Intelligent control system for steel ladle baking

Abstract

The invention relates to the technical field of metal casting, and discloses a steel ladle baking intelligent control system, which comprises a waste gas monitoring unit for collecting humidity, temperature and flow parameters of waste gas in a discharge flue in real time; the feature extraction and stage recognition unit is used for calculating a baking entropy speed increase rate feature index based on the parameter coupling, and dynamically recognizing severe evaporation, evaporation slowing down, pure heating and soaking saturation stages; and the baking control unit sets a target entropy speed increasing rate interval according to the current stage, and homogenized drying and temperature rising are achieved in the steel ladle by adjusting heat output of the heating device. According to the method, the internal thermodynamic state of the steel ladle is dynamically analyzed through the entropy flow characteristics of the waste gas, self-adaptive precise regulation and control of the baking process are achieved, the problems of uneven baking and low energy efficiency caused by black box operation in a traditional method are solved, meanwhile, the penetrability verification capability on residual moisture is achieved, and the baking quality and the energy utilization efficiency are remarkably improved.

Description

Technical Field

[0001] The present invention relates to an intelligent control system for ladle baking, belonging to the technical field of metal casting. Background Art

[0002] In the ladle pretreatment process, traditional baking methods generally adopt a control strategy based on a fixed temperature-time curve or rely on the experience of operators to adjust the gas supply. The core logic is to indirectly infer the internal drying state by monitoring the surface temperature of the ladle.

[0003] However, this apparent parameter control mode has essential limitations in dealing with complex working conditions: 1. The humidity distribution, heat capacity characteristics, and structural differences of the refractory materials in the ladle lining make it difficult for a unified heat input strategy to adapt to individual needs, and significant deviations often occur in the drying processes of newly built ladles and turnover ladles; 2. The coupling effect of water evaporation and material heating is not effectively decoupled. Excessive heat input in the initial stage of baking is likely to cause surface overburning while there is residual moisture inside, and unbalanced heat energy distribution in the later stage leads to low soaking efficiency.

[0004] Although the industry has tried to introduce multi-point temperature monitoring and fuzzy control algorithms for optimization, it is still restricted by the black-box characteristics of the internal state perception of refractory materials and cannot capture the dynamic relationship between water phase change and heat transfer. Especially for ladles with complex wall thickness structures, it is difficult to completely remove the residual moisture under conventional positive pressure baking, and local steam pressure surges are likely to occur after entering the high-temperature stage, endangering the structural integrity of the refractory materials. Therefore, how to avoid the cognitive limitations of apparent parameter control, establish a precise regulation mechanism based on real-time feedback of the internal thermodynamic state of the ladle, and ensure the thoroughness of drying and energy consumption economy under complex working conditions has become a key technical challenge for improving the pretreatment quality of metal casting containers. Summary of the Invention

[0005] The present invention provides an intelligent control system for ladle baking, and its main purpose is to solve the problems of uneven baking, low energy efficiency, and incomplete detection of residual moisture caused by the inability of traditional baking methods to real-time sense the internal thermodynamic state of the ladle.

[0006] To achieve the above object, an intelligent control system for ladle baking provided by the present invention includes: An exhaust gas monitoring unit, configured in the smoke exhaust duct during ladle baking, for real-time collecting the humidity parameter, temperature parameter, and flow parameter of the exhaust gas; A feature extraction and stage recognition unit, connected to the exhaust gas monitoring unit, is configured to: based on the exhaust gas humidity parameter, the exhaust gas temperature parameter, and the exhaust gas flow parameter, calculate and obtain a baking entropy increase rate characteristic index representing the baking process inside the ladle through coupling; by continuously analyzing the temporal variation of the baking entropy increase rate characteristic index, dynamically identify at least one physical and chemical stage among the intense evaporation stage, the evaporation slowdown stage, the pure temperature increase stage, and the stage approaching thermal equilibrium and saturation in the current baking process. A baking control unit, connected to the feature extraction and stage recognition unit and connected to the heating device of the ladle, is configured to: set a corresponding target baking entropy increase rate interval according to the current physical and chemical stage identified by the feature extraction and stage recognition unit; dynamically adjust the total heat output of the heating device according to the deviation between the real-time monitored baking entropy increase rate characteristic index and the target baking entropy increase rate interval to guide the baking process and achieve homogenized drying and temperature increase inside the ladle.

[0007] Preferably, the exhaust gas monitoring unit includes a humidity sensor configured to monitor the exhaust gas humidity parameter, a temperature sensor configured to monitor the exhaust gas temperature parameter, and a flow sensor or a flow estimation module configured to monitor the exhaust gas flow parameter.

[0008] Preferably, the baking entropy increase rate characteristic index includes the moisture evaporation amount per unit time calculated through the exhaust gas humidity parameter, the exhaust gas temperature parameter, and the exhaust gas flow parameter. Among them, the moisture evaporation amount per unit time is higher than the preset initial moisture evaporation threshold value in the initial stage of baking, and shows a downward trend and forms a characteristic inflection point as the free water and part of the bound water in the refractory material are removed.

[0009] Preferably, the feature extraction and stage recognition unit is configured to determine that the baking process transitions from the intense evaporation stage or the evaporation slowdown stage mainly dominated by moisture evaporation to the pure temperature increase stage mainly dominated by the temperature increase of the refractory material matrix by detecting the downward inflection point of the moisture evaporation amount per unit time in time series.

[0010] Preferably, the baking control unit is configured to: in the intense evaporation stage, if the baking entropy increase rate characteristic index is lower than the lower limit of the target baking entropy increase rate interval, increase the heat output of the heating device; if the baking entropy increase rate characteristic index is higher than the upper limit of the target baking entropy increase rate interval, reduce the heat output of the heating device; in the pure temperature increase stage, adjust the heat output strategy according to the change of the exhaust gas temperature parameter to effectively increase the temperature inside the refractory material; when the baking entropy increase rate characteristic index tends to a preset stable value, indicating that the inside of the ladle tends to be in a thermal equilibrium and dry state, the baking control unit automatically switches to the heat preservation stage or the programmed cooling stage.

[0011] Preferably, the system further includes an auxiliary temperature monitoring unit configured to monitor the surface temperature or near-surface temperature of the ladle lining; when adjusting the total heat output of the heating device, the baking control unit refers to the feedback of the auxiliary temperature monitoring unit to ensure that the surface temperature or near-surface temperature of the ladle lining does not exceed a preset upper safety temperature limit.

[0012] Preferably, the baking control unit is configured to select different control strategies according to a preset baking target type, and the baking target type includes fully drying a newly built ladle, rapidly heating a turnover ladle, or energy-saving baking.

[0013] Preferably, the feature extraction and stage identification unit is configured to enable an enhanced discrimination mechanism for residual moisture evaporation response under transient negative pressure perturbation when it is determined that the ladle baking process reaches a preset end-of-drying stage or the initial stage of pure temperature rise; the exhaust gas monitoring unit is configured to monitor the transient change of the exhaust gas humidity parameter in the exhaust duct at a sampling frequency higher than the normal sampling frequency during the application of transient negative pressure perturbation; the feature extraction and stage identification unit is further configured to finally confirm whether the ladle has reached a completely dry state by judging whether the transient change of the exhaust gas humidity parameter presents a specific humidity pulse response indicating that there is residual moisture in the ladle accelerating evaporation under negative pressure induction; wherein, the detection logic of the specific humidity pulse response satisfies: , wherein, is the humidity change amount, is the humidity peak value during the negative pressure perturbation, is the baseline humidity value before the negative pressure perturbation, is the preset humidity pulse identification threshold.

[0014] Preferably, the system further includes a negative pressure application unit, which applies a short-term negative pressure perturbation to the inner cavity of the ladle under the instruction of the feature extraction and stage identification unit by controlling the suction force of the exhaust system connected to the ladle exhaust duct or the combination of bypass valves.

[0015] Preferably, if the feature extraction and stage identification unit detects a specific humidity pulse response, the baking control unit is configured to adjust the subsequent baking strategy, including extending the current low-temperature drying stage or starting to heat up at a gentler rate until there is no specific humidity pulse response when the enhanced discrimination mechanism for residual moisture evaporation response under transient negative pressure perturbation is executed again.

[0016] Compared with the problems in the background art, the beneficial effects of the present invention are: 1. By coupling and analyzing the temporal evolution of waste gas humidity, temperature, and flow parameters, the system can construct core characteristic indicators reflecting the moisture migration and heat absorption states inside the ladle, and accurately divide the physical and chemical stages during the baking process based on this. This mechanism avoids the limitations of traditional apparent parameter control, enabling the heat output strategy to automatically switch according to the internal requirements of the baking stage. It prioritizes ensuring efficient moisture removal during the intense evaporation stage, and focuses on promoting uniform temperature increase inside the refractory during the pure temperature increase stage, achieving dynamic adaptation of energy efficiency distribution and material thermodynamic response, and effectively avoiding the common industry contradictions of over-baking and under-baking.

[0017] 2. The system captures the implicit influence of different initial ladle states (such as residual moisture, ladle age differences) on the baking process through real-time entropy flow characteristic curves, and automatically generates personalized control trajectories matching their heat capacity characteristics. This mechanism transforms the traditional fixed process parameters driven by experience into dynamic decision-making logic based on the feedback of the ladle's own state, enabling different scenarios such as newly built ladles and recycled ladles to obtain optimal baking paths, significantly improving process adaptability and quality consistency.

[0018] 3. At the critical stage at the end of drying, the system actively stimulates the migration response of deep pore moisture through transient negative pressure perturbation, and realizes the penetrability verification of residual moisture based on highly sensitive humidity pulse detection. This mechanism forms a closed-loop verification with the stage recognition module of the main control system. It not only inherits the accurate judgment of the main plan for the macroscopic process, but also strengthens the guarantee ability of drying thoroughness through capturing the microscopic response under physical perturbation. The two work together effectively to solve the industry problem of difficult detection of moisture in dead corners inside ladles with complex structures.

[0019] 4. By integrating waste gas temperature trend analysis, real-time surface temperature monitoring, and dynamic thresholds of entropy increase rate, the system constructs a multi-layer redundant temperature safety protection mechanism. During the intense evaporation stage, the coupled analysis of waste gas heat loss rate and surface temperature can predict local overheating risks in advance; during the pure temperature increase stage, the feedback of the entropy increase rate on the overall heat absorption efficiency of the refractory can suppress ineffective heat input. This multi-dimensional signal fusion mechanism avoids the lag defect of single temperature monitoring, achieving active prediction and precise intervention of thermal shock protection.

[0020] 5. Through the baking target type selection module, the system transforms different requirements such as complete drying and rapid heating into dynamic regulation rules for entropy increase characteristic indicators. For newly built ladles, the system automatically extends the stability criterion for the low-temperature evaporation stage to ensure sufficient removal of bound water; for recycled ladles, it realizes rapid reuse by optimizing the heating rate and soaking time. This flexible architecture enables the same hardware system to internally support multiple process paradigms, significantly improving the equipment reuse efficiency and the intensive level of process management. Description of the Drawings

[0021] Figure 1 Schematic diagram of the hierarchical module composition structure of the present invention; Figure 2 Schematic diagram of the network connection structure between the central control room and the local system of the present invention; Figure 3 Data flow structure diagram of multi-sensor collaborative perception and feature recognition in the exhaust gas monitoring unit of the present invention; Figure 4 Schematic diagram of the evolution trend of each stage in the ladle baking process of the present invention.

[0022] The realization of the purpose, functional characteristics and advantages of the present invention will be further described in conjunction with the embodiments with reference to the accompanying drawings. Specific embodiments

[0023] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0024] The embodiment of the present application provides an intelligent control system for ladle baking, and the system includes: An exhaust gas monitoring unit, configured in the exhaust duct during the ladle baking process, for real-time collecting the humidity parameter, temperature parameter and flow parameter of the exhaust gas; A feature extraction and stage recognition unit, connected to the exhaust gas monitoring unit, and configured to: based on the exhaust gas humidity parameter, exhaust gas temperature parameter and exhaust gas flow parameter, perform coupled calculation to obtain a baking entropy increase rate feature index characterizing the baking process inside the ladle; by continuously analyzing the temporal change of the baking entropy increase rate feature index, dynamically identify at least one physicochemical stage among the violent evaporation stage, evaporation slowdown stage, pure temperature rise stage and approaching thermal saturation stage in the current baking process; A baking control unit, connected to the feature extraction and stage recognition unit and connected to the heating device of the ladle, and configured to: according to the current physicochemical stage identified by the feature extraction and stage recognition unit, set a corresponding target baking entropy increase rate interval; according to the deviation between the real-time monitored baking entropy increase rate feature index and the target baking entropy increase rate interval, dynamically adjust the total heat output of the heating device to guide the baking process to achieve homogenized drying and temperature rise inside the ladle.

[0025] Preferably, the exhaust gas monitoring unit includes a humidity sensor configured to monitor the exhaust gas humidity parameter, a temperature sensor configured to monitor the exhaust gas temperature parameter, and a flow sensor or flow estimation module configured to monitor the exhaust gas flow parameter.

[0026] Preferably, the characteristic index of the baking entropy increase rate includes the moisture evaporation amount per unit time obtained by calculating the waste gas humidity parameter, the waste gas temperature parameter, and the waste gas flow parameter. Among them, the moisture evaporation amount per unit time is higher than the preset initial moisture evaporation threshold value in the initial stage of baking, and shows a downward trend and forms a characteristic inflection point as the free water and part of the bound water in the refractory material are removed.

[0027] Preferably, the feature extraction and stage recognition unit is configured to determine that the baking process transitions from the intense evaporation stage or the evaporation slowdown stage mainly dominated by moisture evaporation to the pure heating stage mainly dominated by the temperature rise of the refractory material matrix by detecting the downward inflection point of the moisture evaporation amount per unit time in time series.

[0028] Preferably, the baking control unit is configured as follows: in the intense evaporation stage, if the characteristic index of the baking entropy increase rate is lower than the lower limit of the target baking entropy increase rate range, increase the heat output of the heating device; if the characteristic index of the baking entropy increase rate is higher than the upper limit of the target baking entropy increase rate range, decrease the heat output of the heating device; in the pure heating stage, adjust the heat output strategy according to the change of the waste gas temperature parameter to effectively increase the internal temperature of the refractory material; when the characteristic index of the baking entropy increase rate tends to the preset stable value, indicating that the inside of the ladle tends to be in a thermal equilibrium and dry state, the baking control unit automatically switches to the heat preservation stage or the programmed cooling stage.

[0029] Preferably, the system further includes an auxiliary temperature monitoring unit configured to monitor the surface temperature or the near-surface temperature of the ladle lining; when adjusting the total heat output of the heating device, the baking control unit refers to the feedback of the auxiliary temperature monitoring unit to ensure that the surface temperature or the near-surface temperature of the ladle lining does not exceed the preset safety temperature upper limit.

[0030] Preferably, the baking control unit is configured to select different control strategies according to the preset baking target type, and the baking target type includes completely drying a newly built ladle, quickly heating a turnover ladle, or energy-saving baking.

[0031] Preferably, the feature extraction and stage recognition unit is configured to enable a discrimination mechanism for enhancing the residual moisture evaporation response under transient negative pressure perturbation when it is judged that the ladle baking process reaches the preset dry end stage or the pure heating initial stage; the waste gas monitoring unit is configured to monitor the transient change of the waste gas humidity parameter in the exhaust flue at a sampling frequency higher than the normal sampling frequency during the application of transient negative pressure perturbation; the feature extraction and stage recognition unit is further configured to confirm the final dry state of the ladle by judging whether the transient change of the waste gas humidity parameter presents a specific humidity pulse response indicating the accelerated evaporation of residual moisture inside the ladle under negative pressure induction; among them, the detection logic of the specific humidity pulse response satisfies: , wherein, is the humidity change amount, is the humidity peak value during the negative pressure perturbation, is the baseline humidity value before the negative pressure perturbation, is the preset humidity pulse recognition threshold; In addition, it should be noted that the moisture evaporation rate per unit time in the system is used as a basic parameter to characterize the dehumidification intensity at the initial stage of baking, mainly to reflect the dynamic migration process of free water and bound water, while the entropy increase rate characteristic index is a characteristic quantity of the system thermodynamic state calculated comprehensively based on multiple parameters such as moisture evaporation, heat input, and smoke exhaust flow. Although they are related in the calculation logic, they are respectively used for the working condition identification and control decision-making in different stages; regarding the detection logic of the specific humidity pulse response, in order to enhance the adaptability and engineering practicality of the system, two types of equivalent criterion expression forms are introduced, where The form is applicable to the standardized modeling scenario, and the 3% humidity sudden increase comes from the empirical determination value under the large sample statistical analysis. Both are reliable and substitutable in actual operation; in terms of the corresponding relationship between the entropy increase rate and the heat regulation strategy, the system adjusts the power output of the heating device through a dynamic feedback loop according to the deviation degree between the entropy increase rate index and the set target interval. This mapping logic constitutes a multivariable adaptive control path in the control module by combining real-time sensing feedback and stage recognition results to ensure the responsiveness and stability of the heat input in each stage; and the threshold parameters used in various physical and chemical stage criteria in the present invention, including the moisture evaporation rate limit, the entropy increase rate fluctuation range, the lower limit of the temperature rise rate, etc., for example, can be derived from the statistical results of the measured data of different types of ladles under multiple furnace types and different ambient temperatures, which all belong to the extended implementation methods known to those of ordinary skill in the art.

[0032] Preferably, the system further includes a negative pressure application unit, which applies a short-term negative pressure perturbation to the inner cavity of the ladle under the instruction of the feature extraction and stage recognition unit by controlling the suction force of the smoke exhaust system connected to the ladle smoke exhaust duct or the bypass valve combination.

[0033] Preferably, if the feature extraction and stage recognition unit detects a specific humidity pulse response, the baking control unit is configured to adjust the subsequent baking strategy, including extending the current low-temperature drying stage or starting to heat up at a gentler rate until there is no specific humidity pulse response when the residual moisture evaporation response enhancement discrimination mechanism under transient negative pressure perturbation is executed again.

[0034] Example 1: In this example, combined with the typical operating environment of a steelmaking plant area, the implementation path of an intelligent ladle baking control system is specifically elaborated. Its core objective is to achieve dynamic adaptive control of the ladle drying and heating processes by real-time sensing and analysis of the thermodynamic characteristics of the exhaust gas, thereby ensuring uniform and thorough baking effects, reasonable energy consumption, and further guaranteeing the integrity of the refractory structure and the service life of the ladle.Before the steel ladle enters the baking area, first, through a composite high-precision sensing device embedded in the smoke exhaust system, the parameters to be collected in the smoke exhaust duct are monitored in real time. The humidity parameter is collected by a capacitive micro-humidity sensor, whose response time is controlled within two seconds and has the ability to operate stably for a long time in an environment of high-temperature water vapor impact; the temperature parameter is measured by an integrated thermocouple module, and its probe position is fixed slightly below the central axis of the main exhaust gas flow after multiple rounds of debugging to ensure that the measured value can truly reflect the heating trend in the central area of the furnace body; the flow parameter is collected by a thermal mass flowmeter and supplemented with a wind pressure fluctuation calibration module for error correction, so as to obtain an accurate trend of the change in the smoke exhaust volume flow rate, ensuring the representativeness and effectiveness of the data. The core function of the feature extraction and stage recognition unit is to continuously analyze the combined change trend of the above three parameters to construct an entropy increase rate feature index reflecting the internal hydrothermal state. This index aims to characterize the phase change intensity and heat transfer effect of free water and bound water driven by the heat inside the ladle per unit time; to enhance the practicality of this index in engineering applications, the system sets a reference evaporation section as the judgment baseline for entropy increase change in the initial stage. Subsequently, by monitoring the change rate of the water evaporation amount per unit time, when the system observes that the fluctuation amplitude of this value is less than the median value of a stable interval obtained through statistical analysis of measured data for five consecutive minutes, it can be determined that the intense evaporation stage has ended and entered the evaporation slowdown or pure heating stage; the calculation of the entropy increase rate feature index adopts the modeling result of a fusion function based on the three parameters of exhaust gas humidity, temperature, and flow rate. Its physical meaning can be regarded as a comprehensive representation of the intensity of water migration and phase change driven by the thermal energy in the system per unit time. According to the acquisition and processing of various parameters, through engineering modeling and training and verification with multiple groups of actual sample data, the rationality and feasibility are ensured. In addition, the relevant numerical thresholds in the stage criterion (such as the evaporation rate interval, the fluctuation amplitude of the entropy increase rate, etc.) all originate from the measured statistical results of typical newly built ladles and turnover ladles under different furnace types, with representativeness and repeatability. Although there is data intersection among the functional modules of the system, such as the auxiliary temperature monitoring unit, which is mainly used for independent temperature control protection and heat loss warning, forming a functional complement with the heat input regulation strategy based on entropy increase feedback in the main control module, enhancing the system safety and response sensitivity from multiple dimensions; although the auxiliary temperature monitoring unit operates independently as a thermal safety guarantee component, when the main control system adjusts the heating power, its feedback data is also used as an auxiliary reference input to correct or confirm the thermal control trend calculated by the entropy increase rate feature index, so as to achieve dynamic regulation under the fusion of multi-source information. Its role is not to directly interfere with the main control logic, but to supplement and redundantly protect the main control strategy with monitoring boundary constraints and response delays, constituting the coordination and complementarity between the inner and outer layer thermal control mechanisms.

[0035] During actual operation, if the system recognizes that the current ladle is a newly built structure, the control module will automatically call the steady-state extension control strategy due to its high initial humidity and large heat capacity. This strategy mainly extends the heat input time in the low-temperature section and limits the initial heating rate to prevent sealing sintering due to excessively rapid surface temperature increase. At this stage, the heat input corresponding to the target range of entropy growth rate set by the system is controlled between 80% and 95% of the basic setting, and the heat input intensity is dynamically corrected according to the time-series change trend of the exhaust humidity to ensure that moisture is released layer by layer from the inside of the ladle and achieves uniform drying; after entering the pure heating stage, the entropy growth rate shows a trend of change from dynamic fluctuations to monotonic slowdown. , the exhaust gas temperature also begins to rise steadily. At this time, the control module adjusts the heat input strategy according to the average change rate of the exhaust gas temperature. When the temperature rise slope is not less than three degrees Celsius per minute for ten consecutive minutes, the system gradually increases the heat source output to the full power level set by the basic setting. At the same time, an auxiliary temperature monitoring mechanism is introduced, and a non-contact infrared temperature measuring device installed near the surface of the ladle lining is used to collect temperature data in real time. When the lining temperature reaches the preset safety upper limit, the system will automatically trigger the output adjustment logic to reduce the heat source input power to prevent the surface material from burning or cracking due to thermal shock; at the end of the drying process, the system starts the transient negative pressure perturbation response mechanism, which cooperates with the exhaust fan system to At the same time, a small vacuum operation is applied inside the ladle for a short time to form a low-amplitude negative pressure disturbance, which stimulates the moisture that may remain in the deep pores of the refractory to evaporate. The duration of the disturbance operation is controlled within 30 seconds, and the vacuum intensity does not exceed 150% of the conventional exhaust state, so as to maintain the stability of the system operation while ensuring the effectiveness of the disturbance. Before and after the disturbance, the humidity sensor samples at a frequency of once per second. The system detects whether a rapid humidity rise of more than 3% occurs within ten seconds after the start of the disturbance. If the humidity pulse response feature is detected, the system determines that there is still residual moisture and will test again after maintaining the current stage for ten to fifteen minutes; if no result is found in two consecutive tests, the humidity sensor will be tested again. If there is a pulse response, the system considers that the drying is completed and can enter the programmed cooling or insulation stage. At the same time, in order to ensure the safety of transient negative pressure disturbance operation and system stability, the control module will synchronously lock the output state of the heating device before applying the disturbance to avoid system fluctuations caused by the superposition of heat input and disturbance process. The system also has upper limits on disturbance intensity and duration to ensure that negative pressure operation does not cause significant interference or sudden temperature changes to the furnace temperature field. In addition, the auxiliary temperature monitoring module continuously tracks the near-surface temperature changes of the ladle lining during the disturbance. Once an abnormal fluctuation in the heating rate is detected, the safety protection logic will be automatically triggered to suspend the heating program and perform system verification to achieve dynamic suppression of disturbance side effects and controllable whole process.

[0036] During the entire operation process, all control logics and parameter settings relied on by the system are based on the accumulation and induction of multiple groups of experimental verifications and on-site operation data. For example, the reference threshold of the entropy increase rate is a key criterion extracted through statistical analysis of measured samples under more than twenty groups of different ladle ages, different lining structures, and different furnace types; the selection of parameters such as the upper limit of temperature control, the adjustment range of heat flux density, and the boundary of disturbance intensity also all stem from physical tests on the high-temperature refractory material test bench, comprehensively considering key factors such as thermal gradient, water vapor diffusion, specific heat of materials, and porosity, ensuring that each parameter setting has a high degree of engineering feasibility and safety margin. At the same time, the system also fully considers the adaptation requirements when multiple process objectives coexist. When receiving a rapid heating instruction, the control module will automatically adjust the control strategy, weaken the time constraint in the dehydration stage, and shift the system response priority to improving the heat penetration efficiency, increasing the heat input rate under the condition of ensuring safety; when switching to the energy-saving baking mode, it realizes the control objective of meeting the drying effect requirements with the lowest energy consumption by reducing the total heat input, extending the residence time of each stage, and relaxing the stage conversion criterion.

[0037] Example 2: In the typical production process of a steel mill, there are significant differences in the lining state of the ladle. Taking a newly built ladle as an example, due to the relatively high initial moisture content of its refractory material, if a traditional unified heating control curve is adopted, it is extremely easy to cause rapid sintering on the surface while the moisture in the deep layer fails to be fully removed. On the other hand, if the long-cycle process of a newly built ladle is still used for a recycled ladle, it is easy to cause energy waste. Therefore, an intelligent control system is constructed that can dynamically identify the ladle baking state based on exhaust gas parameters and adjust the heat input accordingly in real time to achieve personalized control of different ladle structures and ensure the uniformity of the baking effect, the thermal energy utilization efficiency, and the operation safety.

[0038] In this experiment, 6 ladles with the same structure, the same lining material, and a rated capacity of 30 tons were selected as the experimental objects, among which 3 were newly built ladles (the initial moisture content of their refractory materials was about 15%), and 3 were recycled ladles (the moisture content was less than 3%). All ladles independently completed the full-process baking in the same gas heating furnace in sequence to ensure the representativeness and reproducibility of the data; the humidity parameter was collected by a capacitive high-temperature micro-humidity sensor with a sampling frequency of 1 Hz; the temperature parameter was collected by a K-type thermocouple arranged at the lower-axis position of the mainstream of the exhaust pipe, sampling once every 2 seconds; the exhaust gas flow rate was collected by a thermal mass flowmeter, supplemented by a wind pressure fluctuation correction module to improve the measurement accuracy; the auxiliary infrared temperature measurement module was used to non-contact monitor the near-surface temperature of the ladle lining to provide safety temperature control support. The entropy increase rate characteristic index was comprehensively calculated from the moisture evaporation intensity per unit time (unit: g / min), the change trend of the exhaust gas enthalpy, and the exhaust gas flow rate, and was used to characterize the dynamic characteristics of the moisture migration and heat conduction state driven by heat. To achieve stable identification of each stage, the following physical and chemical stage criteria were set in this system:

[0039] Near the end of the pure heating stage, the system automatically activates the transient negative pressure perturbation mechanism to identify the existence of deep residual moisture. During the perturbation, if a typical humidity pulse response is collected (i.e., the humidity suddenly rises by more than 3% within 10 seconds after the perturbation), it is determined that the drying is not completed. Taking the newly built ladle numbered #N1 as an example, its key operating data is shown in the following table: Among them, the newly built ladle sample needs to perform two rounds of negative pressure perturbation to confirm complete drying; while the turnover ladle sample enters the pure heating stage at the 80th minute, the entropy increase rate stabilizes below 0.45, and no characteristic of sudden humidity rise appears. The change trends of the entropy increase rate and the moisture migration intensity collected during the test are highly consistent, the stage switching nodes are clear, the system control strategy response is reasonable, verifying the adaptability of the control system to different ladle states. Through the test process and data analysis, the following conclusions are drawn: The entropy increase rate characteristic index constructed based on the three exhaust gas parameters can effectively reflect the moisture migration and heat absorption states inside the ladle; the physical and chemical characteristics of different baking stages can be stably and reliably identified through the trend of this index; the control system can dynamically adjust the heat output according to this index to achieve effective adaptation to different ladle types such as newly built ladles and turnover ladles; the transient negative pressure perturbation mechanism has obvious penetration verification ability in the final drying stage and can form a closed-loop confirmation of the residual moisture; all parameter settings are derived from the statistical analysis of a large number of actual operation samples and the tests of high-temperature experimental devices, and have good engineering feasibility and safety margins.

[0040] Example 3: This example combines Figures 1 to 4 , and illustrates the implementation of an intelligent control system for ladle baking. As Figure 1 shown, the system structure is divided into three major parts: the perception layer, the control layer, and the monitoring layer. Among them, the perception layer includes hardware components such as a valve matrix, an ignition device, an alarm, multi-source sensors, and detection instruments, which are responsible for real-time sensing of key physical quantities during the ladle baking process. The control layer consists of multiple functional modules. First is the data acquisition and processing module, which specifically includes multi-source sensor acquisition and data preprocessing, a state switching control module that supports manual switching and full automation operation of the system, and an equipment fault diagnosis module that realizes fault diagnosis and alarm, and safety protection functions; in addition, it also includes an advanced baking process control strategy module, which has the capabilities of temperature-flow cascade control, APC temperature control algorithm, and combustion optimization control. The monitoring layer is composed of a smart interaction module, a remote monitoring module, and a process optimization management module. Among them, the smart interaction module provides functions such as baking parameter management, multi-device management, and querying system logs. The remote monitoring module supports remote operation and safety monitoring, and can generate energy consumption reports. The process optimization management module overall improves the overall efficiency of the system operation process.

[0041] As Figure 2 shown, the central control room part includes a client, a main server, a core switch, and an optical-electric converter. The client is used to operate the control interface. The main server is responsible for the management and processing of system operation data. The core switch realizes high-speed network switching between control devices. The optical-electric converter serves as a medium for local communication in the central control room to complete the conversion of optical and electrical signals. The local control part covers the baking site control units of the 1# roaster, 2# roaster, and 3# roaster. Each roaster is respectively configured with an access switch, an HMI, an intelligent controller, and corresponding PLC modules (1# PLC, 2# PLC, 3# PLC); the HMI is used for human-machine interaction interface display and operation. The intelligent controller processes the collected data and executes control strategies. The PLC realizes the underlying control logic of the on-site actuators. The access switch is connected to the aggregation switch to ensure smooth communication between unit modules. The aggregation switch, as the communication aggregation point of the local network, transmits data to the central control room through the optical-electric converter to realize the unified scheduling and centralized monitoring of the system; the entire system architecture supports the parallel management of multiple ladle baking devices and constructs a highly reliable remote and on-site collaborative control network through optical-electric conversion, network switching, and intelligent control to ensure the real-time, accuracy, and stability of system operation.

[0042] As Figure 3 shown, the waste gas monitoring unit includes a temperature sensor, a humidity sensor, a flow sensor or a flow estimation module, a collaborative perception unit, and a feature extraction and stage recognition unit. The temperature sensor is used to collect waste gas temperature parameters in real time. The humidity sensor is used to collect waste gas humidity parameters in real time. The flow sensor or the flow estimation module is used to collect waste gas flow parameters in real time. The data collected by all sensors will be uniformly input into the collaborative perception unit for fusion processing. After the collaborative perception unit performs real-time parameter integration processing on the data transmitted by the temperature sensor, the humidity sensor, and the flow sensor or the flow estimation module, it outputs waste gas temperature parameters, waste gas humidity parameters, and waste gas flow parameters, and further transmits these real-time parameters to the feature extraction and stage recognition unit; based on the above three types of waste gas parameters, the feature extraction and stage recognition unit calculates the baking entropy increase rate characteristic index of the ladle internal baking process through coupling, and accordingly identifies physical and chemical process stages such as the violent evaporation stage, the evaporation slowdown stage, the pure temperature rise stage, and the soaking and saturation stage, realizing the accurate dynamic identification of the ladle drying and heating states.

[0043] As Figure 4As shown in the figure, it demonstrates the dynamic change trends and index evolutions of the humidity curve A, temperature curve B, and entropy increase curve C during the four stages of the ladle baking process, namely, intense evaporation, slowing evaporation, pure temperature rise, and soaking saturation. The horizontal axis represents the baking time (min), and the vertical axis represents the standard values of the humidity, temperature, and entropy increase characteristic indicators. The humidity curve A represents the change in humidity in the ladle exhaust duct, which rapidly decreases during the intense evaporation stage, slows down during the slowing evaporation stage, and gradually stabilizes during the pure temperature rise stage. The temperature curve B shows the trend of the temperature in the center area of the furnace body rising with time, showing a continuous increase throughout the baking process. The entropy increase curve C is the change in the entropy increase rate calculated by the feature extraction and stage recognition unit, with intense fluctuations at the beginning and then gradually stabilizing, entering a plateau state during the soaking saturation stage. In addition, two disturbance events are marked in the figure, namely, disturbance 1 (ΔH = 11%) and disturbance 2 (ΔH = 0%), and the corresponding humidity responses are used to judge whether there is residual moisture in the ladle, thus assisting the system in determining whether the drying is complete.

[0044] Example 4: In a heavy-duty continuous casting workshop with an annual output of over three million tons in a steel mill, aiming at the typical application scenario where the newly built ladle has a high humidity in winter, a tight turnover cycle, and strict requirements for production line coherence, this ladle baking intelligent control system is implemented. In this scenario, the initial moisture content of the ladle lining material usually exceeds 15%. Moreover, due to the tight production organization rhythm, there is a practical need to shorten the baking cycle. If not properly controlled, it is extremely easy to produce the phenomenon of dry surface and wet interior during the heating stage, and then cause the refractory material to crack during the first steel water impact. This example realizes an adaptive control strategy through the dynamic feedback of the internal thermal state of the ladle, so as to ensure the goal of homogeneous drying and heating of the ladle within a limited time. After the ladle is moved into the baking position, the system automatically activates the sensing layer. By arranging a humidity sensing unit, an enthalpy sensing module, and a dynamic flow estimation component in the exhaust system, a three-parameter collaborative sensing matrix of waste gas humidity parameters, waste gas temperature parameters, and waste gas flow parameters is constructed. The collected data is input into the core calculation platform at a frequency of once per second. The feature extraction module generates a thermal state change trend map. This module continuously tracks the moisture evaporation intensity and the change in waste gas heat content per unit time within the initial ten minutes, and thus forms the first batch of entropy increase rate sample data. The system compares this sample data with the entropy increase characteristic curve under the working conditions of similar ladles in the historical database, matches the heat capacity level and moisture thermal decoupling characteristics of the current ladle, and identifies it as a high-water and high-heat-capacity type object, and then automatically calls the steady-state extended control strategy.

[0045] In the initial stage of intense evaporation, the target entropy increase rate range set by the system is based on the lower boundary of the stable evaporation characteristic curve of a typical newly-built ladle. Considering that the furnace temperature is relatively low and the external environmental temperature is about minus five degrees Celsius, the control module actively adjusts the initial heat input to be between 75% and 85% of the standard full-load setting. By periodically adjusting the injection angle of the burner, the spatial uniformity of heat distribution is achieved, thereby reducing the risk of local overheating. At the same time, the auxiliary temperature monitoring device synchronously collects the temperature data near the surface of the ladle. If the correlation deviation between the detected surface temperature change curve and the entropy increase rate change curve exceeds the set warning threshold (i.e., 3%), the slope correction logic is immediately triggered, automatically reducing the heat input rate and extending the duration of the intense evaporation stage; when the system continuously monitors that the moisture evaporation rate per unit time drops below 35 grams per minute, and the fluctuation range of the entropy increase rate is lower than 5% and lasts for more than five minutes, the system determines that it has entered the evaporation slowdown stage, and at the same time enables the decision node strengthening mechanism. This mechanism further verifies the effectiveness of the trend inflection point by performing a first-order derivative fitting on the entropy increase curve to ensure that the stage transition is not misjudged by short-term disturbances.

[0046] In the evaporation slowdown stage, the control module stabilizes the heat input at about 65% of the full-load setting and adjusts the opening ratio of the nozzle array based on the dynamic change trend of the waste gas heat enthalpy to maintain the equilibrium stability of the furnace temperature field. When the waste gas humidity parameter stabilizes below the baseline value and the waste gas temperature continuously rises at a rate of more than 2.5 degrees Celsius per minute, and the entropy increase rate shows a downward trend, the system determines that the ladle is in the pure heating stage. In this stage, considering that the ladle adopts a double-layer composite corundum brick structure with large thermal inertia and relatively low surface thermal conductivity, the system preferably adopts an intermittent incremental heating strategy. Specifically: taking 15 minutes as a unit cycle, gradually increasing the heat input to 95% of the full-load setting. At the same time, an amplitude limit algorithm is introduced to control the response frequency of the heat source power to prevent thermal shock caused by too fast heat input. The surface temperature data collected by the auxiliary temperature monitoring module is input into the proportional-integral-prediction model in the main controller to form a real-time closed-loop feedback system; if it is detected that the surface temperature rise rate is lower than the expected value, it is automatically determined as a situation of excessive thermal resistance, and a pulsed high heat flux intervention strategy is triggered.

[0047] When approaching the end of the pure temperature rise stage, the system starts the residual moisture discrimination program, calls the transient negative pressure application module to perform a low-amplitude air extraction operation on the smoke exhaust passage for no more than 30 seconds, and the extraction force does not exceed 130% of the conventional exhaust negative pressure. During the disturbance application period, the humidity sensing unit monitors the humidity parameters of the smoke exhaust duct at twice the sampling frequency. If the system detects that the humidity suddenly rises by more than 3% within 10 seconds after the disturbance starts, it is determined that there is still deep residual moisture inside the ladle. The control module automatically extends the current holding stage and plans to execute a second round of disturbance operations. If no obvious humidity pulse response is detected in the second disturbance, the system determines that the drying process is completed and then enters the programmed temperature reduction stage; during the entire operation process, the target interval of the entropy increase rate, the heat flow regulation strategy, and the determination thresholds of each stage used by the system all come from the historical sample statistical laws extracted by the data-driven modeling algorithm in the local database, and are optimized and verified in combination with the data of more than 30 actual baking conditions. The setting of all parameters has clear engineering basis and control boundaries to ensure that the system has good implementability and adaptability, and all belong to the extended implementation methods known to those of ordinary skill in the art.

[0048] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention.

[0049] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An intelligent control system for ladle baking, characterized in that, The system includes: An exhaust gas monitoring unit, which is configured in the exhaust duct during the ladle baking process and is used to collect the humidity parameter, temperature parameter and flow parameter of the exhaust gas in real time; A feature extraction and stage identification unit, connected to the exhaust gas monitoring unit, and is configured to: based on the exhaust gas humidity parameter, exhaust gas temperature parameter and exhaust gas flow parameter, calculate and obtain a baking entropy increase rate characteristic index representing the baking process inside the ladle through coupling; by continuously analyzing the temporal variation of the baking entropy increase rate characteristic index, dynamically identify at least one physicochemical stage among the intense evaporation stage, evaporation slowdown stage, pure temperature rise stage and approaching thermal equilibrium and saturation stage in the current baking process; A baking control unit, connected to the feature extraction and stage identification unit and connected to the heating device of the ladle, and is configured to: set a corresponding target baking entropy increase rate interval according to the current physicochemical stage identified by the feature extraction and stage identification unit; according to the deviation between the baking entropy increase rate characteristic index monitored in real time and the target baking entropy increase rate interval, dynamically adjust the total heat output of the heating device to guide the baking process so that the inside of the ladle reaches homogenized drying and temperature rise.

2. The intelligent control system for ladle baking according to claim 1, wherein The exhaust gas monitoring unit includes a humidity sensor configured to monitor the exhaust gas humidity parameter, a temperature sensor configured to monitor the exhaust gas temperature parameter, and a flow sensor or flow estimation module configured to monitor the exhaust gas flow parameter.

3. The intelligent ladle baking control system according to claim 2, wherein, The baking entropy increase rate characteristic index includes the moisture evaporation amount per unit time calculated through the exhaust gas humidity parameter, exhaust gas temperature parameter and exhaust gas flow parameter. Among them, the moisture evaporation amount per unit time is higher than the preset initial moisture evaporation threshold value in the initial stage of baking, and shows a downward trend and forms a characteristic inflection point as the free water and part of the bound water in the refractory material are removed.

4. The intelligent ladle baking control system according to claim 3, wherein The feature extraction and stage identification unit is configured to determine that the baking process transitions from the intense evaporation stage or evaporation slowdown stage mainly dominated by moisture evaporation to the pure temperature rise stage mainly dominated by the temperature rise of the refractory material matrix by detecting the downward inflection point of the moisture evaporation amount per unit time in time series.

5. The intelligent control system for ladle baking according to claim 1, wherein The baking control unit is configured to: in the intense evaporation stage, if the baking entropy increase rate characteristic index is lower than the lower limit of the target baking entropy increase rate interval, increase the heat output of the heating device; if the baking entropy increase rate characteristic index is higher than the upper limit of the target baking entropy increase rate interval, decrease the heat output of the heating device; in the pure temperature rise stage, adjust the heat output strategy according to the change of the exhaust gas temperature parameter; when the baking entropy increase rate characteristic index approaches the preset stable value, indicating that the inside of the ladle tends to be in a thermal equilibrium and dry state, the baking control unit automatically switches to the heat preservation stage or the programmed cooling stage.

6. The intelligent control system for ladle baking according to claim 1, wherein The system further includes an auxiliary temperature monitoring unit, which is configured to monitor the surface temperature or near-surface temperature of the ladle lining; when adjusting the total heat output of the heating device, the baking control unit refers to the feedback of the auxiliary temperature monitoring unit to ensure that the surface temperature or near-surface temperature of the ladle lining does not exceed the preset safety temperature upper limit.

7. The intelligent control system for ladle baking according to claim 1, wherein The baking control unit is configured to select different control strategies according to the preset baking target type, and the baking target type includes completely drying a newly-built ladle, quickly heating a turnover ladle or energy-saving baking.

8. The intelligent control system for ladle baking according to claim 1, characterized in that The feature extraction and stage recognition unit is configured to enable the enhanced discrimination mechanism for the residual moisture evaporation response under transient negative pressure perturbation when it is determined that the ladle baking process reaches the preset end stage of drying or the initial stage of pure temperature rise; the exhaust gas monitoring unit is configured to monitor the transient changes of the exhaust gas humidity parameters in the smoke exhaust duct at a sampling frequency higher than the normal sampling frequency during the application of transient negative pressure perturbation; the feature extraction and stage recognition unit is further configured to finally confirm whether the ladle reaches the completely dry state by judging whether the transient changes of the exhaust gas humidity parameters present a specific humidity pulse response indicating the accelerated evaporation of residual moisture inside the ladle under negative pressure induction; among them, the detection logic of the specific humidity pulse response satisfies: ， wherein, is the humidity change amount, is the humidity peak value during the negative pressure perturbation, is the baseline humidity value before the negative pressure perturbation, is the preset humidity pulse recognition threshold.

9. The intelligent ladle baking control system according to claim 8, characterized in that The system further includes a negative pressure application unit, which applies a short-term negative pressure perturbation to the inner cavity of the ladle under the instruction of the feature extraction and stage recognition unit by controlling the suction force of the smoke exhaust system connected to the ladle smoke exhaust duct or the bypass valve combination.

10. The ladle baking intelligent control system according to claim 8, characterized in that, If the feature extraction and stage recognition unit detects a specific humidity pulse response, the baking control unit is configured to adjust the subsequent baking strategy, including extending the current low-temperature drying stage or starting to heat up at a gentler rate until there is no specific humidity pulse response when the enhanced discrimination mechanism for the residual moisture evaporation response under transient negative pressure perturbation is executed again.

Images