A sintering temperature control system for a medium frequency furnace

Abstract

The present invention relates to the technical field of sintering temperature control of medium frequency furnaces, and discloses a sintering temperature control system for medium frequency furnaces. The system includes a multi-source data acquisition module, which is used to obtain multi-dimensional thermodynamic data in the sintering process in real time, covering temperature distribution, sintering power and heat flow trajectory data; a temperature field analysis module, which performs dynamic compensation and correction processing on the multi-dimensional thermodynamic data and inputs it into the sintering analysis processing layer for feature modeling, and generates sintering temperature control instructions. The sintering analysis processing layer includes a data preprocessing module and a thermodynamic modeling module. The data preprocessing module performs operations such as temperature interval division and outlier removal on the original data, and the thermodynamic modeling module is obtained through collaborative training based on historical data. Through multi-dimensional data acquisition and analysis processing, the system can accurately control the sintering temperature, effectively improve the quality of sintered products and production efficiency, and has good versatility and adaptability.

Description

Technical Field

[0001] The invention relates to the technical field of medium frequency furnace sintering temperature control, in particular to a sintering temperature control system of a medium frequency furnace. Background Art

[0002] In modern industrial production, the medium-frequency (IF) furnace sintering process is widely used in a variety of fields, including metal processing and powder metallurgy. This process achieves specific physical and chemical properties by sintering materials at high temperatures, thereby meeting the manufacturing needs of various industrial products. However, current IF furnace sintering temperature control technology faces numerous challenges.

[0003] Temperature control accuracy is difficult to guarantee. Traditional sintering temperature control systems often rely solely on a single temperature sensor for data collection. The information obtained in this way is extremely limited and cannot fully reflect the complex temperature distribution during the sintering process. For example, in a large medium-frequency furnace, the actual temperature distribution of materials in different areas may have large deviations due to differences in heating positions, heat dissipation conditions and other factors. Control based solely on single-point temperature measurement results can easily lead to insufficient or excessive sintering of some materials, seriously affecting the consistency and stability of product quality. Taking the manufacturing of powder metallurgy parts as an example, if the sintering temperature is not properly controlled, key performance indicators such as the density and hardness of the parts will fluctuate. In subsequent mechanical processing and use, deformation, fracture and other problems are likely to occur, increasing production costs and product rejection rates.

[0004] Insufficient dynamic response capabilities to complex thermal processes. The sintering process is a complex thermal process involving multiple physical and chemical changes. Its parameters such as heat flow trajectory and sintering power will change dynamically with time and material state. The existing control system lacks an effective monitoring and timely response mechanism for these dynamic changes. When the sintering power fluctuates suddenly or the heat flow is abnormally distributed, the traditional system cannot quickly adjust the temperature control strategy, causing the sintering process to deviate from the ideal state, thereby affecting product quality. For example, during the sintering process of ceramic materials, if the heat flow distribution changes suddenly and the control system cannot respond in time, the ceramic products may experience uneven internal stress, leading to surface cracks or internal defects.

[0005] Lack of comprehensive analysis and utilization of multi-dimensional thermodynamic data. In the actual sintering process, there are complex correlations between multi-dimensional thermodynamic data such as temperature distribution, sintering power, and heat flow trajectory. However, current control systems often process this data in isolation, failing to fully tap the potential information behind the data, making it difficult to achieve precise control of the sintering process. For example, in the metal smelting process, although temperature and power data are known, the impact of heat flow trajectory on temperature distribution is not considered. It is impossible to accurately predict the temperature change trend and adjust the control parameters in advance, which affects the smelting efficiency and metal quality.

[0006] Existing systems also have flaws in data processing and modeling. Data preprocessing is crude and simplistic, failing to effectively remove abnormal data and properly standardize the data. This leads to inaccurate subsequent temperature control instructions. Furthermore, established thermodynamic models are often based on simple empirical formulas or fixed parameters, making them incapable of adapting to varying sintering processes and material properties, limiting the system's versatility and adaptability. Summary of the Invention

[0007] The object of the present invention is to provide a sintering temperature control system for a medium frequency furnace to solve the problems raised in the above background technology.

[0008] To achieve the above object, the present invention provides the following technical solution: a sintering temperature control system for an intermediate frequency furnace, the system comprising:

[0009] A multi-source data acquisition module is used to acquire multi-dimensional thermodynamic data during the sintering process in real time. The multi-dimensional thermodynamic data includes a first monitoring sequence corresponding to temperature distribution data, a second monitoring sequence corresponding to sintering power data, and a third monitoring sequence corresponding to heat flow trajectory data. The temperature distribution data includes first temperature field distribution data generated by a multi-zone heating device and second thermal radiation feedback data collected by an infrared array sensor.

[0010] A temperature field analysis module is used to perform dynamic compensation and correction processing on the multi-dimensional thermodynamic data and input the data into a sintering analysis processing layer for feature modeling, and generate sintering temperature control instructions based on the output results of the sintering analysis processing layer;

[0011] The sintering analysis and processing layer includes a data preprocessing module and a thermodynamic modeling module, wherein the data preprocessing module is used to divide the original thermodynamic data stream into temperature intervals and eliminate outliers, and the thermodynamic modeling module is obtained through collaborative training based on historical temperature data and historical power data of multiple sintering cycles; the thermodynamic modeling module includes a heat flow response layer, a compensation correction layer and a temperature decision layer connected in sequence.

[0012] Preferably, the heat flow response layer is used to perform spatial domain correlation processing on different monitoring sequences in the original thermodynamic data stream to generate heat flow coupling characteristic data; the compensation correction layer is used to model the dynamic heat conduction relationship between the heat flow coupling characteristic data corresponding to each monitoring sequence to generate compensated temperature field data; the temperature decision layer is used to perform multi-dimensional integration based on the compensated temperature field data and the heat flow coupling characteristic data to generate sintering temperature control instructions.

[0013] Preferably, the modeling of the dynamic heat conduction relationship between the heat flow coupling characteristic data corresponding to each monitoring sequence to generate the compensated temperature field data includes:

[0014] A dynamic compensation correction algorithm is used to identify temperature key nodes in the heat-flow coupling characteristic data, and a temperature compensation sequence corresponding to each monitoring sequence is determined based on the heating area type corresponding to each temperature key node;

[0015] The deviation coefficient between the temperature nodes of the same heating area in the temperature compensation sequences corresponding to any two monitoring sequences is calculated, and the compensated temperature field data between the any two monitoring sequences is generated based on the deviation coefficient.

[0016] Preferably, the calculation of the deviation coefficient between the temperature nodes of the same heating area in the temperature compensation sequence corresponding to any two monitoring sequences includes:

[0017] When the number of temperature nodes in the temperature compensation sequences corresponding to any two monitoring sequences is inconsistent, virtual node interpolation is performed based on the heating area parameters corresponding to the terminal temperature node of the one with fewer temperature nodes, and the deviation coefficient between the temperature nodes in the same heating area is calculated based on the interpolated data.

[0018] Preferably, the data preprocessing module is specifically used to:

[0019] Performing isogradient division on the first monitoring sequence, the second monitoring sequence, and the third monitoring sequence according to a preset temperature range to generate standardized first temperature field data, standardized second power data, and standardized third heat flow data;

[0020] A dynamic weight adjustment method is used to perform real-time correction on the standardized first temperature field data and the standardized second power data, and a fixed weight correction method is used to perform steady-state optimization on the standardized third heat flow data, and a first correction sequence, a second correction sequence and a third correction sequence are output; wherein, the first correction sequence includes the corrected first temperature field distribution data and the corrected second thermal radiation feedback data.

[0021] Preferably, the data preprocessing module is further used for:

[0022] Calculating the thermal conductivity of the corrected first temperature field distribution data and the corrected second thermal radiation feedback data in a historical sintering cycle;

[0023] Predicting an expected temperature value of the corrected second thermal radiation feedback data in the real-time sintering cycle according to the heat transfer coefficient and the heating parameters of the corrected first temperature field distribution data in the real-time sintering cycle;

[0024] Target heat flux compensation data is generated based on the corrected second thermal radiation feedback data and its expected temperature value, and a monitoring sequence corresponding to the target heat flux compensation data is used as a first correction sequence.

[0025] Preferably, the compensation correction layer specifically includes:

[0026] a heat conduction analysis unit, configured to perform heat flow path tracing for each monitoring sequence in the heat flow coupling characteristic data, so as to extract a corresponding thermal gradient propagation chain from each monitoring sequence;

[0027] The temperature field matching unit is used to spatially map the thermal gradient propagation chain extracted from each monitoring sequence with the corresponding thermal flow coupling characteristic data to generate compensated temperature field data.

[0028] Preferably, the compensation correction layer further comprises:

[0029] The thermal lag elimination unit is used to perform thermal response delay correction processing on the compensated temperature field data.

[0030] Preferably, the temperature decision layer specifically includes:

[0031] A multi-dimensional integration unit, comprising a plurality of temperature control nodes, each temperature control node being connected to each monitoring sequence in the compensated temperature field data and the heat-flow coupling characteristic data through an associated configuration;

[0032] a dynamic weight optimization unit, configured to iteratively adjust the associated configuration using a dynamic weight optimization algorithm to minimize the deviation between the sintering temperature control instruction and the actual temperature distribution;

[0033] The thermal imbalance identification unit is used to locate the sintering abnormality area based on the compensated temperature field data and the heat-flow coupling characteristic data, and generate a sintering temperature control instruction.

[0034] Preferably, the dynamic weight adjustment method specifically includes:

[0035] Generate adaptive correction parameters based on the thermal field distribution characteristics of the real-time sintering environment;

[0036] A sliding window mechanism is used to perform segmented correction processing on the standardized first temperature field data.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] In terms of data acquisition and processing, the multi-source data acquisition module can acquire multi-dimensional thermodynamic data of the sintering process in real time, including the first monitoring sequence corresponding to the temperature distribution data, the second monitoring sequence corresponding to the sintering power data, and the third monitoring sequence corresponding to the heat flow trajectory data. The temperature distribution data covers the first temperature field distribution data generated by the multi-zone heating device and the second thermal radiation feedback data collected by the infrared array sensor. This comprehensive data acquisition method overcomes the limitations of traditional single-point acquisition and provides a rich and accurate information basis for precise control. The data preprocessing module performs operations such as temperature interval division, outlier removal, and standardization on the original thermodynamic data stream, which can effectively improve data quality. The dynamic weight adjustment method is used to perform real-time correction and steady-state optimization of different types of data, ensuring that the data input into subsequent modules can better reflect the actual sintering situation, laying a solid foundation for subsequent precise analysis and regulation.

[0039] From the perspective of temperature field analysis and modeling, the temperature field analysis module dynamically compensates and corrects multi-dimensional thermodynamic data and inputs it into the sintering analysis processing layer for feature modeling. The thermodynamic modeling module in the sintering analysis processing layer is collaboratively trained based on historical temperature and power data from multiple sintering cycles, enabling it to better adapt to different sintering processes and material characteristics. The heat flow response layer spatially correlates different monitoring sequences to generate heat flow coupling feature data, uncovering potential connections between the data. The compensation correction layer generates compensated temperature field data by modeling the dynamic heat conduction relationships between the heat flow coupling feature data. It accounts for the inconsistency of the number of temperature nodes across different monitoring sequences and ensures computational accuracy through virtual node interpolation, significantly improving the simulation and prediction capabilities of complex heat conduction processes. The temperature decision layer generates sintering temperature control commands based on a multi-dimensional integration of the compensated temperature field data and the heat flow coupling feature data. A dynamic weight optimization algorithm iteratively adjusts the correlation configuration to minimize the deviation between the control commands and the actual temperature distribution. It also allows for data-based location of sintering anomalies, further improving the accuracy and timeliness of temperature control.

[0040] In actual application, the system has significantly improved the quality of sintered products. Precise temperature control can ensure that the material is heated evenly throughout the sintering process, avoiding under-sintering or over-sintering, thereby improving the consistency and stability of the product. In the field of metal material processing, it can effectively reduce internal defects in products, improve the mechanical properties of products, reduce scrap rates, and save production costs. In the powder metallurgy industry, it helps to manufacture parts with higher precision and better performance, meeting the strict requirements of high-end manufacturing for material quality. At the same time, the system has a strong dynamic response capability and can promptly respond to changes in parameters such as heat flow and power during the sintering process, ensuring that the sintering process is always in an ideal state, improving production efficiency, and enhancing the company's competitiveness in the market.

[0041] The system also has good versatility and adaptability. Because it is trained and modeled based on a large amount of historical data, it can be flexibly adjusted according to different sintering processes and material characteristics. It is applicable to medium-frequency furnace sintering processes in various industrial fields and provides strong support for the intelligent upgrade of industrial production. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a working principle diagram of the sintering temperature control system of the intermediate frequency furnace of the present invention;

[0043] Figure 2 This is the working principle diagram of each module of the sintering analysis processing layer;

[0044] Figure 3 This is a working diagram for calculating the deviation coefficient (handling inconsistent node numbers);

[0045] Figure 4 This is the schematic diagram of the data preprocessing module;

[0046] Figure 5 This is a diagram showing the working principle of the internal units of the compensation correction layer. DETAILED DESCRIPTION

[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0048] See also Figure 1-Figure 5 The present invention provides a sintering temperature control system for a medium frequency furnace, and the specific implementation steps are as follows:

[0049] The sintering temperature control system of this intermediate frequency furnace is mainly composed of a multi-source data acquisition module and a temperature field analysis module. The function of the multi-source data acquisition module is to obtain multi-dimensional thermodynamic data of the sintering process in real time. In actual operation, when the multi-zone heating device is working, it will generate first temperature field distribution data, which reflects the temperature distribution of different areas in the furnace under the action of the heating device. At the same time, the infrared array sensor is also continuously collecting second thermal radiation feedback data, which can reflect the temperature status in the furnace from the perspective of thermal radiation. These two parts of data together constitute the temperature distribution data and are presented in the form of a first monitoring sequence. In addition, the module also collects sintering power data to form a second monitoring sequence, and heat flow trajectory data to generate a third monitoring sequence. These data comprehensively reflect the thermodynamic information of the sintering process from different dimensions.

[0050] The temperature field analysis module receives multi-dimensional thermodynamic data from the multi-source data acquisition module. First, these data are dynamically compensated and corrected. This is because during the actual acquisition process, the data may be interfered with by various factors and may contain errors or incompleteness. The processed data will be input into the sintering analysis processing layer. The sintering analysis processing layer includes a data preprocessing module and a thermodynamic modeling module. The data preprocessing module divides the original thermodynamic data stream into temperature intervals and eliminates outliers to ensure the accuracy and validity of the data. The thermodynamic modeling module is obtained through collaborative training based on historical temperature data and historical power data of multiple sintering cycles. It includes a heat flow response layer, a compensation correction layer, and a temperature decision layer connected in sequence. After processing by the sintering analysis processing layer, a sintering temperature control instruction will be generated based on the output results, thereby achieving precise control of the sintering temperature of the medium frequency furnace.

[0051] The specific embodiments of the present invention are further described below by means of specific examples:

[0052] Example 1:

[0053] The heat flow response layer spatially correlates the different monitoring sequences in the original thermodynamic data stream. During the actual sintering process in an intermediate frequency furnace, the temperature distribution data, sintering power data, and heat flow trajectory data contained in different monitoring sequences exhibit certain spatial correlations. The heat flow response layer analyzes this data to identify the connections between them, thereby generating heat flow coupling feature data. For example, high-temperature regions in the temperature distribution data correlate with the direction and flow rate of heat flow in the heat flow trajectory data. The heat flow response layer uses a specific algorithm to identify these connections. The compensation correction layer models the dynamic heat conduction relationship between the heat flow coupling feature data corresponding to each monitoring sequence. It first uses a dynamic compensation correction algorithm to identify temperature key nodes in the heat flow coupling feature data. These temperature key nodes are crucial for controlling the sintering temperature, representing sensitive areas within the furnace to temperature changes. Based on the heating zone type corresponding to each temperature key node, the temperature compensation sequence corresponding to each monitoring sequence is determined.

[0054] Next, the deviation coefficient between temperature nodes in the same heating zone in the temperature compensation sequence corresponding to any two monitoring sequences is calculated. If the number of temperature nodes in the temperature compensation sequences corresponding to the two monitoring sequences is inconsistent, virtual node interpolation is performed based on the heating zone parameters corresponding to the terminal temperature node in the sequence with the fewer temperature nodes. The deviation coefficient is then calculated based on this interpolated data, ultimately generating the compensated temperature field data. The temperature decision layer performs multi-dimensional integration based on the compensated temperature field data and the thermal-fluid coupling characteristic data to generate sintering temperature control instructions. The multi-dimensional integration unit contains multiple temperature control nodes, each of which is connected to each monitoring sequence in the compensated temperature field data and the thermal-fluid coupling characteristic data through an association configuration. The dynamic weight optimization unit iteratively adjusts the association configuration using a dynamic weight optimization algorithm to minimize the deviation between the sintering temperature control instructions and the actual temperature distribution. The thermal imbalance identification unit locates sintering anomaly areas based on the compensated temperature field data and the thermal-fluid coupling characteristic data, thereby generating reasonable sintering temperature control instructions.

[0055] Assume that in a specific medium frequency furnace sintering production process, the multi-source data acquisition module continuously collects multi-dimensional thermodynamic data. Among them, the first monitoring sequence of temperature distribution data shows that within a period of time after the start of sintering, different areas in the furnace show different temperature change trends. For example, the temperature of the area close to the heating source rises faster, while the temperature of the area far from the heating source rises relatively slowly. The second monitoring sequence of sintering power data records the power output of the equipment at each moment. As the temperature rises, the power output is gradually adjusted to maintain suitable sintering conditions. The third monitoring sequence of heat flow trajectory data reflects the changes in the flow direction and intensity of the heat flow in the furnace.

[0056] The heat flow response layer starts working and receives data from different monitoring sequences. For a certain moment in the temperature distribution data, such as the 10th minute of sintering, the heat flow response layer finds that in the area where the temperature rises rapidly, the heat flow trajectory data shows that the heat flow intensity is large and the flow direction is concentrated, while in the area where the temperature rises slowly, the heat flow intensity is weak. Through a specific spatial domain association algorithm, the heat flow response layer associates the data of these different monitoring sequences and mines the potential connections therein. For example, it finds that there is a certain correspondence between temperature changes and heat flow intensity and flow direction, that is, the area with large heat flow intensity and concentrated flow direction has a fast temperature rise rate. Based on these associations, the heat flow response layer generates heat flow coupling feature data. For example, information such as the temperature change rate, heat flow intensity and flow direction are integrated together to form a comprehensive feature vector that represents the heat flow coupling characteristics of a specific area at that moment.

[0057] The compensation correction layer begins processing the thermal-flux coupling characteristic data. First, it uses a dynamic compensation correction algorithm to identify the temperature critical nodes in the thermal-flux coupling characteristic data. Assuming that the temperature change in a certain area of ​​the thermal-flux coupling characteristic data has a significant impact on the entire sintering process, the temperature node in this area is identified as a temperature critical node. After analysis, the heating area type corresponding to this temperature critical node is a direct heating zone. Based on this, the compensation correction layer determines the temperature compensation sequence corresponding to each monitoring sequence. For the monitoring sequence of temperature distribution data, the corresponding temperature compensation value is determined based on the temperature change of the heating area; for the monitoring sequence of sintering power data, the temperature compensation value corresponding to the power adjustment is determined based on the relationship between power and temperature.

[0058] Calculate the deviation coefficient between temperature nodes in the same heating region in the temperature compensation sequences corresponding to any two monitoring sequences. For example, compare the temperature nodes in the directly heated region of the temperature distribution data monitoring sequence and the sintering power data monitoring sequence. If the number of temperature nodes in the temperature compensation sequences corresponding to the two monitoring sequences is inconsistent, assuming the temperature distribution data monitoring sequence has more temperature nodes and the sintering power data monitoring sequence has fewer temperature nodes, perform virtual node interpolation based on the heating region parameters corresponding to the terminal temperature node in the sintering power data monitoring sequence. For example, based on parameters such as the heating power and heating time of the terminal temperature node, insert a virtual node in the sintering power data monitoring sequence to ensure that the number of temperature nodes in the two sequences is the same. Then, calculate the deviation coefficient between temperature nodes in the same heating region based on the interpolated data. The deviation coefficient is obtained by calculating the ratio of the temperature difference to the average temperature, and then generating compensated temperature field data. This compensated temperature field data comprehensively accounts for the differences between the different monitoring sequences and more accurately reflects the actual temperature field conditions within the furnace.

[0059] The temperature decision layer performs multi-dimensional integration based on the compensated temperature field data and the heat flow coupling characteristic data to generate sintering temperature control instructions. The temperature control nodes in the multi-dimensional integration unit begin to function. Assume that at a certain moment, a temperature control node is connected to monitoring sequences containing temperature distribution data, sintering power data, and heat flow trajectory data. It makes preliminary control decisions based on the current status of these data, such as excessive temperature, excessive power, or abnormal heat flow. The dynamic weight optimization unit iteratively adjusts the association between the temperature control node and each monitoring sequence based on the actual sintering results. For example, if a deviation between the temperature control instructions and the actual temperature distribution is detected during a sintering process, the dynamic weight optimization unit uses a dynamic weight optimization algorithm to adjust the weights of the temperature control node for different monitoring sequence data. If the temperature distribution data is found to have a more critical impact on temperature control, the weight of the temperature distribution data monitoring sequence is appropriately increased, while the weights of other sequences are reduced to minimize the deviation. The thermal imbalance identification unit locates sintering anomalies based on the compensated temperature field data and the heat flow coupling characteristic data. For example, when the compensated temperature field data shows that the temperature of a certain area deviates significantly from the normal range, or when abnormal heat flow fluctuations occur in the heat flow coupling characteristic data, the thermal imbalance identification unit determines that the area is a sintering abnormality area. Then, the multi-dimensional integration unit and the dynamic weight optimization unit jointly generate reasonable sintering temperature control instructions based on the information provided by the thermal imbalance identification unit. For example, for abnormal areas with too high temperatures, the instructions reduce the heating power of the area; for areas with abnormal heat flow, the direction or intensity of the heat flow is adjusted, etc., to ensure that the sintering process of the medium frequency furnace can be carried out stably and accurately.

[0060] Example 2:

[0061] The data preprocessing module divides the first, second, and third monitoring sequences into isogradient partitions based on preset temperature ranges. These preset temperature ranges are pre-set based on the operating characteristics of the intermediate frequency furnace and the requirements of the sintering process. During this partitioning process, the temperature, power, and heat flux data are processed isogradiently to generate standardized first temperature field data, standardized second power data, and standardized third heat flux data. A dynamic weight adjustment method is then used to perform real-time corrections on the standardized first temperature field data and the standardized second power data. This dynamic weight adjustment method generates adaptive correction parameters based on the thermal field distribution characteristics of the real-time sintering environment. A sliding window mechanism is then used to perform segmented corrections on the standardized first temperature field data. A fixed weight correction method is used for steady-state optimization of the standardized third heat flux data. After these processes, the first, second, and third correction sequences are output. The first correction sequence contains the corrected first temperature field distribution data and the corrected second thermal radiation feedback data. The data preprocessing module also calculates the thermal conductivity coefficients of the corrected first temperature field distribution data and the corrected second thermal radiation feedback data over the historical sintering cycles. Based on this thermal conductivity and the heating parameters of the corrected first temperature field distribution data during the real-time sintering cycle, the expected temperature value of the corrected second thermal radiation feedback data during the real-time sintering cycle is predicted. Target heat flux compensation data is generated based on the corrected second thermal radiation feedback data and its expected temperature value, and the monitoring sequence corresponding to the target heat flux compensation data is used as the first correction sequence.

[0062] In an actual production scenario of metal powder sintering, take the metal powder sintering process required to manufacture high-precision mechanical parts as an example. Before the start of the metal powder sintering operation, the engineer pre-sets the temperature range based on the specifications of the medium frequency furnace, the characteristics of the metal powder, and the desired sintering effect. For example, for this specific metal powder, the temperature range of the entire sintering process is divided from room temperature to 1200°C, and a gradient of 50°C is set. After the sintering process is started, the multi-source data acquisition module starts working and continuously collects multi-dimensional thermodynamic data. The first monitoring sequence records the temperature distribution data, including the first temperature field distribution data generated by the multi-zone heating device and the second thermal radiation feedback data collected by the infrared array sensor; the second monitoring sequence records the sintering power data; and the third monitoring sequence records the heat flow trajectory data.

[0063] The data preprocessing module first performs isogradient segmentation on these raw data. For example, when the collected real-time temperature data falls within the 100-150°C range from the first monitoring sequence, the data preprocessing module normalizes it to a specific numerical form within that range, conforming to pre-set normalization rules and forming standardized first temperature field data. Similarly, the sintering power data from the second monitoring sequence and the heat flow trajectory data from the third monitoring sequence undergo similar processing, generating standardized second power data and standardized third heat flow data, respectively.

[0064] The data preprocessing module uses a dynamic weighting method to perform real-time corrections on the standardized first temperature field data and the standardized second power data. At 30 minutes into the sintering process, the thermal field distribution within the furnace changed. Due to the initial fusion reaction of the metal powder in a specific area, the temperature change trend in that area slightly deviated from the expected trend, altering the thermal field distribution characteristics. The data preprocessing module generates adaptive correction parameters based on these real-time changes in the thermal field distribution. For example, the correction parameters are determined by analyzing factors such as the rate of temperature change, the location of the area, and the temperature difference with surrounding areas. A sliding window mechanism is then used to perform segmented corrections on the standardized first temperature field data. Assuming the sliding window size is set to 10 data points, the window slides sequentially, starting with the earliest data point. As the window moves, the temperature data within the window is adjusted based on the adaptive correction parameters. If the adaptive correction parameters indicate that the overall temperature data for a certain period is too low, the temperature data within that window is increased accordingly to more closely reflect the actual value.

[0065] For the standardized third heat flux data, the data preprocessing module uses a fixed weight correction method for steady-state optimization. Based on previous experience and experimental data, a fixed weight value is determined. For example, a weight of 0.8 is set. A weighted calculation is performed on the standardized third heat flux data to eliminate fluctuations caused by measurement errors or environmental interference, making it more stable. The first, second, and third correction sequences are then output. The first correction sequence includes the corrected first temperature field distribution data and the corrected second thermal radiation feedback data.

[0066] After completing the basic data processing described above, the data preprocessing module further calculates the thermal conductivity coefficients of the corrected first temperature field distribution data and the corrected second thermal radiation feedback data over the historical sintering cycles. For example, by analyzing historical data from the past 10 sintering cycles of the same metal powder, it is calculated that the thermal conductivity coefficient is 0.6 under a specific temperature range and sintering conditions. Then, based on this thermal conductivity coefficient and the heating parameters of the corrected first temperature field distribution data during the real-time sintering cycle, the expected temperature value of the corrected second thermal radiation feedback data during the real-time sintering cycle is predicted. For example, during the current real-time sintering cycle, the corrected first temperature field distribution data shows that the temperature of a certain area increased from 800°C to 850°C within 10 minutes, while the heating power remained constant. Based on the thermal conductivity coefficient and heating parameters, the expected temperature value of the corrected second thermal radiation feedback data for this area after 10 minutes is predicted to be 830°C.

[0067] Finally, the target heat flux compensation data is generated based on the corrected second thermal radiation feedback data and its expected temperature value. If the actual measured corrected second thermal radiation feedback data is 820°C after 10 minutes, which differs by 10°C from the expected temperature value, the data preprocessing module will calculate the heat flux data that needs to be compensated based on this temperature difference and other relevant factors, generate the target heat flux compensation data, and use the monitoring sequence corresponding to the target heat flux compensation data as the first correction sequence. Subsequent system components such as the temperature field analysis module will perform more accurate sintering temperature control based on this more precisely processed first correction sequence, ensuring that the metal powder sintering process can proceed according to the predetermined process requirements and produce mechanical parts that meet precision standards.

[0068] Example 3:

[0069] The heat conduction analysis unit in the compensation correction layer tracks the heat flow path for each monitoring sequence in the heat flow coupling characteristic data. In the complex thermal environment of the medium frequency furnace, the heat flow propagation path is complex and changeable. The heat conduction analysis unit extracts the corresponding thermal gradient propagation chain from each monitoring sequence through specific algorithms and models. These thermal gradient propagation chains reflect the propagation direction and intensity changes of the heat flow in different areas. The temperature field matching unit spatially maps the thermal gradient propagation chain extracted from each monitoring sequence with the corresponding heat flow coupling characteristic data. Through this spatial mapping, the relationship between heat flow and temperature field can be more accurately understood, thereby generating compensated temperature field data. During the spatial mapping process, the corresponding relationship between the spatial positions of the data of different monitoring sequences and the physical laws of heat flow propagation are taken into account to ensure that the generated compensated temperature field data can accurately reflect the actual temperature changes.

[0070] Taking the medium-frequency furnace used in the ceramic firing process as an example, in this ceramic firing scenario, the medium-frequency furnace's multi-source data acquisition module continuously collects multi-dimensional thermodynamic data related to the sintering process. These data will be transmitted to the compensation correction layer for processing.

[0071] After ceramic firing begins, heat flow coupling characteristic data accumulates over time. The heat conduction analysis unit in the compensation layer begins operating, tracing the heat flow path for each monitoring sequence. For example, in the monitoring sequence corresponding to the temperature distribution data, the heat conduction analysis unit focuses on the temperature changes in the area where the ceramic body is located within the medium-frequency furnace. As firing progresses to 30 minutes, the temperature of the ceramic body in a corner of the furnace rises from room temperature to 300°C. By continuously monitoring and analyzing the temperature data in this area and combining it with basic principles of heat transfer, the heat conduction analysis unit discovers that heat is transferred from the heating element to this corner along a specific path. This path involves both convection in the furnace air and heat conduction within the ceramic body itself. The heat conduction analysis unit organizes this data reflecting the heat flow transfer process and extracts the corresponding thermal gradient propagation chain. This thermal gradient propagation chain details the temperature gradient changes in this area as the heat flow propagates. For example, the temperature gradient is larger near the heating element and gradually decreases as the heat flow moves away from the heating element.

[0072] For the monitoring sequence corresponding to the sintering power data, the heat conduction analysis unit also tracks the heat flow path. Assume that during the firing process, the power output of the medium-frequency furnace is adjusted at different stages to ensure uniform heating of the ceramic body. At the 60th minute of firing, the power suddenly increases, and the heat conduction analysis unit finds that this power change causes changes in the generation and propagation of heat flow. By analyzing the relationship between power changes and heat flow changes, combined with the structural characteristics of the furnace, the path of heat flow propagation from the power-increased area to various parts of the ceramic body is determined, and then the thermal gradient propagation chain corresponding to the monitoring sequence is extracted. This thermal gradient propagation chain reflects the impact of the change in heat flow intensity and direction caused by the power change on the temperature gradient.

[0073] After the thermal gradient propagation chain is extracted, the temperature field matching unit begins operation. It spatially maps the thermal gradient propagation chain extracted from each monitoring sequence with the corresponding thermal flow coupling feature data. Taking the thermal gradient propagation chain from the temperature distribution data monitoring sequence as an example, the temperature field matching unit locates the heat flow propagation path and temperature gradient change information recorded in the thermal gradient propagation chain within the spatial range described by the thermal flow coupling feature data. For example, the thermal gradient propagation chain shows that at a certain moment, heat flow propagates downward from the heating element at the furnace top, with the temperature gradient gradually decreasing during the propagation process. The temperature field matching unit locates the corresponding spatial region in the thermal flow coupling feature data and combines the temperature change information in the thermal gradient propagation chain with the thermal flow coupling feature data for that region. Specifically, the temperature values ​​in the thermal gradient propagation chain are matched with the temperature distribution and heat flow intensity information for that region in the thermal flow coupling feature data to generate compensated temperature field data. This spatial mapping provides a more precise understanding of the spatial distribution of heat flow within the furnace and its impact on the temperature field, thereby generating compensated temperature field data that better reflects actual conditions and provides a more accurate basis for subsequent temperature control. In this ceramic firing example, the generated compensated temperature field data can help operators more clearly understand which areas in the furnace have too high or too low temperatures, and whether the distribution of heat flow is reasonable, so as to adjust the firing parameters in time to ensure the quality of ceramic firing.

[0074] Example 4:

[0075] During the sintering process, a certain delay in heat transfer can lead to deviations in temperature control. The thermal lag elimination unit in the compensation correction layer is designed to address this issue. The thermal lag elimination unit corrects the thermal response delay of the compensated temperature field data. It establishes a thermal response delay model through analysis of historical data and feedback from real-time monitoring data. Based on this model, the thermal response delay in the compensated temperature field data is compensated. For example, when a delay in temperature change is detected, the thermal lag elimination unit predicts the temperature change trend based on the model and adjusts the compensated temperature field data in advance, so that the resulting compensated temperature field data can more accurately reflect the current actual temperature conditions, thereby improving the accuracy of temperature control.

[0076] Taking the firing process of glass products as an example, when using a medium-frequency furnace to fire high-precision optical glass, the furnace temperature must be strictly controlled to ensure the optical properties of the glass. Assume that at a certain stage, the target temperature of the glass firing is 850°C, and the temperature fluctuation is required to be controlled within the range of ±5°C.

[0077] During the firing process, the thermal hysteresis elimination unit of the compensation correction layer comes into play. The multi-source data acquisition module continuously collects data. The monitoring sequence corresponding to the temperature distribution data shows that the temperature change in a specific area within the furnace has a heat transfer delay.

[0078] The thermal hysteresis elimination unit establishes a thermal response delay model using the following formula: In this formula, Indicates the temperature correction required due to thermal hysteresis (unit: °C); is the thermal hysteresis coefficient, which is a proportional constant determined by comprehensive factors such as the historical operating data of the medium frequency furnace, the furnace structure, and the characteristics of the glass material. In this firing scenario, after a large number of experiments and data analysis, The value is 0.8 (℃ / s); Indicates the heat transfer delay time (unit: s), which is calculated based on the heat flow trajectory data and temperature change monitoring data. At the current stage, the heat transfer delay time of this area 3s.

[0079] According to the above formula, , that is, the temperature in this area needs to be corrected by 2.4℃ due to thermal hysteresis.

[0080] The thermal hysteresis elimination unit performs thermal response delay correction processing on the compensated temperature field data based on the calculated temperature correction value. In actual operation, it first predicts the temperature change trend of the area over a period of time in the future without thermal hysteresis correction based on the current temperature monitoring data. Assuming that the current temperature in the area is 845°C and the temperature rise rate is 1°C / s, if no thermal hysteresis correction is performed, the temperature is expected to reach 845+1×3=848°C in 3 seconds. However, considering that the thermal hysteresis correction value is 2.4°C, the thermal hysteresis elimination unit will adjust the target temperature of the area to 848+2.4=850.4°C. The thermal hysteresis elimination unit then feeds this corrected temperature information back to the temperature decision layer. The temperature decision layer adjusts the power of the heating device based on the corrected temperature data. For example, the original heating device power was maintained at 600kW. After receiving the temperature data corrected by the thermal lag elimination unit, it was found that the temperature was close to the target temperature upper limit. Therefore, the heating power was appropriately reduced to 550kW to avoid excessive temperature in this area, ensuring that the glass firing process is carried out under precise temperature control and guaranteeing the quality of the glass products.

[0081] Example 5:

[0082] The temperature decision layer's multidimensional integration unit, dynamic weight optimization unit, and thermal imbalance identification unit work in synergy. The temperature decision layer's multidimensional integration unit consists of multiple temperature control nodes, each connected to various monitoring sequences in the compensated temperature field data and thermal-fluid coupling feature data through association configurations. These temperature control nodes act as multiple "control points" for temperature control, acquiring data from different monitoring sequences and making decisions based on this data. The dynamic weight optimization unit iteratively adjusts the association configurations using a dynamic weight optimization algorithm. In the actual sintering process, the sintering environment is constantly changing, and the impact of different monitoring sequences on temperature control also varies. The dynamic weight optimization unit continuously adjusts the association weights between each temperature control node and the monitoring sequence based on real-time sintering conditions to minimize the deviation between the sintering temperature control instructions and the actual temperature distribution. The thermal imbalance identification unit locates sintering anomalies based on the compensated temperature field data and thermal-fluid coupling feature data. When an anomaly is detected in the compensated temperature field data or the thermal-fluid coupling feature data, the thermal imbalance identification unit quickly identifies the abnormal area in the sintering process and feeds this information back to other units. Then, the multi-dimensional integration unit and the dynamic weight optimization unit jointly generate reasonable sintering temperature control instructions based on the information provided by the thermal imbalance identification unit to ensure that the sintering process of the medium frequency furnace can be carried out stably and accurately.

[0083] For example, the sintering process of a new alloy material requires extremely precise control of the sintering temperature, otherwise its final physical and chemical properties will be affected. During this sintering process using an intermediate frequency furnace, the various units in the temperature decision layer involved in this embodiment work together to play a key role.

[0084] During the initial sintering of the new alloy material, the multi-source data acquisition module collected a large amount of multi-dimensional thermodynamic data. The temperature distribution data revealed significant temperature variations across different areas of the furnace, with some areas experiencing rapid temperature increases and others experiencing slow increases. The sintering power data also continuously adjusted with temperature fluctuations, and the heat flow trajectory data also exhibited complex patterns. After preliminary processing, this data was transmitted to the temperature decision-making layer.

[0085] The multi-dimensional integration unit of the temperature decision layer contains multiple temperature control nodes. Assume that the multi-dimensional integration unit has 5 temperature control nodes, marked as A, B, C, D, and E. Node A is connected to the monitoring sequence of the high-temperature area in the temperature distribution data and the power adjustment sequence of the corresponding area in the sintering power data; node B is responsible for processing the monitoring sequence related to the low-temperature area in the temperature distribution data and the heat flux intensity in the area in the heat flux trajectory data. Each temperature control node preliminarily judges the temperature status of the current area based on the monitoring sequence data connected to it. For example, node A finds that the temperature of the high-temperature area it is connected to is close to the preset sintering temperature upper limit, but the power is still increasing. If it continues in the current state, the temperature of the area is likely to exceed the reasonable range, affecting the sintering quality of the alloy material.

[0086] At this point, the dynamic weight optimization unit comes into play. It iteratively adjusts the associated configuration through a dynamic weight optimization algorithm. In this example, the algorithm continuously adjusts the weight of each temperature control node for different monitoring sequence data based on the real-time sintering conditions. Since the temperature in the high-temperature area monitored by node A is close to the upper limit, the dynamic weight optimization unit will increase the node's weight for the temperature distribution data monitoring sequence and reduce the weight for the power data monitoring sequence. Assume that the original weight of node A for the temperature distribution data monitoring sequence is 0.4, and the weight for the power data monitoring sequence is 0.6. After the algorithm adjustment, the weight of the temperature distribution data monitoring sequence is increased to 0.7, and the weight of the power data monitoring sequence is reduced to 0.3. In this way, node A will rely more on temperature distribution data when making decisions, thereby more accurately judging the temperature conditions in the current area and avoiding the problem of ignoring the rapid temperature rise due to excessive focus on power data.

[0087] The thermal imbalance identification unit locates abnormal sintering areas based on the compensated temperature field data and the heat-flux coupling characteristic data. During the sintering process, the thermal imbalance identification unit continuously analyzes the data. For example, it discovers that the compensated temperature field data in a corner of the furnace indicates a significantly lower temperature than other areas, and the heat-flux coupling characteristic data indicates that the heat flux intensity in this area is abnormally weak, significantly different from the heat flux pattern under normal sintering conditions. Based on this, the thermal imbalance identification unit identifies this corner area as an abnormal sintering area and promptly feeds this information back to the multi-dimensional integration unit and the dynamic weight optimization unit.

[0088] The multi-dimensional integration unit and the dynamic weight optimization unit jointly generate reasonable sintering temperature control instructions based on the information provided by the thermal imbalance identification unit. For the abnormal corner area in the furnace discovered by the thermal imbalance identification unit, the relevant temperature control node in the multi-dimensional integration unit (assuming it is node C, which is responsible for the temperature control of the corner area) will make a decision based on the new weight and the specific situation of the abnormal area. Node C will increase the power output instruction of the heating device in this area. At the same time, the dynamic weight optimization unit will further adjust the weight of node C and the relevant monitoring sequence to enable it to more accurately control the temperature of this area. Through such collaborative work, the medium frequency furnace can more accurately control the temperature of each area in the furnace, ensuring that the new alloy material is sintered under appropriate temperature conditions, thereby ensuring the quality and performance of the alloy material.

[0089] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0090] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A sintering temperature control system for a medium frequency furnace, characterized in that: include: A multi-source data acquisition module is used to acquire multi-dimensional thermodynamic data during the sintering process in real time. The multi-dimensional thermodynamic data includes a first monitoring sequence corresponding to temperature distribution data, a second monitoring sequence corresponding to sintering power data, and a third monitoring sequence corresponding to heat flow trajectory data. The temperature distribution data includes first temperature field distribution data generated by a multi-zone heating device and second thermal radiation feedback data collected by an infrared array sensor. A temperature field analysis module is used to perform dynamic compensation and correction processing on the multi-dimensional thermodynamic data and input the data into a sintering analysis processing layer for feature modeling, and generate sintering temperature control instructions based on the output results of the sintering analysis processing layer; The sintering analysis and processing layer includes a data preprocessing module and a thermodynamic modeling module. The data preprocessing module is used to divide the original thermodynamic data stream into temperature intervals and eliminate outliers. The thermodynamic modeling module is obtained through collaborative training based on historical temperature data and historical power data of multiple sintering cycles. The thermodynamic modeling module includes a heat flow response layer, a compensation correction layer, and a temperature decision layer connected in sequence. The heat flow response layer is used to perform spatial domain correlation processing on different monitoring sequences in the original thermodynamic data stream to generate heat flow coupling characteristic data; the compensation correction layer is used to model the dynamic heat conduction relationship between the heat flow coupling characteristic data corresponding to each monitoring sequence to generate compensated temperature field data; the temperature decision layer is used to perform multi-dimensional integration based on the compensated temperature field data and the heat flow coupling characteristic data to generate sintering temperature control instructions.

2. The medium frequency furnace sintering temperature control system according to claim 1, characterized in that: The modeling of the dynamic heat conduction relationship between the heat flow coupling characteristic data corresponding to each monitoring sequence to generate the compensated temperature field data includes: A dynamic compensation correction algorithm is used to identify temperature key nodes in the heat-flow coupling characteristic data, and a temperature compensation sequence corresponding to each monitoring sequence is determined based on the heating area type corresponding to each temperature key node; The deviation coefficient between the temperature nodes of the same heating area in the temperature compensation sequences corresponding to any two monitoring sequences is calculated, and the compensated temperature field data between the any two monitoring sequences is generated based on the deviation coefficient.

3. The medium frequency furnace sintering temperature control system according to claim 2, characterized in that: The calculation of the deviation coefficient between the temperature nodes of the same heating area in the temperature compensation sequence corresponding to any two monitoring sequences includes: When the number of temperature nodes in the temperature compensation sequences corresponding to any two monitoring sequences is inconsistent, virtual node interpolation is performed based on the heating area parameters corresponding to the terminal temperature node of the one with fewer temperature nodes, and the deviation coefficient between the temperature nodes in the same heating area is calculated based on the interpolated data.

4. The medium frequency furnace sintering temperature control system according to claim 1, characterized in that: The data preprocessing module is specifically used for: Performing isogradient division on the first monitoring sequence, the second monitoring sequence, and the third monitoring sequence according to a preset temperature range to generate standardized first temperature field data, standardized second power data, and standardized third heat flow data; A dynamic weight adjustment method is used to perform real-time correction on the standardized first temperature field data and the standardized second power data, and a fixed weight correction method is used to perform steady-state optimization on the standardized third heat flow data, and a first correction sequence, a second correction sequence and a third correction sequence are output; wherein, the first correction sequence includes the corrected first temperature field distribution data and the corrected second thermal radiation feedback data.

5. The medium frequency furnace sintering temperature control system according to claim 4, characterized in that: The data preprocessing module is also used for: Calculating the thermal conductivity of the corrected first temperature field distribution data and the corrected second thermal radiation feedback data in a historical sintering cycle; Predicting an expected temperature value of the corrected second thermal radiation feedback data in the real-time sintering cycle according to the heat transfer coefficient and the heating parameters of the corrected first temperature field distribution data in the real-time sintering cycle; Target heat flux compensation data is generated based on the corrected second thermal radiation feedback data and its expected temperature value, and a monitoring sequence corresponding to the target heat flux compensation data is used as a first correction sequence.

6. The medium frequency furnace sintering temperature control system according to claim 1, characterized in that: The compensation correction layer specifically includes: a heat conduction analysis unit, configured to perform heat flow path tracing for each monitoring sequence in the heat flow coupling characteristic data, so as to extract a corresponding thermal gradient propagation chain from each monitoring sequence; The temperature field matching unit is used to spatially map the thermal gradient propagation chain extracted from each monitoring sequence with the corresponding thermal flow coupling characteristic data to generate compensated temperature field data.

7. The medium frequency furnace sintering temperature control system according to claim 6, characterized in that: The compensation correction layer further includes: The thermal lag elimination unit is used to perform thermal response delay correction processing on the compensated temperature field data.

8. The medium frequency furnace sintering temperature control system according to claim 1, characterized in that: The temperature decision layer specifically includes: A multi-dimensional integration unit, comprising a plurality of temperature control nodes, each temperature control node being connected to each monitoring sequence in the compensated temperature field data and the heat-flow coupling characteristic data through an associated configuration; a dynamic weight optimization unit, configured to iteratively adjust the associated configuration using a dynamic weight optimization algorithm to minimize the deviation between the sintering temperature control instruction and the actual temperature distribution; The thermal imbalance identification unit is used to locate the sintering abnormality area based on the compensated temperature field data and the heat-flow coupling characteristic data, and generate a sintering temperature control instruction.

9. The medium frequency furnace sintering temperature control system according to claim 4, characterized in that: The dynamic weight adjustment method specifically includes: Generate adaptive correction parameters based on the thermal field distribution characteristics of the real-time sintering environment; A sliding window mechanism is used to perform segmented correction processing on the standardized first temperature field data.

Images