Online detection and automatic correction control method for cold spraying water flow of hot continuous rolling layer

Abstract

The invention belongs to the technical field of ferrous metallurgy, and provides a hot continuous rolling layer cold spraying water flow on-line detection and automatic correction control method, which comprises the following steps: adding an execution state closed loop on the basis of a traditional temperature feedback closed loop, comparing the actual spraying state of a layer cooling area with a cooling model point by point, and utilizing a set water flow state and an actually-measured water flow state to determine the temperature of the layer cooling area. Difference is calculated, a normalized index is formed, and when the index exceeds a threshold value, a control action is automatically triggered; when it is detected that water is not sprayed according to the setting or spraying is abnormal, supplementary spraying is conducted immediately, and it is guaranteed that the strip steel cooling curve and the coiling temperature are stable and consistent; according to the method, a double-closed-loop control system of a temperature result closed loop and an execution state closed loop is formed, the laminar cooling control precision, the response speed and the robustness are remarkably improved, energy consumption and manual intervention are reduced, the consistency of the strip steel coiling temperature and the product quality is guaranteed, and the wide industrial popularization value is achieved.

Description

Technical Field

[0001] This invention belongs to the field of iron and steel metallurgy technology, and relates to the intelligent management of temperature control and spraying system in the cold zone of the finishing mill exit layer of a hot continuous rolling coil production line. Specifically, it is a method for online detection and automatic correction control of cold spray water flow in the hot continuous rolling layer. Background Technology

[0002] Hot strip rolling is a core process for steel companies producing high-quality strip and plate products. The laminar cooling process from the finishing mill exit to coiling plays a crucial role in the strip's microstructure, mechanical properties, and straightness. Traditional laminar cooling control primarily relies on preset cooling models and temperature feedback to adjust the water volume, nozzle opening, and strip speed in each zone. While this can achieve some degree of coiling temperature control, under complex conditions such as high-speed operation, multi-variety switching, ultra-wide specifications, and high-strength steel production, discrepancies often arise between the preset spray pattern and the actual spray pattern (i.e., "empty setup"). This leads to cooling curve deviations, product performance fluctuations, and energy waste.

[0003] In actual operation, existing laminar cooling control systems suffer from various potential sources of failure in the laminar cooling spray system. These include damage to the laminar cooling water valve itself, burnt-out coils or solenoid coils, aging or open circuits in the control circuitry, insufficient air pressure, and nozzle blockage or uneven distribution. When such faults occur, although the automated system still outputs spray commands according to the cooling model, it cannot obtain the actual spray status of the nozzles or water outlet chamber in real time. In other words, existing automated control only "issues commands" but cannot "monitor execution," leading to a widespread phenomenon of "empty setup" where the settings do not match reality. As a result, the cooling curve and coiling temperature deviate, the strip's microstructure and properties fluctuate more, and energy and water resources are wasted inefficiently.

[0004] Existing systems largely rely on temperature deviations to indirectly infer spray anomalies, exhibiting significant lag. Once a local valve or nozzle fails, the temperature sensor requires considerable time and distance to detect the deviation, often only being discovered after the strip has already passed the critical cooling zone. This prevents timely adjustments to water flow or switching to backup spray systems. This not only reduces control accuracy but also increases the frequency of manual intervention and maintenance costs. Furthermore, existing alarm methods are mostly centralized or total alarms, making it difficult to accurately and promptly locate faulty valves or manifolds. Maintenance personnel must conduct on-site inspections, disrupting production cycles.

[0005] Furthermore, existing temperature measurement devices are mostly located at the inlet or outlet of the laminar flow cooling zone, making it impossible to directly observe the spray status. The actual operating conditions of the spray valves and nozzles lack visualization and real-time feedback. Even when some flow meters or pressure gauges are introduced, they are mostly zone-level or total flow-level, failing to provide direct evidence at the nozzles. Summary of the Invention

[0006] The purpose of this invention is to provide an online detection and automatic correction control method for cold spray water flow in hot continuous rolling layers. This method can determine whether the spraying is executed according to the settings, compare the execution results with the cooling model settings, and automatically supplement the spraying or switch to the backup spraying, thereby fundamentally solving the problems of inconsistency between automation commands and actual execution, and the inability of the process to be automatically adjusted.

[0007] The technical solution adopted by the present invention to achieve the above objectives is as follows:

[0008] A method for online detection and automatic correction control of cold spray water flow in hot continuous rolling layers includes:

[0009] S1. Temperature acquisition devices are arranged at multiple points in the cooling zone, and laser diffuse reflection light curtain sensors are arranged on the side of the water outlet chamber of each cooling pipe.

[0010] S2. Collect the strip surface temperature and laser diffuse reflection light curtain signals and process parameters to form a complete real-time dataset;

[0011] S3. By using a layered cooling model based on heat transfer and cooling characteristics built within the control system, the surface temperature deviation ΔT(x,t) of the strip steel and the idle index AI(t) of the spray are calculated and compared to monitor the set state and actual execution of the spray.

[0012] The empty field index AI(t) for spraying is to normalize the differences among all sprays:

[0013]

[0014] Where N is the number of nozzles. Sset,i(t) represents the preset water flow state; Smeas,i(t) represents the measured water flow state; Sset,i(t) and Smeas,i(t) can be 0 / 1 to represent the switch state.

[0015] S4. When the temperature deviation exceeds the threshold or the idling index exceeds the threshold, the fault valve is triggered for secondary setting.

[0016] The fault valve control logic is as follows:

[0017] S41. Calculate the corrected control value. The calculation formula is as follows:

[0018]

[0019] Where ui(t) is the control quantity after correction for the i-th partition, and kT,kS are the adjustment coefficients;

[0020] S42. Quickly increase the water flow rate for supplementary spraying based on the corrected control quantity;

[0021] S5. Periodically perform regression analysis on historical data, and the formula for automatically correcting the model parameter θ is as follows:

[0022]

[0023] Where L is the loss function between the predicted temperature Tpred and the measured temperature Tmeas; η is the learning rate.

[0024] Furthermore, the temperature acquisition device employs three sets of infrared scanning thermometers, installed at the inlet, middle, and outlet of the cooling zone, respectively, covering the entire width of the strip and acquiring surface temperature at a frequency of 10Hz or higher.

[0025] Furthermore, the status of the faulty manifold or nozzle zone is displayed on the visual interface, and alarm prompts are provided.

[0026] Further methods to quickly increase water flow include adjusting the opening of the spray valves or increasing the number of sprays.

[0027] The beneficial effects of this invention are:

[0028] Compared with existing technologies, this invention innovatively introduces a laser diffuse reflection light curtain into the cold zone of the finishing mill exit layer in a hot strip rolling mill production line. This enables real-time online detection of the spray water flow from each nozzle or outlet chamber, expanding the traditional single temperature feedback closed loop into a dual closed-loop control system of "temperature result closed loop + execution status closed loop". Its beneficial effects are mainly reflected in the following aspects:

[0029] 1. Completely eliminates the "idle" phenomenon: Traditional systems only output spray commands and lack direct monitoring of the actual spray status. Once the valve body is damaged, the coil burns out, the circuit is broken, or the air source pressure is insufficient, the "spraying is set but not actually spraying" phenomenon can easily occur. This invention directly determines whether the spraying is executed according to the setting through light curtain detection and compares it point by point with the cooling model. Once an abnormality is detected, it immediately triggers automatic supplementary spraying or switches to the backup spray, eliminating the temperature deviation and performance fluctuation caused by idling from the source.

[0030] 2. Significantly improve temperature control accuracy and response speed: Since abnormalities in the spraying state can be detected before temperature deviations appear, the control system of this invention can intervene in advance and quickly adjust valve opening, water volume and zone spraying to achieve dynamic tracking of winding temperature and cooling curve, thereby improving temperature control accuracy and shortening response time.

[0031] 3. Real-time fault location and visual alarm: The system intuitively displays the status of faulty manifolds or nozzle zones on the monitoring interface, using different colors or symbols to indicate normal and abnormal spraying, and automatically records historical data and alarm information. Maintenance personnel can quickly and accurately locate the fault point, reducing manual inspection time and misjudgment, and improving operation and maintenance efficiency.

[0032] 4. Self-learning optimization enhances model robustness: This invention utilizes historical data to perform regression or machine learning optimization on cooling model parameters, achieving adaptive model correction and maintaining high consistency in prediction and control under complex working conditions such as different steel grades, thicknesses, and speeds, thereby further improving product quality stability.

[0033] 5. Reduce energy consumption and manual intervention, and improve the level of intelligent production: Through dual closed-loop control, the system can adjust the spray as needed, avoiding the "over-control" phenomenon of excessive water addition due to local spray failure, reducing water and energy consumption, while reducing manual intervention and adjustment, and improving the automation and intelligence level of the production line. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of the laser diffuse reflection light curtain sensor setup in this invention;

[0035] In the diagram: 1. Cooling rack; 2. Laser diffuse reflection light curtain sensor; 3. Laser; 4. Cooling pipe; 5. Water flow; 6. Control system; 7. Visual monitoring interface. Detailed Implementation

[0036] The present invention will be further described below with reference to the embodiments and accompanying drawings:

[0037] Example

[0038] On a hot-rolled coil production line, the online detection and automatic correction control method for the cold spray water flow of the hot-rolled layers includes:

[0039] S1, such as Figure 1 As shown, temperature acquisition devices are arranged at multiple points in the laminar cooling zone from the exit of the finishing mill stand to the front of the coiler, and laser diffuse reflection light curtain sensors 2 are installed on the side of the water outlet chamber of the laminar cooling pipe 4 of each laminar cooling stand 1.

[0040] The temperature acquisition device uses three sets of infrared scanning thermometers, which are installed at the inlet, middle and outlet of the cooling zone, respectively. The range covers the entire width of the strip and the surface temperature Tmeas(x,t) is collected at a frequency of 10Hz or higher.

[0041] The laser diffuse reflection light curtain sensor 2 uses laser 3 to irradiate water flow 5, detects the state of the spray water flow Smeas,i(t), and determines whether each spray is spraying water according to the set parameters.

[0042] S2. Collect the strip surface temperature and laser diffuse reflection light curtain signals and process parameters to form a complete real-time dataset;

[0043] Process parameters include steel type, thickness, speed, target cooling curve, and spray opening.

[0044] All data is transmitted to control system 6 via fieldbus, providing a basis for model calculation and control decisions;

[0045] S3. By using a layered cooling model based on heat transfer and cooling characteristics built inside the control system, the surface temperature deviation ΔT(x,t) of the strip steel and the idle index AI(t) of the spray are calculated and compared to monitor the spray setting state and actual execution.

[0046] For the instantaneous temperature of the strip unit, the following simplified energy balance formula can be used:

[0047]

[0048] Where hi is the heat transfer coefficient of the i-th partition, Ai is the cooling area, m is the mass of the steel, Cp is the specific heat capacity, and Twater is the cooling water temperature.

[0049] For each spray zone, the model calculates the deviation between the predicted temperature Tpred and the measured temperature Tmeas:

[0050]

[0051] For each spray state, define the "set-actual spray" difference and calculate the spray state difference di(t):

[0052]

[0053] Sset,i(t) is the preset spray water flow state; Smeas,i(t) is the detected spray water flow state; Sset,i(t) and Smeas,i(t) can be 0 / 1 to indicate the on / off state.

[0054] The empty exponent AI(t) is used to normalize the differences in all spray conditions:

[0055]

[0056] Where N is the number of sprayers.

[0057] S4. When the temperature deviation exceeds the threshold |ΔT|>ΔTmax or the idling index exceeds the threshold AI(t)>AImax, the fault valve secondary setting is triggered.

[0058] The logic for the secondary setting of the fault valve is as follows:

[0059] S41. Calculate the corrected control value. The calculation formula is as follows:

[0060]

[0061] Where ui(t) is the control quantity after correction for the i-th partition, and kT,kS are the adjustment coefficients;

[0062] S42. Based on the corrected control quantity, quickly add a set of cold water layers or switch to the standby spray to perform immediate replenishment spraying and ensure that the strip coiling temperature remains stable within the set range.

[0063] S5. Periodically perform regression analysis on historical data, and the formula for automatically correcting the model parameter θ is as follows:

[0064]

[0065] Where L is the loss function between the predicted temperature Tpred and the measured temperature Tmeas; η is the learning rate.

[0066] Regularly performing regression analysis or machine learning modeling on historical data can improve prediction accuracy and control stability, adapting to changes in different steel grades, thicknesses, and production rhythms. In this way, the laminar cooling model can continuously adaptively optimize during long-term operation, improving prediction-control consistency.

[0067] In addition, the visual monitoring interface 7 displays the working status of each sprinkler in real time with different colors, records the sprinkler status, temperature distribution, vacancy index and correction amount, and pops up an alarm prompt when a fault occurs, marking the location of the faulty manifold or sprinkler zone to remind maintenance personnel to handle it in a timely manner.

[0068] This invention adds an "execution state" closed loop to the traditional temperature feedback closed loop. It compares the actual spraying state in the laminar cooling zone with the cooling model settings point by point. Using the set state Sset,i and the measured state Smeas,i, it calculates the difference di(t) = |Sset,i−Smeas,i| and forms a normalized index AI(t) = ∑di(t) / N. When AI(t) exceeds a threshold, it automatically triggers a control action. When it detects that the water spraying is not according to the settings or an abnormal spraying pattern, the system automatically adds a set of laminar cooling water or switches to a backup spray, achieving immediate supplementary spraying and ensuring that the strip cooling curve is stable and consistent with the coiling temperature.

[0069] In addition, by performing regression or machine learning modeling on historical operating data, the cooling model parameters are automatically corrected to improve prediction accuracy and control stability, adapting to different steel grades, thicknesses and production rhythms.

[0070] Field operation results show that after adopting the method of the present invention, the temperature control accuracy of strip steel outlet is improved, the plate shape defects are reduced, the cooling water consumption is reduced, the spray fault location time is shortened, the accuracy, response speed and robustness of layer cooling control are significantly improved, energy consumption and manual intervention are reduced, and the intelligence level of the production line and the consistency of product quality are improved.

Claims

1. A method for online detection and automatic correction control of cold spray water flow in hot continuous rolling layers, characterized in that: include: S1. Temperature acquisition devices are arranged at multiple points in the laminar cooling zone from the exit of the finishing mill stand to the front of the coiler, and laser diffuse reflection light curtain sensors are arranged on the side of the water outlet chamber of each laminar cooling pipe. S2. Collect the strip surface temperature and laser diffuse reflection light curtain signals and process parameters to form a complete real-time dataset; S3. By using a layered cooling model based on heat transfer and cooling characteristics built within the control system, the surface temperature deviation ΔT(x,t) of the strip steel and the idle index AI(t) of the spray are calculated and compared to monitor the set state and actual execution of the spray. The vacancy index AI(t) of the spray is used to normalize the differences among all sprays: Where N is the number of nozzles, and di(t) is the difference in spray conditions: Sset,i(t) is the preset spray water flow state; Smeas,i(t) is the detected nozzle water flow state; Sset,i(t) and Smeas,i(t) can be 0 / 1 to indicate the on / off state. S4. When the temperature deviation exceeds the threshold or the idling index exceeds the threshold, the fault valve is triggered for secondary setting. S5. Periodically perform regression analysis on historical data, and the formula for automatically correcting the model parameter θ is as follows:

2. Where L is the loss function between the predicted temperature Tpred and the measured temperature Tmeas; η is the learning rate.

3. The method for online detection and automatic correction control of cold spray water flow in hot continuous rolling layers according to claim 1, characterized in that: The temperature acquisition device uses three sets of infrared scanning thermometers, which are installed at the inlet, middle and outlet of the cooling zone, respectively, covering the entire width of the strip and acquiring the surface temperature at a frequency of 10Hz or higher.

4. The method for online detection and automatic correction control of cold spray water flow in hot continuous rolling layers according to claim 1, characterized in that: The logic for the secondary setting of the fault valve in S4 is as follows: S41. Calculate the corrected control value. The calculation formula is as follows: Where u i (t) represents the corrected control quantity for the i-th partition, k T ,k S This is the adjustment coefficient; S42. Quickly increase the water flow rate for supplementary spraying based on the corrected control quantity.

5. The method for online detection and automatic correction control of cold spray water flow in hot continuous rolling layers according to claim 3, characterized in that: The methods for rapidly increasing water flow include adjusting valve opening or increasing spray.

6. The method for online detection and automatic correction control of cold spray water flow in hot continuous rolling layers according to claim 1, characterized in that: The control system displays the status of faulty manifolds or sprinklers on a visual interface and provides alarm prompts.

Images