoring system more sensitive to abnormal changes.

[0036] In the above steps, the initial temperature sequence The data is clustered. At the time of collection, the temperature data Merge into the initial temperature sequence as the last element to get the real-time temperature sequence For real-time temperature series First, clustering is performed using the same clustering method as the initial temperature series, and then the real-time temperature series is calculated Any temperature data The comprehensive similarity of the temperature data corresponding to the center point of each cluster in the clustering result is calculated as follows: ; in Indicates The temperature data at the time of collection is The comprehensive similarity of the temperature data corresponding to the cluster centers; Indicates Temperature at the time of collection The rate of change of Indicates The temperature corresponding to the cluster center The rate of change of Indicates The collection time point corresponding to the collection moment. It should be noted that in this embodiment, all collection time points are stored in the system in Datetime format to facilitate calculation; Indicates The cluster centers correspond to the temperature data The collection time point; is the normalization function.

[0037] when The smaller the time, the more The temperature at the time of collection is The smaller the difference between the temperature change rates corresponding to the cluster centers, the greater the comprehensive similarity between the two, and the closer the change trends are. The bigger; when The smaller the time, the more Collection time and The closer the collection time points of the temperatures corresponding to the cluster centers are, the greater the comprehensive similarity between the temperature data corresponding to the two collection times. The bigger.

[0038] Based on the comprehensive similarity, all temperature data before the current collection time are classified to complete the secondary clustering. For example, there is a temperature data , and the real-time temperature sequence composed of all current temperature data has 3 clusters. The comprehensive similarity between them and the temperature data corresponding to each cluster center is calculated by the above comprehensive similarity calculation method, which are respectively , as well as . Indicates the comprehensive similarity between the temperature data at the fifth collection moment and the temperature data corresponding to the first cluster center; Indicates the comprehensive similarity between the temperature data at the fifth collection moment and the temperature data corresponding to the second cluster center; It represents the comprehensive similarity between the temperature data at the fifth collection time and the temperature data corresponding to the third cluster center. And , then the temperature data Classify it into the cluster where the first cluster center is located to complete the secondary clustering of temperature data.

[0039] By performing secondary clustering on all temperature data before the current collection time, the division of the metal melting temperature change stage is completed, that is, each cluster after secondary clustering corresponds to a temperature change stage. Exemplary explanation: There is a temperature data, which belongs to the second cluster after secondary clustering, then it means that the collection time corresponding to the temperature data is the rapid heating stage.

[0040] Cluster analysis is performed before curve fitting. By calculating the comprehensive similarity between the temperature and the cluster center at each acquisition time, accurate classification of temperature data is achieved. The temperature data is merged into the cluster where the cluster center with the highest similarity is located, which helps to improve the effect of curve fitting. This method dynamically adjusts the clustering results, enhances the flexibility and adaptability of data analysis, and enables the monitoring system to respond to temperature changes in real time. Compared with traditional static clustering methods, the present invention provides higher accuracy and flexibility, making data processing more intelligent and efficient.

[0041] S4. Obtain the error factor at the acquisition time.

[0042] In step S3, the temperature change stage of each temperature data is obtained. In each temperature change stage, although the change trend of the temperature data is similar, there may be vibrations generated when other equipment is used, electromagnetic interference generated by electric heating elements, etc., which may cause the obtained temperature data to be inaccurate, thereby causing some temperature data measured by the instrument to deviate from the actual temperature. Therefore, in this embodiment, the data in each temperature change stage can be curve fitted using the least squares method, and the residual between each data point and the fitting curve can be obtained to obtain the error factor at the time of acquisition. The above least squares method belongs to a well-known technology and will not be described in detail here.

[0043] In one embodiment, the real-time acquisition Temperature data at the time of collection , if it is clustered twice in step S3, it belongs to the clusters, and for the After curve fitting for clusters, we get The fitting temperature value corresponding to the acquisition time is , then the calculation method of the error factor at the acquisition time is: ; in Indicates The first Error factor at the time of acquisition; Indicates Temperature data at the time of collection; Indicates its The first The fitting temperature value corresponding to the acquisition time; Indicates In all clusters except Other than the collection time Temperature data at the time of collection; Indicates its In all clusters except Other than the collection time The fitting temperature value at the time of acquisition; Indicates that the final clustering clusters except The number of all other collection moments other than the moment; It represents the parameter adjustment factor, and in this embodiment, the empirical value 0.001 can be taken.

[0044] when The larger the value, the The first The greater the difference between the temperature data at the time of collection and the fitted temperature value at the time of collection, the more likely the actual temperature data collected at that time will have errors, and the error factor The bigger; when The smaller the time, the In all clusters except The smaller the discreteness of the temperature data corresponding to the other acquisition times and the fitting temperature value, the smaller the discreteness of the temperature data corresponding to the other acquisition times and the fitting temperature value. If the temperature data at the time of collection deviates more from the temperature data at other times of collection, the error factor The bigger.

[0045] In another embodiment, The calculation method of the error factor at the acquisition time is: ; in Indicates The first Error factor at the time of acquisition; Indicates Temperature data at the time of collection; Indicates its The first The fitted temperature value at the time of acquisition.

[0046] The present invention provides a method for accurately measuring the error of temperature data through an error factor, wherein the error factor is related to the difference between the temperature and the fitted temperature value, thereby ensuring accurate correction of the temperature data. By calculating the mean of the error factor at other times except the current acquisition time, the accuracy of the current temperature data can be dynamically adjusted. This method effectively reduces the impact of inaccurate temperature data, ensures that the fitted temperature value is closer to the actual situation, significantly improves the reliability and data accuracy of the monitoring system, and has a stronger error correction capability than traditional methods.

[0047] S5. Correct the temperature data based on the error factor and determine whether the furnace temperature reaches a preset target temperature according to the corrected temperature data.

[0048] For the temperature data collected in real time, if the error factor corresponding to the collection time is large, it means that the data collected at this time is likely to deviate from the actual temperature. In this embodiment, the temperature data collected in real time is corrected based on the error factor. The specific correction process is as follows: First, the secondary clustering results of the real-time collected temperature data are obtained; then, the temperature change stage of the real-time collected temperature data is obtained according to the secondary clustering results; the error factor of the real-time collected temperature data in its temperature change stage is further calculated; finally, the difference between the error factor and the error factors of the remaining temperature data in its change stage is evaluated, and based on the difference, it is determined whether to use the temperature data fitting value to replace the real-time collected temperature data.

[0049] In one embodiment, there is a temperature data belonging to the third temperature change stage The mean error factor of all temperature data in the temperature change stage is , the standard deviation is , and the temperature data The error factor Greater than or less than , then it is considered that the temperature data deviates from the actual temperature data, and the fitting temperature value obtained by step S4 is used as the current real-time collected temperature data, otherwise the original temperature data is kept unchanged and used as the actual temperature data.

[0050] In another embodiment, there is a temperature data belonging to the third temperature change stage. , the mean value of all temperature data error factors during the temperature change stage is recorded as When the temperature data The error factor Greater than or less than When the temperature data deviates from the actual temperature data, the fitting temperature value obtained in step S4 is used as the current real-time collected temperature data; otherwise, the original temperature data is kept unchanged and used as the actual temperature data.

[0051] After each temperature data correction, it is compared with the preset target temperature. If the corrected temperature data is greater than or equal to the preset target temperature, it means that the melting point of the metal raw material has been reached. The system obtains an early warning sound through early warning information to prompt the operator to perform subsequent processing operations such as pouring the furnace or refining.

[0052] The present invention collects temperature data at set time intervals and analyzes the temperature change rate, thereby realizing accurate monitoring of the metal melting process. In this process, cluster analysis is performed using the change rate sequence, and curve fitting is performed on the temperature data, which greatly improves the sensitivity and accuracy of temperature changes. The error factor calculation method further corrects the real-time temperature data, and uses the fitted temperature value as the actual temperature to ensure monitoring accuracy. When the actual temperature reaches the preset target temperature, the system will automatically issue an early warning to remind the staff to perform subsequent operations. Compared with the prior art, this automated and intelligent temperature monitoring mechanism significantly improves the accuracy of abnormal detection and response, reduces the need for manual intervention, and improves the quality control level of the metal melting process.

[0053] Example of temperature change monitoring system during melting of metal raw materials: On the other hand, the present invention also provides a system for monitoring temperature changes during the melting process of metal raw materials. Figure 3 As shown, a temperature change monitoring system during the melting process of metal raw materials includes a processor and a memory, wherein the memory stores computer program instructions. When the computer program instructions are executed by the processor, a temperature change monitoring method during the melting process of metal raw materials according to the first aspect of the present invention is implemented.

[0054] The temperature change monitoring system during the melting of metal raw materials also includes other components familiar to those skilled in the art, such as a communication interface, and the configuration and functions thereof are known in the art, so they will not be described in detail here.

[0055] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented using computer-readable / executable instructions that may be stored or otherwise maintained by such a computer-readable medium.

[0056] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0057] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A method for monitoring temperature changes during the melting process of a metal raw material, characterized in that: The method comprises: Collect all temperatures from the start of furnace heating to the current collection time within a set time interval; use the melting point of the metal raw material as the preset target temperature; Obtaining the change rate of each temperature to obtain a change rate sequence, clustering the change rate sequence to obtain a plurality of clusters; obtaining a temperature set corresponding to each cluster, and performing curve fitting on each temperature set; calculating an error factor for each temperature, wherein the error factor is positively correlated with the difference between the temperature and the fitted temperature value; In response to the error factor of the current temperature being greater than or less than a threshold, the fitting temperature value corresponding to the current temperature is taken as the actual temperature; in response to the error factor of the current temperature being equal to the threshold, the current temperature is taken as the actual temperature; the threshold is the average of the error factors of all temperatures in the temperature set except the current temperature; When the actual temperature reaches the preset target temperature, the staff is warned to perform subsequent processing operations.

2. The method for monitoring temperature changes during the melting process of a metal raw material according to claim 1, characterized in that: Before curve fitting for each temperature set, it also includes: Obtain the comprehensive similarity between each temperature and the temperature corresponding to each cluster center, and merge each temperature into the cluster where the cluster center with the largest comprehensive similarity is located; the calculation method of the comprehensive similarity is specifically as follows: ； in Indicates temperature With The temperature corresponding to the cluster center The comprehensive similarity of Indicates temperature The rate of change of Indicates temperature The rate of change of Indicates temperature The corresponding collection time; Indicates temperature The collection time; is the normalization function.

3. The method for monitoring temperature changes during the melting process of a metal raw material according to claim 1, characterized in that: The error factor acquisition method is: ； in Indicates The error factor of each temperature; Indicates Temperature; Indicates The fitted temperature value for a temperature.

4. The method for monitoring temperature changes during the melting process of a metal raw material according to claim 1, characterized in that: The error factor acquisition method is: ； in Indicates The error factor of each temperature; Indicates Temperature; Indicates The fitting temperature value of each temperature; Indicates the temperature concentration except the The rest of the temperature Temperature; Indicates the temperature concentration except the Except the temperature The fitting temperature value corresponding to each temperature; Indicates the temperature concentration except the The total number of temperatures other than the temperature; Represents the tuning factor.

5. The method for monitoring temperature changes during the melting process of a metal raw material according to claim 1, characterized in that: The method for obtaining the change rate is: ； in Indicates The rate of change of temperature; Indicates Temperature; Indicates Temperature; Indicates the preset target temperature; Represents the tuning factor.

6. The method for monitoring temperature changes during the melting process of a metal raw material according to claim 1, characterized in that: Clustering the change rate sequence to obtain a clustering result includes: Based on the acquisition time and temperature in the change rate sequence, the DBSCAN clustering algorithm is used to perform clustering and obtain the clustering results.

7. The method for monitoring temperature changes during the melting process of a metal raw material according to claim 1, characterized in that: Perform curve fitting for each temperature set, including: The temperatures included in each temperature set are curve fitted using the least squares method.

8. A temperature change monitoring system during the melting process of metal raw materials, characterized in that: The method comprises a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the method for monitoring temperature changes in the metal raw material melting process as described in any one of claims 1 to 7 is implemented.

Images