A controlled cooling system cooling medium intelligent connection control method and system

Abstract

The application discloses a kind of cooling system cooling medium intelligent connection control method and system, the method includes: the multimodal process parameters collected by the sensing array deployed in the water cooling and air cooling connection area are acquired;According to the multimodal process parameters, determine process stability index;When the process stability index is lower than the preset threshold, trigger dynamic compensation mechanism, according to the real-time position, real-time speed and real-time temperature of the rolled piece, generate dynamic compensation control instruction;The dynamic compensation control instruction is executed, at least one of the start-stop timing, flow regulation curve or medium mixing ratio of water spraying and air spraying is adjusted, to connect water cooling and air cooling process.The application solves the technical problem of cooling discontinuity in the water cooling and air cooling transition area, realizes the intelligent connection and dynamic compensation of the cooling medium, effectively improves the cooling quality and process stability of rolling process.

Description

Technical Field

[0001] This invention relates to the field of steel production technology, and in particular to a method and system for intelligent connection and control of cooling medium in a controlled cooling system. Background Technology

[0002] Current high-temperature linear cooling systems mainly employ the following technical solutions: first, fixed-sequence water-cooling / air-cooling switching control; second, static adjustment based on simple temperature detection; and third, dynamic adjustment through manual experience intervention. These existing technologies have the following significant drawbacks:

[0003] The lack of a systematic control strategy makes it difficult to achieve a smooth transition between water cooling and air cooling zones, resulting in discontinuous cooling and the formation of "uncooled sections." This, in turn, leads to process failures such as steel accumulation bottlenecks, roller table thermal deformation, and scrap steel in the wire drawing area. Furthermore, the existing system is insufficiently responsive to dynamic changes in process parameters such as workpiece temperature and equipment status. Fault handling measures are passive and lack scientific decision-making basis. Process improvement has long relied on on-site manual observation and experience accumulation, which restricts further improvements in finished product quality and yield. Summary of the Invention

[0004] In order to overcome the above-mentioned shortcomings of the prior art, the purpose of this invention is to provide a method and system for intelligent connection control of cooling medium in a controlled cooling system.

[0005] The technical solution adopted by this invention to solve its technical problem is: a method for intelligent connection control of cooling medium in a controlled cooling system, comprising:

[0006] Acquire multimodal process parameters collected by the sensor array deployed in the area where water cooling and air cooling meet;

[0007] Based on the multimodal process parameters, determine the process stability index;

[0008] When the process stability index is lower than a preset threshold, a dynamic compensation mechanism is triggered to generate a dynamic compensation control command based on the real-time position, real-time speed and real-time temperature of the rolled piece.

[0009] The dynamic compensation control command is executed to adjust at least one of the following: the start-stop sequence of water spray and air spray, the flow rate adjustment curve, or the medium mixing ratio, so as to connect the water cooling and air cooling processes.

[0010] As a further improvement of the present invention: the acquisition of multimodal process parameters collected by the sensor array deployed in the water-cooling and air-cooling junction area includes:

[0011] The temperature field distribution data of the rolled piece collected by the infrared temperature measurement array is obtained, and the two-dimensional temperature measurement data is combined with the movement trajectory of the rolled piece to perform three-dimensional reconstruction of the temperature field and construct a dynamic temperature gradient model.

[0012] Acquire the geometric shape data of the roller conveyor collected by the laser rangefinder;

[0013] Acquire cooling medium flow data collected by the flow sensor;

[0014] Acquire mechanical vibration spectrum data collected by the vibration detection module.

[0015] As a further improvement of the present invention: the step of determining the process stability index based on the multimodal process parameters includes:

[0016] The multimodal process parameters are fused and processed by a deep neural network to output a process stability index, which is a standardized score that includes four dimensions: temperature consistency, cooling uniformity, equipment safety, and production rhythm.

[0017] The process stability index is adjusted based on the prediction results by using a long short-term memory network to predict the process trend within a preset time period.

[0018] As a further improvement of the present invention: the generation of dynamic compensation control instructions includes:

[0019] The preset empirical delay parameters are used as initial values ​​and input into the model predictive control framework.

[0020] Within a preset control cycle, the action sequence for a future preset time period is optimized by using the workpiece heat conduction model, the cooling medium flow model, and the equipment dynamic response model.

[0021] The parameters of the workpiece heat conduction model, cooling medium flow model, and equipment dynamic response model are identified online using real-time data to update the model parameters.

[0022] Based on the optimization results, dynamic compensation control commands are generated.

[0023] As a further improvement to the present invention, it also includes:

[0024] Acquire axial temperature distribution data of the roller conveyor monitored by the distributed fiber optic temperature measurement system;

[0025] Based on the axial temperature distribution data of the roller conveyor, the thermal stress field and predicted deformation of the roller conveyor are calculated by finite element analysis.

[0026] When the predicted deformation exceeds the safety threshold, the cooling strategy is adjusted or the roller conveyor is triggered to rotate.

[0027] As a further improvement to the present invention, it also includes:

[0028] A multi-objective optimization algorithm is adopted, with temperature control accuracy, yield, energy consumption and equipment life as optimization objectives, to solve for the Pareto optimal solution set;

[0029] Control parameters are selected from the Pareto optimal solution set according to the production plan and then sent to the execution system.

[0030] As a further improvement to the present invention, it also includes:

[0031] When an anomaly is detected, the fault diagnosis process is initiated, real-time data and device history records are input into the Bayesian diagnostic network, the probability distribution of each fault cause is calculated, and a diagnostic report is generated.

[0032] When the fault is repairable, a pre-defined self-healing procedure is executed.

[0033] This invention also provides an intelligent control system for the cooling medium of a cooling system, comprising:

[0034] The sensing and detection module is deployed in the area where water cooling and air cooling meet, and is used to collect multimodal process parameters;

[0035] The decision control module is used to determine the process stability index based on the multimodal process parameters. When the process stability index is lower than the preset threshold, a dynamic compensation mechanism is triggered, and a dynamic compensation control command is generated based on the real-time position, real-time speed and real-time temperature of the rolled piece.

[0036] The execution module is used to execute the dynamic compensation control command and adjust at least one of the following: the start-stop sequence of water spray and air spray, the flow rate adjustment curve, or the medium mixing ratio, so as to connect the water cooling and air cooling processes.

[0037] As a further improvement of the present invention: the sensing and detection module includes:

[0038] Infrared temperature measurement array is used to collect temperature field distribution data of rolled parts;

[0039] Laser rangefinders are used to collect geometric data of the roller conveyor.

[0040] Flow sensor, used to collect cooling medium flow data;

[0041] Vibration detection module, used to collect mechanical vibration spectrum data;

[0042] The decision control module is also used to combine the two-dimensional temperature measurement data collected by the infrared temperature measurement array with the movement trajectory of the rolled piece to perform three-dimensional reconstruction of the temperature field and construct a dynamic temperature gradient model.

[0043] As a further improvement of the present invention: the decision control module includes:

[0044] Long Short-Term Memory (LSTM) network units are used to predict process trends within a preset time period in the future, and adjust the process stability index based on the prediction results;

[0045] The online parameter identification subunit is used to update the parameters of the workpiece heat conduction model, cooling medium flow model, and equipment dynamic response model using real-time data.

[0046] Compared with the prior art, the beneficial effects of the present invention are:

[0047] This invention acquires multimodal process parameters in the transition zone between water cooling and air cooling, and uses a dynamic compensation mechanism triggered intelligently based on the process stability index to precisely adjust the start-stop sequence of water and air spraying, flow rate adjustment curves, and media mixing ratio. This ensures a smooth transition between the water cooling and air cooling processes, effectively eliminating the "uncooled section" phenomenon and avoiding the risk of steel buildup due to discontinuous cooling. By sensing changes in the position, speed, and temperature of the rolled piece in real time, the system can dynamically respond to process fluctuations and achieve adaptive compensation control. This systematic compensation control strategy replaces traditional manual experience intervention, making process adjustments more scientific and precise, resulting in substantial improvements in finished product quality and yield. Attached Figure Description

[0048] Figure 1 This is a schematic diagram of the process of the present invention.

[0049] Figure 2 This is a flowchart illustrating an embodiment of the present invention. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0051] In order to solve the technical problems in the prior art, the present invention will now be further described in conjunction with the accompanying drawings and embodiments:

[0052] like Figure 1 As shown, this invention discloses an intelligent connection control method for cooling medium in a controlled cooling system, comprising:

[0053] S1: Acquire the multimodal process parameters collected by the sensor array deployed in the water-cooled and air-cooled junction area;

[0054] In some embodiments, acquiring the multimodal process parameters collected by the sensor array deployed in the water-cooling and air-cooling junction area includes:

[0055] The temperature field distribution data of the rolled piece collected by the infrared temperature measurement array is obtained, and the two-dimensional temperature measurement data is combined with the movement trajectory of the rolled piece to perform three-dimensional reconstruction of the temperature field and construct a dynamic temperature gradient model.

[0056] Acquire the geometric shape data of the roller conveyor collected by the laser rangefinder;

[0057] Acquire cooling medium flow data collected by the flow sensor;

[0058] Acquire mechanical vibration spectrum data collected by the vibration detection module.

[0059] S2: Determine the process stability index based on the multimodal process parameters;

[0060] In some implementations, the process stability index is determined based on multimodal process parameters, including:

[0061] By fusing and processing multimodal process parameters through deep neural networks, a process stability index is output. The process stability index is a standardized score that includes four dimensions: temperature consistency, cooling uniformity, equipment safety, and production rhythm.

[0062] The process trend is predicted within a preset time period by using a long short-term memory network, and the process stability index is adjusted based on the prediction results.

[0063] S3: When the process stability index is lower than the preset threshold, the dynamic compensation mechanism is triggered, and dynamic compensation control commands are generated based on the real-time position, real-time speed and real-time temperature of the rolled piece.

[0064] In some implementations, dynamic compensation control commands are generated, including:

[0065] The preset empirical delay parameters are used as initial values ​​and input into the model predictive control framework.

[0066] Within a preset control cycle, the action sequence for a future preset time period is optimized by using the workpiece heat conduction model, the cooling medium flow model, and the equipment dynamic response model.

[0067] Online parameter identification is performed on the parameters of the workpiece heat conduction model, cooling medium flow model, and equipment dynamic response model using real-time data to update the model parameters;

[0068] Based on the optimization results, dynamic compensation control commands are generated.

[0069] S4: Execute dynamic compensation control commands to adjust at least one of the following: the start-stop sequence of water spray and air spray, the flow rate regulation curve, or the medium mixing ratio, so as to connect the water cooling and air cooling processes.

[0070] By deploying a sensor array in the transition zone between water cooling and air cooling to collect multimodal process parameters in real time, and introducing a process stability index as a trigger condition for dynamic compensation, the controlled cooling system can adaptively adjust the start-stop sequence of water and air sprays, flow rate adjustment curves, and media mixing ratios based on changes in the real-time position, speed, and temperature of the rolled piece. Compared to traditional switching methods that rely on fixed timing or manual experience, this invention achieves a smooth transition between water cooling and air cooling media in the transition zone, eliminating cooling discontinuity problems caused by process parameter mismatches from a control mechanism perspective, and effectively preventing steel pile-up accidents.

[0071] Meanwhile, this method transforms process adjustment from passive response and experience-based reliance to proactive intelligent compensation based on multi-parameter fusion perception, significantly improving the adaptability of the controlled cooling process to process fluctuations, fundamentally improving the control precision and stability of the cooling process, and thereby ensuring the stability of finished product quality and the improvement of yield.

[0072] In some implementations, it also includes:

[0073] Acquire axial temperature distribution data of the roller conveyor monitored by the distributed fiber optic temperature measurement system;

[0074] Based on the axial temperature distribution data of the roller conveyor, the thermal stress field and deformation of the roller conveyor are calculated and predicted by finite element analysis.

[0075] When the predicted deformation exceeds the safety threshold, adjust the cooling strategy or trigger the roller conveyor to rotate.

[0076] In some implementations, it also includes:

[0077] A multi-objective optimization algorithm is adopted, with temperature control accuracy, yield, energy consumption and equipment life as optimization objectives, to solve for the Pareto optimal solution set;

[0078] Control parameters are selected from the Pareto optimal solution set according to the production plan and then issued to the execution system.

[0079] In some implementations, it also includes:

[0080] When an anomaly is detected, the fault diagnosis process is initiated, real-time data and device history records are input into the Bayesian diagnostic network, the probability distribution of each fault cause is calculated, and a diagnostic report is generated.

[0081] When the fault is repairable, a pre-defined self-healing procedure is executed.

[0082] Furthermore, a dynamic compensation mechanism is triggered, including: acquiring the rolling mill tracking signal; and, based on the rolling mill tracking signal and the preset safety interlock logic, performing a speed reduction operation or a full-line emergency stop operation when a steel accumulation risk is detected.

[0083] Based on the same inventive concept, the present invention also provides an intelligent connection control system for cooling medium in a cooling system, comprising:

[0084] The sensing and detection module is deployed in the area where water cooling and air cooling meet, and is used to collect multimodal process parameters;

[0085] The decision control module is used to determine the process stability index based on multimodal process parameters. When the process stability index is lower than the preset threshold, a dynamic compensation mechanism is triggered, and dynamic compensation control commands are generated based on the real-time position, real-time speed and real-time temperature of the rolled piece.

[0086] The execution module is used to execute dynamic compensation control commands to adjust at least one of the following: the start-stop sequence of water spray and air spray, the flow rate regulation curve, or the medium mixing ratio, so as to connect the water cooling and air cooling processes.

[0087] In some implementations, the sensing module includes:

[0088] Infrared temperature measurement array is used to collect temperature field distribution data of rolled parts;

[0089] Laser rangefinders are used to collect geometric data of the roller conveyor.

[0090] Flow sensor, used to collect cooling medium flow data;

[0091] Vibration detection module, used to collect mechanical vibration spectrum data;

[0092] Furthermore, the decision control module is also used to combine the two-dimensional temperature measurement data collected by the infrared temperature measurement array with the movement trajectory of the rolled piece to perform three-dimensional reconstruction of the temperature field and construct a dynamic temperature gradient model.

[0093] Furthermore, the decision control module includes a deep neural network unit for fusing and processing multimodal process parameters and outputting a process stability index. The process stability index is a standardized score that includes four dimensions: temperature consistency, cooling uniformity, equipment safety, and production rhythm.

[0094] In some implementations, the decision control module includes a model prediction control unit, which uses a preset empirical delay parameter as an initial value, optimizes the action sequence within a preset time period through a workpiece heat conduction model, a cooling medium flow model, and an equipment dynamic response model within a preset control period, and generates dynamic compensation control commands based on the optimization results.

[0095] Furthermore, the decision control module includes:

[0096] Long Short-Term Memory (LSTM) network units are used to predict process trends within a preset time period and adjust the process stability index based on the prediction results.

[0097] The online parameter identification subunit is used to update the parameters of the workpiece heat conduction model, cooling medium flow model, and equipment dynamic response model using real-time data.

[0098] In some embodiments, the system of the present invention further includes:

[0099] The roller conveyor thermal deformation protection module is connected to the decision control module and includes a distributed fiber optic temperature measurement system and a finite element analysis unit.

[0100] The distributed fiber optic temperature measurement system is used to monitor the axial temperature distribution data of the roller conveyor.

[0101] The finite element analysis unit is used to calculate the thermal stress field of the roller conveyor and predict the deformation based on the axial temperature distribution data of the roller conveyor.

[0102] The decision control module is also used to adjust the cooling strategy or trigger the roller conveyor rotation when the predicted deformation exceeds the safety threshold.

[0103] Furthermore, the system also includes:

[0104] The safety interlock protection module, connected to the decision control module, is used to acquire the rolling mill tracking signal and, based on the rolling mill tracking signal and the preset safety interlock logic, to perform a speed reduction operation or a full-line emergency stop operation when a risk of steel piling is detected.

[0105] Implementation Case 1:

[0106] like Figure 2 As shown, this embodiment discloses an intelligent cooling medium connection control system for a controlled cooling system, mainly applied to the controlled cooling process in high-speed wire rod rolling, addressing the steel buildup problem caused by process parameter mismatch in the transition zone between water cooling and air cooling. Through innovative technologies such as multimodal sensing detection, dynamic compensation control, and process parameter optimization, it effectively solves the technical problem that the uncooled section can never be completely eliminated, significantly improving finished product quality and yield. Specifically, it includes the following steps:

[0107] Step 1: System Overall Architecture Design

[0108] The intelligent compensation system for the controlled cooling system adopts a three-layer architecture of "detection-decision-execution". The detection layer consists of an infrared temperature measurement array, laser rangefinder, flow sensor, and vibration detection module deployed in the connecting area, collecting the following parameters in real time: workpiece temperature field distribution, roller conveyor geometry, cooling medium flow rate, and mechanical vibration spectrum. The decision-making layer includes edge computing nodes and a process optimization platform. The edge nodes are responsible for generating real-time compensation instructions, while the cloud platform enables in-depth analysis and strategy optimization. The execution layer includes a water / air spray control system, a roller conveyor drive system, and an alarm output module. The system communicates with the PLC system via a PROFIBUS-DP bus to achieve coordinated control with other equipment on the rolling line. All process data is stored in a time-series database, supporting long-term quality analysis and process improvement.

[0109] Preferably, the infrared temperature measurement array uses a mid-wave infrared detector with a wavelength of 3.9μm, a temperature measurement range of 400-1200℃, a spatial resolution of 1cm, and a frame rate of 60Hz. The array is arranged at a density of 2 groups per meter of rolling mill line, and water mist interference is eliminated through multispectral fusion technology. The innovation lies in the three-dimensional temperature field reconstruction algorithm, which combines two-dimensional temperature measurement data with the movement trajectory of the rolled piece to construct a dynamic temperature gradient model. This system overcomes the problem of unstable cooling effect and provides accurate process status information for intelligent compensation.

[0110] Step 2: Multimodal Process Condition Monitoring System

[0111] This module develops a process status assessment algorithm based on multi-sensor fusion. The system monitors the following key parameters in real time: temperature gradient at the head / tail of the rolled piece, uniformity of cooling medium coverage, thermal deformation of the roller table, and vibration amplitude of the rolled piece. These parameters are fused and processed through a deep neural network to output a Process Stability Index (PSI). When the PSI falls below a threshold, a compensation mechanism is automatically triggered. The monitoring system pays special attention to the issue of "water causing roller deformation," predicting roller table deformation trends through thermo-mechanical coupling analysis. All monitoring data is visualized, enabling process engineers to quickly grasp the system status.

[0112] Preferably, the process stability index is a standardized score from 0 to 100, encompassing four dimensions: temperature consistency, cooling uniformity, equipment safety, and production rhythm. The index calculation uses the analytic hierarchy process (AHP) to determine weights and fuzzy logic to handle uncertainties. The innovation lies in the index prediction module, which uses an LSTM network to predict process trends over the next 30 seconds, achieving proactive control. The system encodes experience gained from repeated trials into evaluation rules, ensuring a high degree of alignment with production process requirements.

[0113] Step 3: Dynamic Compensation Control System

[0114] For empirical parameters such as "17 stands open with steel signal, then close after a 15-second delay," this embodiment designs an adaptive dynamic compensation algorithm. The system dynamically calculates the optimal compensation strategy based on real-time parameters such as workpiece position, speed, and temperature, including: water / air spray start / stop sequence, flow rate adjustment curve, and medium mixing ratio. The control algorithm employs a model predictive control (MPC) framework, optimizing the action sequence for the next 5 seconds within a 100ms control cycle to ensure continuous and stable cooling. The system supports multiple compensation modes, which can be flexibly switched according to steel grade, specifications, and quality requirements.

[0115] Preferably, the model predictive control algorithm comprises three core models: a workpiece heat conduction model, a cooling medium flow model, and a equipment dynamic response model. The innovation lies in the online parameter identification mechanism, which continuously calibrates model parameters using real-time data to adapt to process changes. The control algorithm uses empirical delay parameters (such as "delay 5 seconds to turn on, delay 10 seconds to turn off") as initial values, and then dynamically optimizes them based on actual results, forming a continuously improving control strategy.

[0116] Step 4: Roller Conveyor Thermal Deformation Protection System

[0117] To address the issue of water causing roller deformation, this embodiment develops a real-time monitoring and protection system for roller conveyor thermal deformation. The system monitors the axial temperature distribution of the roller conveyor using distributed fiber optic temperature measurement technology, and calculates the thermal stress field and deformation amount using finite element analysis. When the predicted deformation exceeds a safety threshold, the system automatically adjusts the cooling strategy or triggers roller conveyor rotation to evenly distribute the heat load. The system establishes a thermal history file for each roller conveyor, calculates the cumulative heat load, and guides preventative maintenance plans.

[0118] Preferably, the distributed fiber optic temperature measurement system employs OFDR technology, with a spatial resolution of 1 cm and a temperature measurement accuracy of ±1℃. The optical fiber is spirally wound along the roller conveyor axis to comprehensively monitor the temperature field. The innovation lies in the thermal deformation prediction algorithm, which considers influencing factors such as material creep characteristics and the number of cooling cycles to achieve high-precision prediction. The system can automatically identify risk patterns such as roller conveyor deformation and take protective measures in advance to avoid production interruptions and equipment damage.

[0119] Step 5: Multi-objective optimization control platform

[0120] This embodiment constructs a multi-objective optimization platform considering quality, efficiency, and safety. Optimization objectives include temperature control accuracy, yield, energy consumption, and equipment lifespan. The platform uses the NSGA-II genetic algorithm to solve for the Pareto optimal solution set and selects the optimal equilibrium point based on the production plan. The optimization results automatically generate control parameters and are sent to the execution system. The platform supports both offline simulation and online optimization modes, facilitating process testing and parameter tuning. Through continuous optimization, the system addresses the core needs of improving quality and yield.

[0121] Preferably, the multi-objective optimization platform adopts a cloud computing architecture, supporting large-scale parallel computing. Its innovation lies in a knowledge-guided optimization strategy, which encodes process experience (such as the evolution of water spraying, air spraying, and water spraying in air-cooled roller conveyors) into optimization constraints, reducing the search space and improving optimization efficiency. The platform provides interactive visualization tools to help process engineers understand the trade-offs between various objectives and make informed decisions.

[0122] Step Six: Safety Interlock Protection System

[0123] To ensure safety, this embodiment employs a multi-layered safety interlocking protection system. Primary protection monitors process parameters to prevent anomalies; intermediate protection reduces speed upon detecting dangerous signs; and advanced protection triggers a full-line emergency stop and initiates emergency cooling. Safety functions are designed to IEC 61508 SIL2 level, with a response time of <100ms. The system pays special attention to risks such as scrap steel in the wire-spinning area, incorporating dedicated anti-scraping control logic and emergency response procedures.

[0124] Preferably, the anti-stacking control logic comprises three subsystems: roll tracking, speed matching, and gap control. The innovation lies in the machine vision-based precise roll head and tail positioning technology, achieving a positioning accuracy of ±10mm. The system is deeply integrated with hardware points such as the 17 steel-holding signals and the finishing mill steel-holding signals, forming a reliable interlocking protection network. All safety functions are regularly tested and verified to ensure effective prevention of dangerous situations under any operating conditions.

[0125] Step 7: Fault Diagnosis and Self-Healing System

[0126] When the system detects an anomaly, it automatically initiates the diagnostic process. The system has a pre-built feature library of common fault modes, such as nozzle blockage, pipeline leakage, and sensor failure. The diagnostic process combines real-time data and equipment history records, using a Bayesian network to calculate the probability distribution of various fault causes and generate a diagnostic report. For repairable faults, the system automatically executes preset self-healing procedures, such as switching to backup nozzles and adjusting backup pipelines, to minimize production interruptions.

[0127] Preferably, the Bayesian diagnostic network employs a dynamic structure learning algorithm, where nodes represent fault symptoms or root causes, and edges represent causal relationships. Network parameters are obtained through training on historical fault data and continuously optimized with new cases. The innovation lies in the network's integration of equipment structural knowledge (such as electrical cable layout) and process knowledge, improving diagnostic accuracy. The system can automatically identify various problem patterns and recommend validated solutions.

[0128] Step 8: Process Knowledge Management System

[0129] This embodiment constructs a structured process knowledge base, transforming "long-term on-site observation" experience into reusable knowledge assets. The knowledge base includes the following: typical process cases, troubleshooting solutions, parameter optimization records, and equipment maintenance archives. The knowledge is organized using an ontology-based approach, supporting semantic retrieval and associative reasoning. The system provides knowledge mining tools to discover potential patterns from massive amounts of process data, continuously enriching the knowledge base content.

[0130] Preferably, the process knowledge ontology defines core concepts such as rolled products, equipment, and process parameters, as well as their relationships. The innovation lies in the knowledge verification mechanism; new knowledge must be verified through simulation or actual production before being added to the database. The system uses technological evolution processes (such as the improvement process from water spraying to air spraying back to water spraying) as typical knowledge cases to help process engineers deeply understand technical principles and optimization directions.

[0131] Step Nine: Digital Twin Simulation Platform

[0132] To verify the system's effectiveness, this embodiment constructs a digital twin model of the controlled cooling system. The model accurately simulates the thermal-fluid-structure interaction behavior of the cooling process using multiphysics simulation technology. The platform supports fault injection testing, such as simulating the impact of nozzle blockage on cooling uniformity, to verify the system's compensation strategy. Operators can practice anomaly handling skills in a virtual environment without affecting actual production. The digital twin system is also updated synchronously with actual production line data to maintain model accuracy.

[0133] Preferably, the multiphysics simulation employs the ANSYS-MATLAB co-simulation platform, integrating computational fluid dynamics (cooling medium flow), heat transfer (roller temperature field), and structural mechanics (roller deformation) analysis. Model parameters are calibrated using actual measurement data, such as process parameters under various operating conditions. The simulation platform supports virtual experiments to explore the feasibility of innovative solutions, such as water spraying on the second section of the air-cooled roller conveyor, reducing the cost and risk of actual testing.

[0134] The intelligent compensation system employed in this embodiment reduces temperature fluctuations in the water-cooled / air-cooled transition zone by over 70%, completely resolving the issue of uncooled sections, increasing the yield by over 1.2 percentage points, and significantly improving process quality. Multi-level safety protection effectively avoids risks associated with scrap steel handling and processing high-temperature rolled pieces in the wire-rolling area, reducing the frequency of high-risk operations. The automation system reduces scrap steel hoisting, cutting, and related processing work, freeing up valuable human resources for higher-value tasks. The heat deformation protection system extends the roller conveyor's service life by over 30%, significantly reducing equipment maintenance costs and downtime losses.

[0135] In summary, after reading this invention document, those skilled in the art can make various other corresponding modifications to the technical solutions and concepts based on this invention without creative mental effort, and all of these modifications fall within the scope of protection of this invention.

Claims

1. A method for intelligent connection control of cooling medium in a controlled cooling system, characterized in that, include: Acquire multimodal process parameters collected by the sensor array deployed in the area where water cooling and air cooling meet; Based on the multimodal process parameters, determine the process stability index; When the process stability index is lower than a preset threshold, a dynamic compensation mechanism is triggered to generate a dynamic compensation control command based on the real-time position, real-time speed and real-time temperature of the rolled piece. The dynamic compensation control command is executed to adjust at least one of the following: the start-stop sequence of water spray and air spray, the flow rate adjustment curve, or the medium mixing ratio, so as to connect the water cooling and air cooling processes.

2. The intelligent connection control method for cooling medium in a controlled cooling system according to claim 1, characterized in that, The acquisition of multimodal process parameters collected by the sensor array deployed in the water-cooling and air-cooling junction area includes: The temperature field distribution data of the rolled piece collected by the infrared temperature measurement array is obtained, and the two-dimensional temperature measurement data is combined with the movement trajectory of the rolled piece to perform three-dimensional reconstruction of the temperature field and construct a dynamic temperature gradient model. Acquire the geometric shape data of the roller conveyor collected by the laser rangefinder; Acquire cooling medium flow data collected by the flow sensor; Acquire mechanical vibration spectrum data collected by the vibration detection module.

3. The intelligent connection control method for cooling medium in a controlled cooling system according to claim 1, characterized in that, The step of determining the process stability index based on the multimodal process parameters includes: The multimodal process parameters are fused and processed by a deep neural network to output a process stability index, which is a standardized score that includes four dimensions: temperature consistency, cooling uniformity, equipment safety, and production rhythm. The process stability index is adjusted based on the prediction results by using a long short-term memory network to predict the process trend within a preset time period.

4. The intelligent connection control method for cooling medium in a controlled cooling system according to claim 1, characterized in that, The generation of dynamic compensation control commands includes: The preset empirical delay parameters are used as initial values ​​and input into the model predictive control framework. Within a preset control cycle, the action sequence for a future preset time period is optimized by using the workpiece heat conduction model, the cooling medium flow model, and the equipment dynamic response model. The parameters of the workpiece heat conduction model, cooling medium flow model, and equipment dynamic response model are identified online using real-time data to update the model parameters. Based on the optimization results, dynamic compensation control commands are generated.

5. The intelligent connection control method for cooling medium in a controlled cooling system according to claim 1, characterized in that, Also includes: Acquire axial temperature distribution data of the roller conveyor monitored by the distributed fiber optic temperature measurement system; Based on the axial temperature distribution data of the roller conveyor, the thermal stress field and predicted deformation of the roller conveyor are calculated by finite element analysis. When the predicted deformation exceeds the safety threshold, the cooling strategy is adjusted or the roller conveyor is triggered to rotate.

6. The intelligent connection control method for cooling medium in a controlled cooling system according to claim 1, characterized in that, Also includes: A multi-objective optimization algorithm is adopted, with temperature control accuracy, yield, energy consumption and equipment life as optimization objectives, to solve for the Pareto optimal solution set; Control parameters are selected from the Pareto optimal solution set according to the production plan and then sent to the execution system.

7. The intelligent connection control method for cooling medium in a controlled cooling system according to claim 1, characterized in that, Also includes: When an anomaly is detected, the fault diagnosis process is initiated, real-time data and device history records are input into the Bayesian diagnostic network, the probability distribution of each fault cause is calculated, and a diagnostic report is generated. When the fault is repairable, a pre-defined self-healing procedure is executed.

8. A smart control system for the cooling medium in a cooling system, characterized in that, include: The sensing and detection module is deployed in the area where water cooling and air cooling meet, and is used to collect multimodal process parameters; The decision control module is used to determine the process stability index based on the multimodal process parameters. When the process stability index is lower than the preset threshold, a dynamic compensation mechanism is triggered, and a dynamic compensation control command is generated based on the real-time position, real-time speed and real-time temperature of the rolled piece. The execution module is used to execute the dynamic compensation control command and adjust at least one of the following: the start-stop sequence of water spray and air spray, the flow rate adjustment curve, or the medium mixing ratio, so as to connect the water cooling and air cooling processes.

9. The intelligent connection control system for cooling medium in a cooling system according to claim 8, characterized in that, The sensing and detection module includes: Infrared temperature measurement array is used to collect temperature field distribution data of rolled parts; Laser rangefinders are used to collect geometric data of the roller conveyor. Flow sensor, used to collect cooling medium flow data; Vibration detection module, used to collect mechanical vibration spectrum data; The decision control module is also used to combine the two-dimensional temperature measurement data collected by the infrared temperature measurement array with the movement trajectory of the rolled piece to perform three-dimensional reconstruction of the temperature field and construct a dynamic temperature gradient model.

10. The intelligent connection control system for cooling medium in a cooling system according to claim 8, characterized in that, The decision control module includes: Long Short-Term Memory (LSTM) network units are used to predict process trends within a preset time period and adjust the process stability index based on the prediction results. The online parameter identification subunit is used to update the parameters of the workpiece heat conduction model, cooling medium flow model, and equipment dynamic response model using real-time data.

Images