Dynamic Temperature Control System Based on Temperature Control Algorithm of Multistage Series Electric Heating Furnace

Abstract

The present disclosure relates to a dynamic temperature control system based on a temperature control algorithm for a multi-stage series electric heating furnace, and relates to the technical field of electric heating furnaces. The system includes: obtaining the basic design information and position serial number identification of the K-level electric heaters of the target electric heating furnace, identifying a multi-stage temperature control scheme, generating a multi-stage temperature control scheme, generating K horizontal dynamic temperature control constraints and K vertical dynamic temperature control constraints, obtaining the central monitoring temperatures of the K electric heaters and K real-time temperature rise values, performing double-constraint optimization on the K allocated powers to obtain K target allocated powers, and performing dynamic temperature control. The present disclosure solves the technical problems of low temperature control reliability and low dynamic temperature control accuracy in the prior art for multi-stage series electric heating furnaces, and achieves the technical effect of improving temperature control reliability.

Description

Technical Field

[0001] The present disclosure relates to the technical field of electric heating furnaces, and particularly to a dynamic temperature control system based on a temperature control algorithm for a multi-stage series electric heating furnace. Background Art

[0002] During the production process, when the temperature sensor of a certain stage of electric heater in a multi-stage series electric heating furnace fails, the actual temperature of the heater cannot be monitored in a timely manner. When the temperature is much higher than the preset value, it is easy to cause overheating damage to the metal material, and then trigger the safety interlock protection system, resulting in the shutdown of the entire production line for several hours for maintenance. At the same time, due to the thermal coupling effect between the electric heaters of each stage, when the temperature control of a certain stage of electric heater has a deviation, it will directly affect the subsequent electric heaters of each stage, making the temperature control of the entire system complex and difficult to stabilize. The prior art has the technical problems of low temperature control reliability and low dynamic temperature control accuracy for multi-stage series electric heating furnaces. Summary of the Invention

[0003] The present disclosure provides a dynamic temperature control system based on a temperature control algorithm for a multi-stage series electric heating furnace, which is used to solve the technical problems of low temperature control reliability and low dynamic temperature control accuracy for multi-stage series electric heating furnaces in the prior art.

[0004] In view of the above problems, the present disclosure provides a dynamic temperature control system based on a temperature control algorithm for a multi-stage series electric heating furnace, and the system includes:

[0005] A basic design information acquisition module, which is used to obtain the basic design information and position serial number identification of the K-stage electric heaters of the target electric heating furnace;

[0006] A multi-stage temperature control scheme generation module, which is used to obtain a preset output temperature threshold, and identify a multi-stage temperature control scheme by combining the basic design information and the position serial number identification, and generate a multi-stage temperature control scheme;

[0007] A dynamic temperature control constraint setting module, which is used to generate K horizontal dynamic temperature control constraints and K vertical dynamic temperature control constraints according to the multi-stage temperature control scheme;

[0008] A real-time temperature rise value acquisition module, which is used to monitor the inlet and outlet temperatures and the surface multi-point temperatures of the K-stage electric heaters in real time, and obtain K central monitoring temperatures of the electric heaters and K real-time temperature rise values;

[0009] A target allocation power acquisition module, which is used to perform double-constraint optimization on the K allocation powers in the multi-stage temperature control scheme based on the K central monitoring temperatures of the electric heaters and the K real-time temperature rise values, and obtain K target allocation powers;

[0010] A dynamic control module, which is used to transmit the K target allocation powers to a dynamic adjustment unit to complete power reallocation and dynamic temperature control.

[0011] One or more technical solutions provided in the present disclosure have at least the following technical effects or advantages:

[0012] In the present disclosure, by obtaining the basic design information and position serial number identification of the K-level electric heaters of the target electric heating furnace, then obtaining the preset output temperature threshold, identifying a multi-level temperature control scheme by combining the basic design information and the position serial number identification, generating a multi-level temperature control scheme, and then generating K horizontal dynamic temperature control constraints and K vertical dynamic temperature control constraints according to the multi-level temperature control scheme, and then performing real-time monitoring on the inlet and outlet temperatures and multi-point surface temperatures of the K-level electric heaters to obtain the central monitoring temperatures of the K electric heaters and the K real-time temperature rise values, performing double-constraint optimization on the K allocated powers in the multi-level temperature control scheme based on the central monitoring temperatures of the K electric heaters and the K real-time temperature rise values to obtain the K target allocated powers, and then transmitting the K target allocated powers to the dynamic adjustment unit to complete power reallocation and dynamic temperature control. The technical effect of improving the temperature control reliability of the multi-level series electric heating furnace is achieved. Description of the Drawings

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present disclosure. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0014] Figure 1 Schematic structural diagram of a dynamic temperature control system based on a multi-level series electric heating furnace temperature control algorithm provided by an embodiment of the present disclosure;

[0015] Figure 2 Schematic flow diagram of a dynamic temperature control method based on a multi-level series electric heating furnace temperature control algorithm provided by an embodiment of the present disclosure.

[0016] Explanation of reference numerals: Basic design information acquisition module 11, multi-level temperature control scheme generation module 12, dynamic temperature control constraint setting module 13, real-time temperature rise value acquisition module 14, target allocated power acquisition module 15, dynamic control module 16. Detailed Embodiments

[0017] The present disclosure provides a dynamic temperature control system based on a multi-level series electric heating furnace temperature control algorithm, which is used to solve the technical problems of low temperature control reliability and low dynamic temperature control accuracy in the prior art for multi-level series electric heating furnaces.

[0018] Next, the technical solutions in the embodiments of the present disclosure will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present disclosure without creative efforts shall fall within the protection scope of the present disclosure.

[0019] It should be noted that the terms "including" and "having" are intended to cover non-exclusive inclusion. For example, a process, system, product, or server that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, systems, products, or devices.

[0020] An embodiment is as Figure 1 shown. The present disclosure provides a dynamic temperature control system based on a multi-stage series electric heating furnace temperature control algorithm for executing a dynamic temperature control method based on a multi-stage series electric heating furnace temperature control algorithm as Figure 2 shown. The system includes:

[0021] A basic design information acquisition module 11 for obtaining the basic design information and position serial number identification of the K-level electric heaters of the target electric heating furnace;

[0022] In a possible embodiment, the target electric heating furnace is a multi-stage series electric heating furnace including K-level electric heating furnaces. The basic design information acquisition module 11 is used to extract K basic design information of the K-level electric heating furnaces from the structural design information of the target electric heating furnace, providing data support for subsequent temperature control. Preferably, the target temperatures to be heated by each level of electric heating furnace are different, so the models of each level of electric heating furnace are different, and each has a corresponding basic design information. Among them, the K basic design information reflects the structural design situation of the K-level electric heating furnaces, including the rated power, heating area, etc. of the electric heaters.

[0023] In an embodiment, each level of electric heater is an independent heating unit, which may have different characteristics such as power, size, heating efficiency, etc. Each electric heater has a unique position serial number identification in the electric heating furnace, which is used to distinguish and identify different electric heaters, helping the system accurately obtain the specific position of each electric heater in the multi-stage series electric heating furnace.

[0024] A multi-level temperature control scheme generation module 12 for obtaining a preset output temperature threshold, identifying a multi-level temperature control scheme by combining the basic design information and the position serial number identification, and generating a multi-level temperature control scheme;

[0025] Furthermore, the steps executed by the multi-level temperature control scheme generation module 12 further include:

[0026] Obtain multiple sample preset output temperature thresholds, K sample basic design information, K sample position serial number identifiers, and multiple sample multi-level temperature control schemes as training data;

[0027] Supervisedly train the encoder and the decoder based on the training data, and continuously update the network parameters of the encoder and the decoder during the training until the preset number of iterations is satisfied, to obtain a trained temperature control scheme recognizer;

[0028] Use the temperature control scheme recognizer to recognize the multi-level temperature control scheme for the preset output temperature threshold, basic design information, and the position serial number identifier, and generate the multi-level temperature control scheme.

[0029] In one embodiment, extract the preset output temperature threshold according to the heating task of the target electric heating furnace. Wherein, the preset output temperature threshold is the output temperature range of the target electric heating furnace. Furthermore, based on the preset output temperature threshold, K basic design information, and the K position serial number identifiers, use the temperature control scheme recognizer to determine the heating temperature control scheme for the K-level electric heaters in the target electric heating furnace, to obtain the multi-level temperature control scheme. Wherein, the multi-level temperature control scheme is a scheme that defines the power allocated to the K-level electric heaters and the heating conditions.

[0030] Optionally, obtain multiple sample preset output temperature thresholds, K sample basic design information, K sample position serial number identifiers, and multiple sample multi-level temperature control schemes as training data. Furthermore, construct an encoder (Encoder) for encoding the input data (preset output temperature threshold, basic design information, and position serial number identifier) into an intermediate representation, and a decoder (Decoder) for decoding the multi-level temperature control scheme (including K temperature rise thresholds, K temperature thresholds, and K allocated powers) from this intermediate representation. Use the prepared training data to supervise the training of the encoder and the decoder. During the training process, the encoder converts the input data into an internal representation, and the decoder attempts to recover the multi-level temperature control scheme from this representation. By comparing the multi-level temperature control scheme output by the decoder with the true multi-level temperature control scheme in the training data, calculate the loss function (such as mean square error, cross entropy, etc.), and update the network parameters of the encoder and the decoder based on the gradient of the loss function. The training process will iterate continuously, and each iteration will adjust the network parameters according to the value of the current loss function to reduce the prediction error. The iteration will continue until the preset number of iterations is satisfied or other stopping conditions are reached (such as the value of the loss function no longer decreases significantly). After sufficient iterations, the encoder and the decoder will learn how to extract effective information from the input data and generate a reasonable multi-level temperature control scheme. At this time, a trained temperature control scheme recognizer is obtained.

[0031] Furthermore, the trained temperature control scheme recognizer is used to process the preset output temperature threshold, K basic design information, and the K position sequence number identifiers. The input data is encoded into an intermediate representation by an encoder, and then a multi-level temperature control scheme is decoded from this representation by a decoder. This scheme includes the temperature rise threshold, temperature threshold, and allocated power for each level of the electric heater. By combining the supervised training method of deep learning, it is realized to automatically learn how to generate a reasonable multi-level temperature control scheme from the input data. The technical effects of improving the accuracy and efficiency of the temperature control scheme and reducing manual intervention and trial-and-error costs are achieved.

[0032] The dynamic temperature control constraint setting module 13 is used to generate K horizontal dynamic temperature control constraints and K vertical dynamic temperature control constraints according to the multi-level temperature control scheme;

[0033] The real-time temperature rise value obtaining module 14 is used to perform real-time monitoring on the inlet and outlet temperatures and the multi-point surface temperatures of the K-level electric heaters, and obtain the central monitoring temperatures of the K electric heaters and the K real-time temperature rise values;

[0034] In a possible embodiment, the K temperature rise thresholds are respectively used as K horizontal dynamic temperature control constraints, realizing the goal of horizontally constraining the temperature control of the target electric heating furnace from the temperature rise ranges that the K-level electric heaters can heat in sequence. Furthermore, the K temperature thresholds are used as K vertical dynamic temperature control constraints, realizing the goal of vertically constraining from the ranges that the heating temperatures of the K-level electric heaters can meet during each level of heating.

[0035] In a possible embodiment, temperature sensors are used to perform real-time temperature monitoring at the inlets and outlets of the K-level electric heaters and at the K initial surface monitoring point arrays, obtaining the K real-time surface monitoring temperature sets, the K real-time outlet temperatures, and the K real-time inlet temperatures. And, by further processing the K real-time surface monitoring temperature sets (such as weighted average, interpolation calculation, etc.), the central monitoring temperature of each level of electric heater is determined. These central monitoring temperatures will be used for subsequent temperature control. Among them, the central monitoring temperatures of the K electric heaters are temperature values reflecting the actual temperature conditions of the K-level electric heaters.

[0036] Furthermore, the steps executed by the real-time temperature rise value obtaining module 14 further include:

[0037] Obtain the preset number of initial surface monitoring points;

[0038] Based on the K basic design information, extract K surface areas, identify the monitoring point layout intervals according to the K surface areas and the preset number of initial surface monitoring points, and perform initial surface monitoring point layout according to the identification results to obtain the K initial surface monitoring point arrays;

[0039] In a preset monitoring window, continuously monitor the temperature of the K-level electric heaters according to the K initial surface monitoring point arrays, and obtain K initial temperature monitoring sequence sets;

[0040] Calculate the temperature gradients of the K initial temperature monitoring sequence sets respectively to obtain K temperature gradient sets;

[0041] Based on the K temperature gradient sets, screen the monitoring points to determine K surface monitoring point arrays;

[0042] Perform real-time multi-point surface temperature monitoring according to the K surface monitoring point arrays, generate K real-time surface monitoring temperature sets, and identify the central monitoring temperatures of the K electric heaters from the K real-time surface monitoring temperature sets.

[0043] Furthermore, the steps performed by the real-time temperature rise value obtaining module 14 further include:

[0044] Extract K minimum temperature gradients from the K temperature gradient sets respectively, and use the K minimum temperature gradients as simulated irrigation points for water injection. As the water surface rises, until it reaches the preset temperature gradient, obtain K first water surfaces, and use the multiple temperature gradients submerged by the K first water surfaces as K first temperature gradient sets, where the K temperature gradient sets include K initial surface monitoring point positioning identifiers;

[0045] Perform nearest neighbor fusion on the K first temperature gradient sets to generate K first region segmentation ridge lines, where the K first region segmentation ridge lines rise as the water surface rises;

[0046] Continue to inject water until the rising amplitude of the water surface reaches the preset temperature gradient to obtain K second water surfaces, and use the multiple temperature gradients submerged by the K second water surfaces as K second temperature gradient sets;

[0047] Perform nearest neighbor fusion on the K second temperature gradient sets to generate K second region segmentation ridge lines, where the K second region segmentation ridge lines rise as the water surface rises;

[0048] After multiple nearest neighbor fusions, until the water surface submerges the K maximum temperature gradients in the K temperature gradient sets, obtain K region segmentation ridge line sets;

[0049] Based on the K region segmentation ridge line sets, screen the K temperature gradient sets to obtain the K surface monitoring point arrays.

[0050] Furthermore, the steps performed by the real-time temperature rise value obtaining module 14 further include:

[0051] Use the set of ridgelines divided by the K regions to divide the K sets of temperature gradients, generating K sets of temperature gradient divided regions;

[0052] Calculate the average temperature gradient within each of the K sets of temperature gradient divided regions respectively, obtaining K sets of divided region means;

[0053] Calculate the ratio of each divided region mean in the K sets of divided region means to the sum of the means of the corresponding set of divided region means respectively, and take the calculation result as the distribution coefficient of the monitoring points in the divided regions, obtaining K sets of distribution coefficients of the monitoring points in the divided regions;

[0054] Obtain the preset number of real-time surface monitoring points, multiply each of the K sets of distribution coefficients of the monitoring points in the divided regions by the preset number of real-time surface monitoring points respectively, obtaining K sets of real-time surface monitoring point numbers in the divided regions;

[0055] Based on the K sets of real-time surface monitoring point numbers in the divided regions, randomly extract monitoring points in the K sets of temperature gradient divided regions, generating the K surface monitoring point arrays.

[0056] In one embodiment, according to the experience of those skilled in the art or design requirements, preset an initial number of surface monitoring points for arranging monitoring points on the surface of the electric heater, that is, the preset initial number of surface monitoring points. Furthermore, based on K basic design information (such as the size, shape, heating element layout, etc. of the electric heater), extract K surface areas, and calculate the layout interval of the monitoring points according to the preset initial number of surface monitoring points and the surface area. Optionally, divide the surface area by the preset initial number of surface monitoring points to obtain the monitoring area area of each monitoring point, set the monitoring area as a square, and obtain the side length of the area according to the monitoring area area, that is, the monitoring point layout interval. According to the identified monitoring point layout interval, arrange the initial monitoring points on the surface of the electric heater to form K initial surface monitoring point arrays.

[0057] Within a preset monitoring window (a continuously monitored time period preset by those skilled in the art), use temperature sensors to continuously monitor the temperatures of the K initial surface monitoring point arrays respectively, record the temperature data of each monitoring point, and form K sets of initial temperature monitoring sequences. Among them, the K sets of initial temperature monitoring sequences reflect the monitoring temperature change situations of each initial surface monitoring point in the K initial surface monitoring point arrays within the preset monitoring window.

[0058] Calculate the temperature gradients of the K sets of initial temperature monitoring sequences respectively to evaluate the change trend and distribution of the surface temperature of the electric heater. The set of temperature gradients reflects the spatial distribution characteristics of the surface temperature of the electric heater.

[0059] Furthermore, based on the K sets of temperature gradients, the initial surface monitoring points are screened, and those points that can more accurately reflect the central temperature or representative temperature of the electric heater are retained to form an array of K surface monitoring points. The optimized array of K surface monitoring points is used to perform real-time monitoring of the surface multi-point temperature of the electric heater, generating K sets of real-time surface monitoring temperatures. By further processing the K sets of real-time surface monitoring temperatures (such as weighted averaging, interpolation calculation, etc.), the central monitoring temperature of each stage of the electric heater is identified. These central monitoring temperatures will be used for subsequent temperature control.

[0060] In one embodiment, since different mountains in the mountainous area have different heights, when water infiltrates from the lowest point of the mountainous area, two different parts will appear in the mountainous area. One part is the water-collecting area, and the other part is the dividing ridge line. The dividing ridge line is used to separate the mountains with water from the mountains without water infiltration. And as continuous water injection continues and the water surface rises, the dividing ridge line will also rise until all mountainous areas are submerged by the water surface. Thus, the goal of separating mountainous areas with different heights is achieved.

[0061] Simulating the temperature gradient as the mountain height, then based on the K sets of temperature gradients, K mountainous areas can be simulated, and there are multiple temperature gradient points in each area. The temperature gradient value of each point can correspond to a certain height on the mountain. The temperature gradient reflects the temperature change situation of each stage of the electric heater at a certain monitoring point. Exemplarily, the temperature gradient can be 5°C, 10°C, etc., and the temperature gradient sets within each mountainous area vary from 0°C to 50°C. Therefore, the mountain height within each mountainous area can range from the flat ground (0°C) to the highest point (50°C). Thus, different mountain forms of the K mountainous areas can be obtained according to the K sets of temperature gradients.

[0062] Irrigate from the lowest points of the K mountainous regions (i.e., the lowest points of the valleys or mountainous regions). For example, the temperature gradient corresponding to the lowest point of the first mountainous region is 0°C, the temperature gradient corresponding to the lowest point of the second mountainous region is 5°C, etc. As continuous irrigation progresses, the water level will continuously rise until it submerges the highest points of the mountainous regions. Therefore, extract the minimum values of K temperature gradients from the K sets of temperature gradients respectively, and use these minimum values of the K temperature gradients as the simulated irrigation points for water injection. As the water level rises, until it rises to a preset temperature gradient (the maximum temperature gradient that those skilled in the art can set by themselves when it can be divided into the same area as the minimum values of the K temperature gradients, such as 15°C), K first water surfaces are obtained. Take the multiple temperature gradients submerged by the K first water surfaces as the K first sets of temperature gradients. That is, summarize all the temperature gradients between the temperature gradient at the lowest point and the temperature gradient after adding 15°C to the temperature gradient at the lowest point, so as to obtain the K first sets of temperature gradients. Among them, the K sets of temperature gradients include K positioning identifiers for initial surface monitoring points. Perform nearest neighbor fusion on the K first sets of temperature gradients to generate K first regional segmentation ridge lines. Among them, the K first regional segmentation ridge lines rise as the water level rises. Optionally, identify adjacent temperature gradients of the K first sets of temperature gradients according to the K positioning identifiers for initial surface monitoring points, fuse the adjacent temperature gradients into one area, and then take the edge of each area as part of the segmentation ridge line to obtain the K first regional segmentation ridge lines.

[0063] Continue to inject water until the rising amplitude of the water level reaches the preset temperature gradient to obtain K second water surfaces. Take the multiple temperature gradients submerged by the K second water surfaces as the K second sets of temperature gradients. Based on the same principle as the K first regional segmentation ridge lines, perform nearest neighbor fusion on the K second sets of temperature gradients to generate K second regional segmentation ridge lines. Among them, the K second regional segmentation ridge lines rise as the water level rises.

[0064] After multiple nearest neighbor fusions, until the water level submerges the maximum values of the K temperature gradients in the K sets of temperature gradients, a set of K regional segmentation ridge lines is obtained. Among them, the set of K regional segmentation ridge lines divides the K sets of temperature gradients into multiple different areas, and the ranges of temperature gradients within each area are different. Based on the set of K regional segmentation ridge lines, screen the monitoring points of the K sets of temperature gradients to obtain the K surface monitoring point arrays.

[0065] In a possible embodiment, the K temperature gradient sets are partitioned by using the ridge line set partitioned into K regions to generate a set of K temperature gradient partition regions. The temperature gradient ranges within each temperature gradient partition region are different. For example, in a set of temperature gradient partition regions, the temperature gradient in a certain temperature gradient partition region is between 5°C and 20°C, and the temperature gradients in the remaining temperature gradient partition regions are between 21°C and 35°C, between 36°C and 50°C, etc. Furthermore, the average value of the temperature gradients within each of the K temperature gradient partition regions is calculated respectively to obtain a set of K partition region average values. For example, if the temperature gradient in a temperature gradient partition region is between 5°C and 20°C and includes 10 temperature gradients, which are 5°C, 15°C, 6°C, 9°C, 12°C, 16°C, 11°C, 15°C, 13°C, and 11°C respectively, then the average value of the partition region is 11.3°C. Among them, the set of K partition region average values reflects the average temperature gradient situation of each temperature gradient partition region in the set of K temperature gradient partition regions.

[0066] The ratio of each partition region average value in the set of K partition region average values to the sum of the average values of the corresponding partition region average value set is calculated respectively, and the calculation result is used as the distribution coefficient of the monitoring points in the partition region to obtain a set of K distribution coefficients of the monitoring points in the partition region. Each distribution coefficient of the monitoring points in the partition region reflects the importance of the temperature fluctuation degree of each partition region to the electric heater. The larger the coefficient, the higher the importance.

[0067] The preset number of real-time surface monitoring points (the number of surface monitoring points set in advance by those skilled in the art during real-time monitoring) is obtained. The set of K distribution coefficients of the monitoring points in the partition region is multiplied by the preset number of real-time surface monitoring points respectively to obtain a set of K real-time surface monitoring point numbers in the partition region. Furthermore, based on the set of K real-time surface monitoring point numbers in the partition region, monitoring points are randomly selected in the set of K temperature gradient partition regions to generate the K surface monitoring point arrays. The technical effect of reliably setting the surface monitoring points according to the real-time temperature gradient distribution of each stage of the electric heater is achieved.

[0068] In an embodiment, by calculating the difference between the real-time outlet temperature and the real-time inlet temperature of each electric heater, K real-time temperature rise values are obtained. These temperature rise values reflect the current heating efficiency or effect of the electric heater. The goal of providing reliable data support for subsequent double-constraint optimization is achieved.

[0069] The target assigned power obtaining module 15 is configured to perform double-constraint optimization on the K assigned powers in the multi-stage temperature control scheme based on the K central monitoring temperatures of the electric heaters and the K real-time temperature rise values to obtain K target assigned powers;

[0070] In a possible embodiment, by using the target power allocation obtaining module 15, it is ensured that the power allocated by the system to the K-level electric heaters of the target electric heating furnace can simultaneously satisfy the dynamic temperature control constraints in the horizontal direction (i.e., temperature control between different electric heaters) and the vertical direction (i.e., temperature control inside a single electric heater).

[0071] Optionally, in the double-constraint optimization, the K real-time temperature rise values of the K-level electric heaters during the optimization process need to satisfy the K horizontal dynamic temperature control constraints, and the monitored temperatures at the centers of the K electric heaters need to satisfy the K vertical dynamic temperature control constraints. That is to say, when optimizing the K allocated powers based on the monitored temperatures at the centers of the K electric heaters and the K real-time temperature rise values, the working effects of the K-level electric heaters for power reallocation need to satisfy the K horizontal dynamic temperature control constraints and the K vertical dynamic temperature control constraints. The technical effect of improving the reliability of dynamic temperature control is achieved.

[0072] Furthermore, the steps executed by the target power allocation obtaining module 15 further include:

[0073] Using the K vertical dynamic temperature control constraints to perform constraint identification on the monitored temperatures at the centers of the K electric heaters, and determining K vertical deviation factors and K vertical deviation directions;

[0074] Using the K horizontal dynamic temperature control constraints to perform constraint identification on the K real-time temperature rise values, and determining K horizontal deviation factors and K horizontal deviation directions;

[0075] Based on the K horizontal deviation factors, K horizontal deviation directions, K vertical deviation factors, and K vertical deviation directions, perform double-constraint optimization on the K allocated powers to obtain the K target allocated powers, where in the double-constraint optimization, the K real-time temperature rise values of the K-level electric heaters during the optimization process need to satisfy the K horizontal dynamic temperature control constraints, and the monitored temperatures at the centers of the K electric heaters need to satisfy the K vertical dynamic temperature control constraints.

[0076] Furthermore, the steps executed by the target power allocation obtaining module 15 further include:

[0077] Using the power allocation adjustment unit to identify the K horizontal deviation factors, K horizontal deviation directions, K vertical deviation factors, K vertical deviation directions, and the K allocated powers, and determining K adjusted allocated powers;

[0078] Based on the K adjusted allocated powers, perform temperature control simulation on the K-level electric heaters, and determine whether the K temperature control simulation results satisfy the K horizontal dynamic temperature control constraints and the K vertical dynamic temperature control constraints;

[0079] If so, use the K adjusted allocated powers as the K target allocated powers.

[0080] Further, the steps performed by the target allocated power obtaining module 15 further include:

[0081] If not, perform deviation identification based on the K temperature control simulation results, K lateral dynamic temperature control constraints, and K longitudinal dynamic temperature control constraints to obtain K simulated lateral deviation factors, K simulated lateral deviation directions, K simulated longitudinal deviation factors, and K simulated longitudinal deviation directions;

[0082] Use the allocated power adjustment unit to identify the K simulated lateral deviation factors, K simulated lateral deviation directions, K simulated longitudinal deviation factors, K simulated longitudinal deviation directions, and the K adjusted allocated powers to determine K stage allocated powers;

[0083] When the temperature control simulation results of the K stage allocated powers satisfy the K lateral dynamic temperature control constraints and the K longitudinal dynamic temperature control constraints, use the K stage allocated powers as the K target allocated powers.

[0084] In a possible embodiment, use the K longitudinal dynamic temperature control constraints to perform constraint identification on the K monitored temperatures at the centers of the electric heaters to determine K lateral deviation factors and K longitudinal deviation directions. That is, calculate the differences between the K monitored temperatures at the centers of the electric heaters and the K temperature thresholds, and use the calculation results as the K longitudinal deviation factors. When the K monitored temperatures at the centers of the electric heaters are greater than the K temperature thresholds, the K longitudinal deviation directions are positive; when the K monitored temperatures at the centers of the electric heaters are less than or equal to the K temperature thresholds, the K longitudinal deviation directions are negative.

[0085] Based on the same principle, use the K lateral dynamic temperature control constraints to perform constraint identification on the K real-time temperature rise values to determine K lateral deviation factors and K lateral deviation directions. That is, calculate the differences between the K real-time temperature rise values and the K temperature rise thresholds, and use the calculation results as the K lateral deviation factors. When the K real-time temperature rise values are greater than the K temperature rise thresholds, the K lateral deviation directions are positive; when the K real-time temperature rise values are less than or equal to the K temperature rise thresholds, the K lateral deviation directions are negative. The goal of providing an optimization scale and an optimization direction for subsequent optimization is achieved.

[0086] Use the allocated power adjustment unit to identify the K lateral deviation factors, K lateral deviation directions, K longitudinal deviation factors, K longitudinal deviation directions, and the K allocated powers to determine K adjusted allocated powers. Among them, the allocated power adjustment unit is used to intelligently adjust and optimize the K allocated powers and output the K adjusted allocated powers.

[0087] Furthermore, temperature control simulation is performed on the K-level electric heaters using the K adjusted allocated powers to determine the heating conditions of the K-level electric heaters at the K adjusted allocated powers, and K temperature control simulation results are obtained. Exemplarily, a temperature control simulation environment is set up for the K-level electric heaters through COMSOL Multiphysics or MATLAB, and the temperature response of the electric heaters can be simulated according to the given power inputs (i.e., the K adjusted allocated powers). The calculated K adjusted allocated powers are used as inputs and respectively allocated to the K-level electric heaters. In the simulation environment, each electric heater will simulate the temperature change according to the allocated power. In the simulation environment, the temperature control simulation process is started. This process will calculate and simulate the temperature change of each electric heater over time based on the physical characteristics of the electric heater (such as heat capacity, thermal conductivity, etc.) and the input power. Thus, the temperature control simulation results are obtained.

[0088] After the temperature control simulation is completed, the simulated temperature data of each electric heater are collected and recorded. These data may include the time series of temperature, the central monitored temperature, the real-time temperature rise value, etc. Among them, the K temperature control simulation results include K sets of simulated surface monitored temperatures, K simulated outlet temperatures, and K simulated inlet temperatures. Based on the same analysis principle as above, K simulated temperature rise values and K simulated central monitored temperatures of the electric heaters are obtained. Furthermore, it is judged whether the K simulated temperature rise values and the K simulated central monitored temperatures of the electric heaters meet the K lateral dynamic temperature control constraints and the K longitudinal dynamic temperature control constraints. When they are met, the K adjusted allocated powers are used as the K target allocated powers.

[0089] If not, deviation identification is performed according to the K temperature control simulation results, the K lateral dynamic temperature control constraints, and the K longitudinal dynamic temperature control constraints to obtain K simulated lateral deviation factors, K simulated lateral deviation directions, K simulated longitudinal deviation factors, and K simulated longitudinal deviation directions. Optionally, the differences between the K simulated central monitored temperatures of the K electric heaters in the K temperature control simulation results and the K temperature thresholds are calculated, and the calculation results are used as the K simulated longitudinal deviation factors. When the K simulated central monitored temperatures of the K electric heaters are greater than the K temperature thresholds, the K simulated longitudinal deviation directions are positive; when the K simulated central monitored temperatures of the K electric heaters are less than or equal to the K temperature thresholds, the K simulated longitudinal deviation directions are negative.

[0090] Optionally, the differences between the K simulated real-time temperature rise values and the K temperature rise thresholds are calculated, and the calculation results are used as the K simulated lateral deviation factors. When the K simulated real-time temperature rise values are greater than the K temperature rise thresholds, the K simulated lateral deviation directions are positive; when the K simulated real-time temperature rise values are less than or equal to the K temperature rise thresholds, the K simulated lateral deviation directions are negative.

[0091] The K analog lateral deviation factors, K analog lateral deviation directions, K analog longitudinal deviation factors, K analog longitudinal deviation directions, and the K adjusted allocation powers are identified by using the allocation power adjustment unit to determine the K-stage allocation powers. When the temperature control simulation results of the K-stage allocation powers satisfy the K lateral dynamic temperature control constraints and the K longitudinal dynamic temperature control constraints, the K-stage allocation powers are used as the K target allocation powers.

[0092] Optionally, a plurality of sample lateral deviation factor sets, a plurality of sample lateral deviation direction sets, a plurality of sample longitudinal deviation factor sets, a plurality of sample longitudinal deviation direction sets, a plurality of sample allocation power sets, and a plurality of sample adjusted allocation power sets are obtained as training data, and a framework constructed based on a convolutional neural network is supervised and trained until the output converges to obtain the trained allocation power adjustment unit.

[0093] Preferably, the allocation power adjustment unit includes an input layer, a convolutional layer, a pooling layer, and a fully connected layer. The loss function is used to perform loss analysis on the training process of the allocation power adjustment unit, where the loss function is , where is the sample allocation power, is the predicted allocation power output during the training of the allocation power adjustment unit, and N is the number of training samples. When the loss function reaches the minimum value, the training converges to obtain the allocation power adjustment unit.

[0094] The dynamic control module 16 is configured to transmit the K target allocation powers to the dynamic adjustment unit for power reallocation to perform dynamic temperature control.

[0095] Optionally, the dynamic adjustment unit is configured to reallocate the powers of the K-level electric heaters according to the K target allocation powers. After the allocation is completed, the heating conditions of the K-level electric heaters can meet the preset output temperature threshold, the temperature rise thresholds of each level, and the temperature threshold of the target electric heating furnace. The technical effect of improving the temperature control accuracy of the target electric heating furnace is achieved.

[0096] In summary, the embodiments of the present disclosure have at least the following technical effects:

[0097] The present disclosure obtains the basic design information and position serial number identification of the K-level electric heaters of the target electric heating furnace, then obtains the preset output temperature threshold, identifies a multi-level temperature control scheme by combining the basic design information and the position serial number identification, generates a multi-level temperature control scheme, and then generates K horizontal dynamic temperature control constraints and K vertical dynamic temperature control constraints according to the multi-level temperature control scheme. Then, real-time monitoring of the inlet and outlet temperatures and the multi-point surface temperatures of the K-level electric heaters is carried out to obtain the central monitoring temperatures of the K electric heaters and the K real-time temperature rise values. Based on the central monitoring temperatures of the K electric heaters and the K real-time temperature rise values, double-constraint optimization of the K allocated powers in the multi-level temperature control scheme is performed to obtain the K target allocated powers, and then the K target allocated powers are transmitted to the dynamic adjustment unit to complete power reallocation and dynamic temperature control. The technical effect of improving the temperature control reliability of the multi-level series electric heating furnace is achieved.

[0098] It should be noted that the above sequence of embodiments of the present disclosure is only for description and does not represent the superiority or inferiority of the embodiments. And the above description of specific embodiments of this specification has been made. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0099] The above are only the preferred embodiments of the present disclosure and are not intended to limit the present disclosure. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present disclosure shall be included within the protection scope of the present disclosure.

[0100] This specification and the drawings are only exemplary descriptions of the present disclosure and are considered to have covered any and all modifications, variations, combinations, or equivalents within the scope of the present disclosure. Obviously, those skilled in the art can make various changes and modifications to the present disclosure without departing from the scope of the present disclosure. Thus, if these modifications and variations of the present disclosure fall within the scope of the present disclosure and its equivalent technologies, the present disclosure is intended to include these changes and modifications.

Claims

1. A dynamic temperature control system based on a temperature control algorithm for a multi-stage series electric heating furnace, characterized in that: The system comprises: A basic design information acquisition module is used to obtain the basic design information and position serial number identification of the K-class electric heater of the target electric heating furnace; A multi-level temperature control scheme generation module is used to obtain a preset output temperature threshold, identify a multi-level temperature control scheme based on basic design information and position serial number identification, and generate a multi-level temperature control scheme; The dynamic temperature control constraint setting module is used to generate K horizontal dynamic temperature control constraints and K vertical dynamic temperature control constraints according to the multi-level temperature control scheme. The horizontal direction refers to the temperature control between different electric heaters, and the vertical direction refers to the internal temperature control of a single electric heater. A real-time temperature rise value acquisition module is used to monitor the inlet and outlet temperatures and surface multi-point temperatures of the K-level electric heater in real time, and obtain the central monitoring temperatures of K electric heaters and K real-time temperature rise values; A target allocated power acquisition module is used to perform dual-constraint optimization on K allocated powers in a multi-stage temperature control scheme based on the K electric heater center monitoring temperatures and the K real-time temperature rise values ​​to obtain K target allocated powers; The dynamic control module is used to transmit the K target allocated powers to the dynamic adjustment unit to complete power redistribution and dynamic temperature control.

2. The dynamic temperature control system based on the temperature control algorithm of the multi-stage series electric heating furnace according to claim 1 is characterized in that: The steps performed by the real-time temperature rise value obtaining module also include: Arrange initial surface monitoring points to obtain K initial surface monitoring point arrays, perform continuous temperature monitoring and temperature gradient calculation within a preset monitoring window to obtain K temperature gradient sets, and perform monitoring point screening to determine K surface monitoring point arrays; Real-time monitoring of multi-point surface temperature is performed according to the K surface monitoring point arrays, K real-time surface monitoring temperature sets are generated, central monitoring temperature identification is performed, and central monitoring temperatures of the K electric heaters are obtained.

3. The dynamic temperature control system based on the temperature control algorithm of the multi-stage series electric heating furnace according to claim 2 is characterized in that: The steps performed by the real-time temperature rise value obtaining module also include: Extracting K temperature gradient minimum values ​​from the K temperature gradient sets respectively, performing watering simulation, and obtaining K first temperature gradient sets; Perform nearest neighbor fusion on the K first temperature gradient sets to generate K first region segmentation ridge lines; Continue to inject water, after multiple nearest neighbor fusions, until the K maximum temperature gradients in the K temperature gradient sets are submerged, obtain K regional segmentation ridge line sets, and screen the monitoring points of the K temperature gradient sets to obtain the K surface monitoring point arrays.

4. The dynamic temperature control system based on the temperature control algorithm of the multi-stage series electric heating furnace according to claim 3 is characterized in that: The steps performed by the real-time temperature rise value obtaining module also include: Using the K regional segmentation ridge line sets to divide the K temperature gradient sets into regions, and calculating the mean of the temperature gradient in the regions to obtain the mean value sets of the K divided regions, and obtaining the distribution coefficient sets of the monitoring points of the K divided regions based on the mean value sets of the K divided regions; Divide the number of monitoring points according to the preset number of real-time surface monitoring points and the K divided area monitoring point distribution coefficient sets to obtain the K divided area real-time surface monitoring point number sets; Monitoring points are randomly selected based on the number sets of real-time surface monitoring points in the K divided areas to generate the K surface monitoring point arrays.

5. The dynamic temperature control system based on the temperature control algorithm of the multi-stage series electric heating furnace according to claim 1 is characterized in that: The steps performed by the target allocated power acquisition module also include: K longitudinal dynamic temperature control constraints and K transverse dynamic temperature control constraints are used to identify the central monitoring temperatures of K electric heaters and K real-time temperature rise values, respectively, to determine the deviation factor and deviation direction; The K allocated powers are optimized with dual constraints based on the deviation factor and the deviation direction to obtain the K target allocated powers.

6. The dynamic temperature control system based on the temperature control algorithm of the multi-stage series electric heating furnace according to claim 5 is characterized in that: The steps performed by the target allocated power acquisition module also include: Using the allocated power adjustment unit to identify the deviation factor and the deviation direction, and determine K adjusted allocated powers, wherein the deviation factor includes K lateral deviation factors and K longitudinal deviation factors, and the deviation direction includes K lateral deviation directions and K longitudinal deviation directions; Performing temperature control simulation on the K-level electric heater based on the K adjusted allocated powers, and determining whether the K temperature control simulation results satisfy the K lateral dynamic temperature control constraints and the K longitudinal dynamic temperature control constraints; If so, the K adjusted allocated powers are used as the K target allocated powers.

7. The dynamic temperature control system based on the temperature control algorithm of the multi-stage series electric heating furnace according to claim 6 is characterized in that: The steps performed by the target allocated power acquisition module also include: If not, performing deviation identification according to the K temperature control simulation results and the K lateral dynamic temperature control constraints and the K longitudinal dynamic temperature control constraints to obtain a simulation deviation factor and a simulation deviation direction; Using the allocated power adjustment unit to identify the simulated deviation factor, the simulated deviation direction and the K adjusted allocated powers, and determine the K stage allocated powers; When the temperature control simulation results of the K stage allocated powers satisfy the K lateral dynamic temperature control constraints and the K longitudinal dynamic temperature control constraints, the K stage allocated powers are used as the K target allocated powers.

8. The dynamic temperature control system based on the temperature control algorithm of the multi-stage series electric heating furnace according to claim 1 is characterized in that: The steps performed by the multi-stage temperature control scheme generation module also include: A temperature control identifier is pre-built to identify a multi-level temperature control solution based on the preset output temperature threshold, basic design information and the position serial number identifier, and generate the multi-level temperature control solution.

Images