Parameter self-adaptive based intelligent control method and system for ceramic spraying

Abstract

The embodiment of the application relates to the computer technical field, and specifically discloses a ceramic spraying intelligent regulation and control method and system based on parameter self-adaption, which realizes global optimization control of a ceramic spraying flow line process through integration of real-time collection and dynamic regulation and control mechanism of multi-dimensional equipment parameters. The embodiment of the application breaks through the limitation of traditional single-device independent control, constructs an intelligent spraying control system with environment self-adaptive characteristics through parameter collaborative optimization of equipment groups and bidirectional data interaction of an MES system, and realizes comprehensive control processing with multi-parameter collaborative optimization and system-level intelligent decision-making capability.

Description

Technical Field

[0001] This application relates to the field of computer technology, specifically to a method and system for intelligent control of ceramic spraying based on parameter adaptation. Background Technology

[0002] In the field of industrial ceramics manufacturing, automated spraying control technology is a core element in achieving uniformity of product surface coatings and process stability. Current mainstream spraying production lines typically integrate key equipment such as burners, pressure pumps, and circulating fans, using digital panels for basic parameter setting and status monitoring. With increasingly stringent environmental requirements and the development of intelligent manufacturing, the industry's demand for precise control of spraying processes has shifted from optimizing the performance of single equipment to multi-equipment collaborative control and comprehensive safety and energy efficiency management.

[0003] Current technologies generally employ independent PID control modules to separately regulate the burner fuel supply, water pump pressure, and fan speed. This distributed control architecture struggles to effectively handle the dynamic coupling effects between equipment. More significantly, existing spraying control systems and MES (Manufacturing Execution System) production management systems suffer from data silos, resulting in a disconnect between process parameter optimization and production scheduling decisions. Electrical control cabinet configuration strategies are often based on fixed empirical values, failing to dynamically respond to changes in raw material characteristics or environmental disturbances, leading to difficulties in effectively controlling energy consumption and scrap rates. These technical deficiencies severely restrict the application effectiveness of ceramic spraying processes in intelligent manufacturing scenarios. Therefore, there is an urgent need to develop a comprehensive control scheme with multi-parameter collaborative optimization and system-level intelligent decision-making capabilities. Summary of the Invention

[0004] This application provides a method and system for intelligent control of ceramic spraying based on parameter adaptation, which can realize comprehensive control processing with multi-parameter collaborative optimization and system-level intelligent decision-making capabilities.

[0005] In a first aspect, embodiments of this application provide a parameter-adaptive intelligent control method for ceramic spraying, applied to an intelligent control system for ceramic spraying. The method includes: acquiring a set of equipment operating states from a digital panel of the spraying production line process, the set of equipment operating states including fuel supply efficiency and combustion temperature data of the burner, water pressure fluctuation data of the water pump, speed deviation value of the fan, water and gas consumption data, and harmful gas concentration monitoring indicators; calling a trained adaptive control model to perform dynamic parameter analysis processing on the set of equipment operating states, generating fuel supply control parameters for the burner, water pressure compensation parameters for the water pump, and speed correction parameters for the fan; based on... The fuel supply control parameters dynamically adjust the fuel valves of the burner, the water pressure compensation parameters dynamically calibrate the motor power of the water pump, and the speed correction parameters dynamically adjust the blade angle of the blower. Real-time data collection is also used to update the equipment operating status set. The updated equipment operating status set, the harmful gas concentration monitoring indicators, and the powder booth safety indicators are synchronized to the MES system, triggering the MES system to optimize the production data configuration of the electrical control cabinet and update the real-time process parameters in the digital process panel of the spraying production line.

[0006] Secondly, embodiments of this application provide a ceramic spraying intelligent control system, comprising:

[0007] processor;

[0008] Storage device, on which computer programs are stored,

[0009] When the computer program is executed by the processor, the processor enables the processor to implement any of the aforementioned intelligent control methods for ceramic spraying based on parameter adaptation.

[0010] This application provides a readable storage medium storing a program or instructions, which, when executed by a processor, implement the steps of the parameter-adaptive intelligent control method for ceramic spraying.

[0011] Therefore, the embodiments of this application have the following beneficial effects: By integrating a real-time acquisition and dynamic control mechanism for multi-dimensional equipment parameters, global optimization control of the ceramic spraying production line process is achieved. Through the adaptive control model's collaborative analysis of burner fuel supply efficiency, water pump pressure fluctuations, and fan speed deviations, precise control parameters matching different process stages can be dynamically generated under complex operating conditions. The adaptive control model performs nonlinear parameter compensation based on the coupling relationship of equipment operating states, effectively suppressing the interference of combustion temperature fluctuations on spray uniformity. Simultaneously, real-time calibration of water pressure compensation parameters reduces the impact of pulsation effects on atomization quality. The dynamic adjustment mechanism of the fan blade angle, combined with speed correction parameters, can simultaneously optimize airflow distribution and powder deposition efficiency, ensuring the stability of coating thickness. Through continuous iterative updates of fuel flow feedback data, a self-learning capability for equipment operating states is formed, significantly improving the dynamic response accuracy of the control strategy. Furthermore, the synchronous integration of harmful gas concentration monitoring indicators and powder room safety indicators achieves multi-dimensional mapping between process safety parameters and MES production data, enabling the optimized configuration of the electrical control cabinet to both ensure production continuity and proactively avoid environmental risks. With this design, the embodiments of this application break through the limitations of traditional single-device independent control. Through parameter collaborative optimization of device groups and bidirectional data interaction with the MES system, an intelligent spraying control system with environmental adaptive characteristics is constructed, realizing comprehensive control processing with multi-parameter collaborative optimization and system-level intelligent decision-making capabilities. Attached Figure Description

[0012] Figure 1 This is a flowchart of a parameter-adaptive intelligent control method for ceramic spraying provided in an embodiment of this application.

[0013] Figure 2 This is a schematic diagram of the basic structure of a ceramic spraying intelligent control system provided in an embodiment of this application. Detailed Implementation

[0014] See Figure 1 As shown, this figure is a flowchart of a parameter-adaptive intelligent control method for ceramic spraying provided in an embodiment of this application. This method can be applied to an intelligent control system for ceramic spraying. Figure 1 As shown, the method includes steps S110-S140.

[0015] Step S110: Obtain the set of equipment operating statuses from the digital panel of the spraying production line process. The set of equipment operating statuses includes the fuel supply efficiency and combustion temperature data of the burner, the water pressure fluctuation data of the water pump, the speed deviation value of the fan, the water and gas consumption data, and the monitoring indicators of harmful gas concentration.

[0016] In a ceramic spraying production line scenario, the sensor network of the electrical control cabinet is responsible for collecting data from various equipment. For the burner, the sensor network collects real-time fuel flow rate data and combustion temperature data. For example, if the real-time fuel flow rate of the burner is collected as 50 liters / hour, and the preset fuel threshold is 45 liters / hour, the fuel supply efficiency is generated based on the difference between the real-time fuel flow rate data and the preset fuel threshold, i.e., 50-45=5 liters / hour. At the same time, if the combustion temperature is collected as 1200 degrees Celsius, this combustion temperature data is correlated with the fuel supply efficiency (those skilled in the art will know that the combustion temperature data and fuel supply efficiency are standardized during the correlation process to eliminate dimensional differences), forming the burner status characteristics.

[0017] For water pumps, the fluctuation difference between their input and output water pressure is collected. For example, if the input water pressure is 20 Pascals and the output water pressure is 25 Pascals, the fluctuation difference is 25-20=5 Pascals. Water pressure fluctuation data is generated by combining the historical average water flow rate and real-time water flow rate from the water and gas consumption data. If the historical average water flow rate is 30 cubic meters per hour and the real-time water flow rate is 32 cubic meters per hour, water pressure fluctuation data is generated by combining these data and analyzing the changing trends and their proportional relationships. Furthermore, those skilled in the art can further refine and detail the water pressure fluctuation data by combining it with time series analysis (such as sliding window statistical fluctuation standard deviation), which is not limited here.

[0018] For the fan, the matching degree between its real-time rotational speed data and the preset rotational speed curve is obtained. For example, if the fan's real-time rotational speed is 1500 rpm, and the preset rotational speed curve specifies that the speed should be 1450 rpm at that moment, the difference between the two is calculated to determine the rotational speed deviation value. Simultaneously, the nitrogen oxide concentration in the hazardous gas concentration monitoring index is correlated. For example, if the nitrogen oxide concentration is 50 ppm, the rotational speed deviation value is correlated with this nitrogen oxide concentration (those skilled in the art will know that the rotational speed deviation value and the nitrogen oxide concentration are standardized during the correlation process to eliminate dimensional differences), generating the fan's status characteristics.

[0019] Abnormal action events and the number of actions performed in the powder coating booth are extracted from safety indicators. For example, statistics show that there were 3 spraying interruptions, 1 instance of the robotic arm exceeding its limits, and 2 abnormal start-up and shutdown events of the ventilation system. These abnormal action events are then correlated and mapped with hazardous gas concentration monitoring indicators to generate powder coating booth environmental safety data.

[0020] Finally, the burner status characteristics, water pressure fluctuation data, fan status characteristics, water and gas consumption data, and powder room environmental safety data are integrated into a set of equipment operating status data and synchronized to the digital process panel of the spraying production line.

[0021] In an optional embodiment, the step of obtaining the set of equipment operating statuses in the digital panel of the spraying production line process includes:

[0022] Step S111: Collect real-time fuel flow data and combustion temperature data of the burner through the sensor network of the electrical control cabinet, generate fuel supply efficiency based on the difference between the real-time fuel flow data and the preset fuel threshold, and generate burner status characteristics by associating the combustion temperature data with the fuel supply efficiency.

[0023] In this ceramic coating production line, a sensor network continuously monitors the burner's operation. For example, if real-time fuel flow rate data is collected at 60 liters / hour and the preset fuel threshold is 55 liters / hour, the fuel supply efficiency (which can also be understood as fuel supply deviation or flow rate difference) is obtained by calculating 60-55=5 liters / hour. Simultaneously, the combustion temperature is collected at 1300 degrees Celsius. Correlation between the combustion temperature of 1300 degrees Celsius and the fuel supply efficiency of 5 liters / hour generates burner status characteristics, which include operational information regarding fuel supply and combustion temperature.

[0024] Step S112: Collect the fluctuation difference between the input and output water pressure of the water pump, and generate water pressure fluctuation data by combining the historical average water flow and real-time water flow in the water and gas consumption data.

[0025] During the same production process, operational data of the water pump is collected. For example, if the pump's input pressure is 22 Pascals and its output pressure is 28 Pascals, the fluctuation difference between the input and output pressures is 28 - 22 = 6 Pascals. In the water and gas consumption data, the historical average water flow rate is 35 cubic meters per hour, and the real-time water flow rate is 38 cubic meters per hour. Combining this fluctuation difference of 6 Pascals with the two water flow rate data points, and analyzing their numerical relationships and trends—such as considering the ratio of real-time water flow rate to the historical average and the variation pattern of the fluctuation difference throughout the production process—water pressure fluctuation data is generated. This data reflects the dynamic changes in water pressure experienced by the pump and helps to understand whether the pump's operating status is stable.

[0026] Step S113: Obtain the matching degree between the real-time speed data of the fan and the preset speed curve, determine the speed deviation value based on the matching degree, and generate the fan status characteristics by associating the nitrogen oxide concentration in the harmful gas concentration monitoring index.

[0027] Optionally, the fan operation is continuously observed, and the real-time fan speed is collected as 1600 rpm. The preset speed curve specifies that the speed should be 1550 rpm at this time. By comparing and analyzing the difference between the real-time speed and the preset speed curve at this moment, the matching degree is calculated. Based on this matching degree, the speed deviation value is determined. The speed deviation value can be determined by calculating the ratio of the difference between the two to the preset speed. For example, if the nitrogen oxide concentration in the harmful gas concentration monitoring index is 60 ppm at this time, the determined speed deviation value is correlated with the nitrogen oxide concentration of 60 ppm to generate a fan status characteristic. This characteristic integrates information on the fan speed and the nitrogen oxide concentration in the powder chamber, providing a basis for assessing the impact of the fan on the powder chamber environment.

[0028] Step S114: Extract abnormal action events and the number of powder room actions from the powder room safety indicators, and associate and map the abnormal action events with the harmful gas concentration monitoring indicators to generate powder room environmental safety data.

[0029] During the operation of the paint booth, safety indicators were analyzed. Statistics revealed four spraying interruptions, two instances of the robotic arm exceeding its limits, and three abnormal start-up / shutdown events in the ventilation system. These were identified as abnormal events. For example, the nitrogen oxide concentration, a key indicator for monitoring hazardous gas concentrations, fluctuated during this period, reaching a peak of 70 ppm. These abnormal events were correlated with changes in hazardous gas concentrations, such as analyzing the specific numerical changes and trends of nitrogen oxide concentration at each abnormal event. This generated paint booth environmental safety data, reflecting the overall safety status of the paint booth and crucial for ensuring its safe operation.

[0030] Step S115: Integrate the burner status characteristics, water pressure fluctuation data, fan status characteristics, water and gas consumption data, and powder room environmental safety data into a set of equipment operating status data, and synchronize it to the digital panel of the spraying production line process.

[0031] Optionally, the previously generated burner status characteristics (including information related to fuel supply efficiency and combustion temperature), water pressure fluctuation data (reflecting dynamic changes in water pump pressure), fan status characteristics (combining fan speed and nitrogen oxide concentration), water and gas consumption data (recording water and gas usage), and powder booth environmental safety data (reflecting powder booth safety-related information) can be integrated. For example, existing data structures (such as arrays and structures to store data from different devices) and algorithms (such as using a weighted average algorithm to fuse different types of data) can be used to combine the above-mentioned different types of data (standardization is required during the combination process to eliminate dimensional differences) to form a set of equipment operating statuses. Then, through the existing data transmission network, this set of equipment operating statuses is synchronized to the digital panel of the spraying production line process to achieve visualization of the set of equipment operating statuses, allowing operators to intuitively understand the operating status of the entire production equipment and providing comprehensive data support for subsequent control operations.

[0032] Step S120: Call the trained adaptive control model to perform dynamic parameter parsing processing on the set of equipment operating states, and generate fuel supply control parameters for the burner, water pressure compensation parameters for the water pump, and speed correction parameters for the fan.

[0033] In this ceramic coating production line scenario, after acquiring the set of equipment operating states, a trained adaptive control model is invoked for processing. This model possesses the ability to analyze input data and generate corresponding control parameters. First, the burner control branch of the adaptive control model processes the fuel supply efficiency and combustion temperature data from the set of equipment operating states. For example, if the fuel supply efficiency is 8 liters / hour and the combustion temperature is 1350 degrees Celsius, a multi-scale feature extraction method is used to analyze this data from different time scales and data dimensions, generating combustion fluctuation features. These features contain information on the changes in fuel supply and combustion temperature at different levels.

[0034] Next, the generated combustion fluctuation characteristics are input into the parallel fully connected layer of the burner control branch for nonlinear mapping, outputting the dynamic adjustment coefficient of the fuel valve opening and the combustion temperature compensation gradient. For example, the dynamic adjustment coefficient of the fuel valve opening is 1.2, and the combustion temperature compensation gradient is 0.5. Then, these two values ​​are fused to generate the fuel supply control parameters.

[0035] For the water pump compensation branch, fuel supply control parameters and water pressure fluctuation data are input together. The water pressure fluctuation data shows significant current fluctuations. A sliding time window is used to decompose the water pressure fluctuation data, separating the water pressure trend term and the water pressure residual term. For example, analysis shows that the water pressure trend term exhibits an upward trend, while the water pressure residual term shows some random fluctuations. The water pressure trend term and water pressure residual term are input into a bidirectional long short-term memory network for pressure fluctuation prediction, generating a water pressure fluctuation prediction result. For example, the prediction result shows that the water pressure will continue to rise in the near future. Simultaneously, the real-time water flow rate and historical average value from the water and gas consumption data are input into the difference calculation unit to generate a dynamic compensation coefficient. Finally, the water pressure fluctuation prediction result and the dynamic compensation coefficient are weighted and fused to generate the water pressure compensation parameters.

[0036] For the fan coordination module, water pressure compensation parameters and speed deviation values ​​are input. For example, if the speed deviation value is 50 rpm, a joint feature space is constructed between the speed deviation value and the hazardous gas concentration monitoring index. Topological association features between the fan nodes and the powder room gas concentration monitoring nodes in the joint feature space are extracted using a graph convolutional network. For example, a dynamic correlation pattern between fan speed changes and hazardous gas concentration in the powder room is analyzed. The topological association features are then input into a temporal attention mechanism for dynamic weighting to generate speed adjustment demand weights. Based on these weights, the ventilation compensation duration is calibrated using the water pressure compensation parameters, generating speed correction parameters that include blade angle adjustment, speed change rate, and ventilation compensation duration.

[0037] In an optional embodiment, a trained adaptive control model is invoked to perform dynamic parameter parsing processing on the set of equipment operating states, generating fuel supply control parameters for the burner, water pressure compensation parameters for the water pump, and speed correction parameters for the fan, including:

[0038] Step S121: The combustion efficiency and combustion temperature data are extracted using the combustion engine control branch of the adaptive control model to generate combustion fluctuation features; the combustion fluctuation features are input into the parallel fully connected layer of the combustion engine control branch for nonlinear mapping, and the dynamic adjustment coefficient of the fuel valve opening and the combustion temperature compensation gradient are output. The dynamic adjustment coefficient and the combustion temperature compensation gradient are then fused to generate the fuel supply control parameters.

[0039] In ceramic coating production lines, when the equipment operating status set is transmitted to the burner control branch of the adaptive control model, operations are performed on the fuel supply efficiency and combustion temperature data. For example, the current fuel supply efficiency is 10 liters / hour, and the combustion temperature is 1400 degrees Celsius. The multi-scale feature extraction process analyzes these data from different time scales and data feature dimensions. From a time scale perspective, it observes the rapid changes in fuel supply efficiency and combustion temperature over short periods, as well as the overall trends over long periods. From a data feature dimension perspective, it analyzes the correlation between fuel supply efficiency and combustion temperature, and their respective magnitudes of change. Through the above processing and analysis, combustion fluctuation characteristics can be generated.

[0040] Next, the generated combustion fluctuation characteristics are input into the parallel fully connected layer of the burner control branch. In this layer, through nonlinear calculations, such as transforming the input data using activation functions, the dynamic adjustment coefficient of the fuel valve opening and the combustion temperature compensation gradient are output. For example, the dynamic adjustment coefficient of the fuel valve opening is 1.5, and the combustion temperature compensation gradient is 0.8. Finally, these two values ​​are fused using existing methods, such as weighted fusion based on their importance in control, to generate fuel supply control parameters. These parameters comprehensively consider both fuel supply and combustion temperature, providing a basis for subsequent adjustment of the burner's fuel valves.

[0041] Step S122: Input the fuel supply control parameters and the water pressure fluctuation data into the pump compensation branch of the adaptive control model. Perform trend decomposition on the water pressure fluctuation data through a sliding time window to separate the water pressure trend term and the water pressure residual term. Input the water pressure trend term and the water pressure residual term into a bidirectional long short-term memory network for pressure fluctuation prediction to generate a water pressure fluctuation prediction result. Input the real-time water flow rate and historical average value in the water and gas consumption data into the difference calculation unit to generate a dynamic compensation coefficient. Perform weighted fusion of the water pressure fluctuation prediction result and the dynamic compensation coefficient to generate the water pressure compensation parameters.

[0042] In the pump compensation branch of the model, the previously generated fuel supply control parameters and the current water pressure fluctuation data are used as inputs. The water pressure fluctuation data exhibits a certain complexity, which can be decomposed using a sliding time window method. The sliding time window slides across the water pressure fluctuation data sequence, extracting a certain length of data segment for analysis each time. For example, a 5-minute time window is used to analyze the water pressure change trend within each time window. In this way, the water pressure trend term and the water pressure residual term are separated. The water pressure trend term reflects the overall direction of water pressure change, while the water pressure residual term contains some random, short-term fluctuations.

[0043] The separated water pressure trend term and water pressure residual term are input into a bidirectional long short-term memory (LSTM) network. This network can process time series data and predict subsequent water pressure fluctuations by learning from and analyzing historical data. For example, the bidirectional LSM network predicts that the water pressure will first rise and then slightly decrease within the next 10 minutes, which is the generated water pressure fluctuation prediction result.

[0044] Simultaneously, the real-time water flow rate and historical average value from the water and gas consumption data are input into the difference calculation unit. For example, if the real-time water flow rate is 45 cubic meters per hour and the historical average value is 40 cubic meters per hour, the difference is calculated as 45-40=5 cubic meters per hour. Then, this difference is converted into a dynamic compensation coefficient through a time-series weighted mapping rule (the time-series weight can be selected according to the actual situation and is not limited here).

[0045] Finally, the water pressure fluctuation prediction results are weighted and fused with the dynamic compensation coefficient. Based on the importance of the water pressure fluctuation prediction results and the dynamic compensation coefficient in pump control, different weights are assigned to them, and then a fusion calculation is performed to generate water pressure compensation parameters. These parameters can comprehensively consider factors such as water pressure change trends and real-time water flow.

[0046] Step S123: Input the water pressure compensation parameters and the rotational speed deviation value into the fan coordination module of the adaptive control model to construct a joint feature space of the rotational speed deviation value and the harmful gas concentration monitoring index; extract the topological association features of the fan nodes and the powder room gas concentration monitoring nodes in the joint feature space through a graph convolutional network; input the topological association features into a temporal attention mechanism for dynamic weighting to generate rotational speed adjustment demand weights; calibrate the ventilation compensation duration of the water pressure compensation parameters according to the rotational speed adjustment demand weights to generate the rotational speed correction parameters that include blade angle adjustment amount, rotational speed change rate, and ventilation compensation duration.

[0047] In the fan coordination module, water pressure compensation parameters and speed deviation values ​​are used as inputs. For example, if the speed deviation value is 60 rpm, a joint feature space is constructed between the speed deviation value and the hazardous gas concentration monitoring index. This joint feature space integrates relevant information such as speed deviation and hazardous gas concentration to form a comprehensive feature representation. For example, considering that speed deviation may affect the diffusion and concentration distribution of hazardous gases in the powder room, these factors are correlated to construct the joint feature space.

[0048] Next, the joint feature space is processed using a graph convolutional network. Graph convolutional networks can analyze the node relationships and features in graph-structured data. In this scenario, it extracts the topological association features between the fan nodes and the powder room gas concentration monitoring nodes in the joint feature space. For example, it finds a connection between changes in fan speed and changes in harmful gas concentration in different areas of the powder room; these connections can be understood as the extracted topological association features.

[0049] Topological association features are input into a temporal attention mechanism, which dynamically weights them based on the importance of data over time. For example, the impact of fan speed on harmful gas concentration may vary at different production times. The temporal attention mechanism can assign different weights to the topological association features based on these changes, generating weights for speed adjustment needs.

[0050] Finally, the ventilation compensation duration is calibrated based on the weighting of the speed adjustment demand for the water pressure compensation parameters. Combining the pump operating status reflected by the water pressure compensation parameters with the weighting of the speed adjustment demand, a suitable ventilation compensation duration is calculated, along with the blade angle adjustment and speed change rate. The blade angle adjustment, speed change rate, and ventilation compensation duration are then integrated (in the integration process, the blade angle adjustment, speed change rate, and ventilation compensation duration are first standardized to eliminate dimensional differences, then assigned corresponding integration weights to obtain the three weighted integration results, and finally integrated based on these results) to generate speed correction parameters. These parameters enable the fan to make reasonable speed and blade angle adjustments considering the concentration of harmful gases in the powder chamber and the pump operating status.

[0051] Step S130: Dynamically adjust the fuel valve of the burner based on the fuel supply control parameters, dynamically calibrate the motor power of the water pump based on the water pressure compensation parameters, dynamically adjust the blade angle of the blower based on the speed correction parameters, and collect the fuel flow feedback data of the burner, the output water pressure calibration data of the water pump, and the speed adjustment data of the blower in real time to update the equipment operating status set.

[0052] During the operation of the ceramic coating production line, the fuel valve of the burner is adjusted based on the generated fuel supply control parameters. For example, the dynamic adjustment coefficient in the fuel supply control parameters is 1.3, and the combustion temperature compensation gradient is 0.6. These control parameters are input into the burner's control unit, which generates a fuel valve opening command based on these parameters. Through the combined effect of the dynamic adjustment coefficient and the combustion temperature compensation gradient, the required opening degree of the fuel valve is calculated. For example, considering the characteristics of the equipment and production process requirements, the corresponding opening degree of the fuel valve can be calculated, driving the proportional solenoid valve of the fuel valve to perform the opening and closing action. For example, the standardized calculation method is: Adjusted opening degree = Base opening degree × Dynamic adjustment coefficient + Combustion temperature compensation gradient × (Current combustion temperature - Ideal combustion temperature). After the action is executed, real-time fuel flow feedback data and combustion temperature data are collected. For example, if the real-time fuel flow rate is collected as 65 liters / hour and the combustion temperature is 1380 degrees Celsius, the fuel supply efficiency is recalculated based on this data, generating an updated fuel supply efficiency.

[0053] For the water pump, the water pressure compensation parameters and the updated fuel supply efficiency are input into the pump's frequency converter control module. For example, if the dynamic compensation coefficient in the water pressure compensation parameters is 1.2, the frequency converter control module adjusts the pulse width modulation signal of the motor power according to this coefficient. By changing the parameters of the pulse width modulation signal, the pump motor is driven to change its output torque. After the operation is executed, the output water pressure calibration data and real-time water flow rate are collected. For example, if the output water pressure calibration data is 30 Pascals and the real-time water flow rate is 42 cubic meters per hour, combined with the water and gas consumption data, calibrated water pressure fluctuation data is generated.

[0054] For the fan, the calibrated water pressure fluctuation data and speed correction parameters are input into the fan's servo controller. For example, the blade angle adjustment in the speed correction parameters is 10 degrees, and the ventilation compensation duration is 15 minutes. The servo controller generates blade deflection pulse signals based on these parameters, driving the blade angle actuator to perform positioning calibration. After execution, real-time speed adjustment data and harmful gas concentration monitoring indicators are collected, and a dynamic ventilation compensation coefficient is generated. For example, the real-time speed adjustment data is 1650 rpm, and the nitrogen oxide concentration in the harmful gas concentration monitoring indicators is 55 ppm.

[0055] Finally, the updated fuel supply efficiency, calibrated water pressure fluctuation data, real-time speed adjustment data, and dynamic ventilation compensation coefficient are fused into a multi-dimensional dataset. Using existing data fusion algorithms, these different data types are integrated to generate an updated equipment operating status set that includes fuel flow feedback data, output water pressure calibration data, and speed adjustment data. This updated equipment operating status set is then synchronized to the digital panel of the spraying production line process, enabling real-time updates and monitoring of the equipment's operating status.

[0056] Step S131: Input the fuel supply control parameters into the burner control unit, generate a fuel valve opening command through the dynamic adjustment coefficient and combustion temperature compensation gradient in the fuel supply control parameters, drive the proportional solenoid valve of the fuel valve to perform opening and closing actions, and collect real-time fuel flow feedback data and combustion temperature data after execution to generate an updated fuel supply efficiency.

[0057] In a ceramic coating production line, once the fuel supply control parameters are generated, they are input into the burner's control unit. For example, the dynamic adjustment coefficient in the fuel supply control parameters might be 1.4, and the combustion temperature compensation gradient might be 0.7. The control unit then generates a fuel valve opening command based on these two parameters. First, the control unit calculates the fuel valve opening command according to pre-set rules and algorithms, combining the dynamic adjustment coefficient and the combustion temperature compensation gradient. For example, the pre-set rules and algorithms can consider and process factors such as the deviation between the current combustion temperature and the ideal combustion temperature, as well as the impact of fuel supply efficiency on combustion performance, to determine the specific opening value that the fuel valve needs to be adjusted to, thus generating the fuel valve opening command.

[0058] Next, the command is transmitted to the proportional solenoid valve of the fuel valve, driving it to open and close. The proportional solenoid valve precisely controls the valve opening according to the received command. After the valve action is completed, sensors installed on the burner collect real-time fuel flow feedback data and combustion temperature data. For example, the collected real-time fuel flow is 70 liters / hour, and the combustion temperature is 1420 degrees Celsius. Based on this collected data, the fuel supply efficiency is recalculated. An exemplary calculation method is to compare the real-time fuel flow with a preset standard flow value, and combine this with changes in combustion temperature to obtain an updated fuel supply efficiency. For example, by comparison, it is found that the current real-time fuel flow has increased by a certain percentage compared to the preset standard flow value. Considering the increase in combustion temperature, the updated fuel supply efficiency is calculated to be 12 liters / hour (it should be noted that this embodiment is only an example; actual calculations will be more complex and based on specific production processes and equipment characteristics. Those skilled in the art can select / adjust the corresponding rules and algorithms according to the actual situation). This updated fuel supply efficiency will serve as important data for subsequent analysis and control, reflecting the current fuel supply status of the burner.

[0059] Step S132: Input the water pressure compensation parameters and the updated fuel supply efficiency into the frequency converter control module of the water pump, adjust the pulse width modulation signal of the motor power through the dynamic compensation coefficient in the water pressure compensation parameters, drive the water pump motor to change the output torque, and collect the output water pressure calibration data and real-time water flow after execution, and generate calibrated water pressure fluctuation data by combining water and gas consumption data.

[0060] During the pump control process, the water pressure compensation parameters and the updated fuel supply efficiency are input into the pump's frequency converter control module. For example, the dynamic compensation coefficient in the water pressure compensation parameters is 1.3, and the updated fuel supply efficiency is 12 liters / hour. The frequency converter control module adjusts the pulse width modulation signal for motor power based on the dynamic compensation coefficient. The pulse width modulation signal is a key signal for controlling motor power, and the dynamic compensation coefficient affects parameters such as the duty cycle of this signal. For example, based on the dynamic compensation coefficient, the frequency converter control module will correspondingly increase or decrease the pulse width of the pulse width modulation signal, thereby changing the motor's input voltage and current, and thus adjusting the motor power.

[0061] The above method drives the water pump motor to change its output torque. The water pump motor adjusts its output torque according to the adjusted signal to meet the water pressure requirements during production. After the motor operates, a sensor installed at the water pump output end collects output water pressure calibration data and real-time water flow rate. For example, the collected output water pressure calibration data is 32 Pascals, and the real-time water flow rate is 45 cubic meters per hour.

[0062] Next, calibrated water pressure fluctuation data is generated by combining water and gas consumption data. This requires analyzing the relationship between historical water flow data, current real-time water flow, and output water pressure calibration data within the water and gas consumption data. For example, the trend of water flow changes over a past period can be analyzed, along with the impact of these changes on water pressure. By comparing the current real-time water flow with the historical average, and considering the differences between the output water pressure calibration data and previous water pressure data, the calibrated water pressure fluctuation data can be determined. This calibrated water pressure fluctuation data more accurately reflects the actual water pressure fluctuation of the pump after adjustment, providing a reliable basis for subsequent production process monitoring and further control.

[0063] Step S133: Input the calibrated water pressure fluctuation data and the speed correction parameters into the servo controller of the fan. Generate a blade deflection pulse signal through the blade angle adjustment amount and ventilation compensation duration in the speed correction parameters, drive the blade angle actuator to perform positioning calibration, and collect the real-time speed adjustment data and harmful gas concentration monitoring indicators after execution to generate a dynamic ventilation compensation coefficient.

[0064] Once the calibrated water pressure fluctuation data and speed correction parameters are determined, they are input into the fan's servo controller. For example, the blade angle adjustment in the speed correction parameters is 12 degrees, and the ventilation compensation duration is 20 minutes. The servo controller generates blade deflection pulse signals based on these parameters. This signal is a precise control signal that determines the pulse frequency, amplitude, and duration based on parameters such as the blade angle adjustment and ventilation compensation duration. For instance, based on a blade angle adjustment of 12 degrees, the number and frequency of pulses to be sent are calculated to ensure the blades are accurately adjusted to the specified angle; simultaneously, combined with the ventilation compensation duration, the duration of the pulse signals is adjusted to ensure the fan maintains appropriate ventilation for the corresponding time.

[0065] The generated blade deflection pulse signal is transmitted to the blade angle actuator, driving it to perform positioning calibration. Upon receiving the signal, the blade angle actuator adjusts the blade angle to the designated position via a mechanical transmission device. After the blade angle adjustment is complete, sensors installed on the wind turbine collect real-time speed adjustment data and simultaneously monitor harmful gas concentration indicators.

[0066] Based on the collected data, a dynamic ventilation compensation coefficient is generated. The generation process requires comprehensive consideration of factors such as real-time speed adjustment data, changes in harmful gas concentration, and the fan's ventilation capacity. For example, the difference between the real-time speed and the preset speed is analyzed, along with the effect of this difference on reducing harmful gas concentration. If a significant decrease in harmful gas concentration is observed after increasing the speed, the dynamic ventilation compensation coefficient is calculated based on this relationship and a preset algorithm. This coefficient reflects a quantitative indicator of the fan's ventilation compensation effect under current operating conditions, helping to assess the fan's ventilation performance and its impact on the powder room environment, providing a reference for subsequent production control. For instance, when calculating the dynamic ventilation compensation coefficient, a base coefficient is set, and the difference between the real-time and preset speeds is analyzed to calculate the speed change. The harmful gas concentrations before and after the speed increase are compared to obtain the concentration decrease. These two factors are then input into a preset algorithm with certain weights, such as using a linear combination calculation, to finally obtain the dynamic ventilation compensation coefficient, thus quantifying the ventilation compensation effect.

[0067] Step S134: The updated fuel supply efficiency, calibrated water pressure fluctuation data, real-time speed adjustment data and dynamic ventilation compensation coefficient are fused into multi-dimensional data to generate an updated set of equipment operating status that includes fuel flow feedback data, output water pressure calibration data and speed adjustment data, and the updated set of equipment operating status is synchronized to the digital panel of the spraying production line process.

[0068] After completing the control and data collection of the burner, water pump, and fan, multi-dimensional data fusion was initiated. First, the updated fuel supply efficiency, calibrated water pressure fluctuation data, real-time speed adjustment data, and dynamic ventilation compensation coefficient were integrated. These data reflect the operating status of the burner, water pump, and fan, as well as their impact on the overall production environment.

[0069] During the fusion process, existing data fusion algorithms will be employed. For example, considering the varying importance of different data points in reflecting equipment operating status and the production environment, different weights will be assigned to each data point. For fuel supply efficiency, which directly impacts burner operation and product quality, a higher weight may be assigned; for water pressure fluctuation data, appropriate weights will be assigned based on their impact on coating quality and production stability; and for real-time speed adjustment data and dynamic ventilation compensation coefficients, appropriate weights will be determined according to their roles in ventilation effectiveness and powder chamber environmental safety.

[0070] Then, the data are weighted according to these weights. For example, the updated fuel supply efficiency is multiplied by its corresponding weight, the calibrated water pressure fluctuation data is multiplied by its weight, the real-time speed adjustment data is multiplied by its weight, and the dynamic ventilation compensation coefficient is multiplied by its weight. These weighted results are then concatenated. In this way, an updated set of equipment operating status data, including fuel flow feedback data, output water pressure calibration data, and speed adjustment data, is generated.

[0071] Finally, through data transmission networks and related communication protocols, the updated equipment operating status is synchronized to the digital panel of the spraying production line. This allows operators to visually view the latest equipment operating status information on the digital panel, including fuel flow feedback from the burner, output water pressure calibration of the water pump, and fan speed adjustment results. This real-time updated data helps operators promptly identify problems in equipment operation and make corresponding adjustments and decisions, ensuring the stable and efficient operation of the ceramic spraying production line.

[0072] Step S140: Synchronize the updated equipment operating status set, the harmful gas concentration monitoring index, and the powder booth safety index to the MES system, trigger the MES system to optimize the production data of the electrical control cabinet, and update the real-time process parameters in the digital panel of the spraying production line process.

[0073] During the production process of the ceramic coating production line, after the equipment operating status set is updated, the updated equipment operating status set, hazardous gas concentration monitoring indicators, and powder booth safety indicators are synchronized to the MES system. For example, the updated equipment operating status set includes detailed information such as the burner's fuel flow feedback data of 75 liters / hour, the water pump's output water pressure calibration data of 35 Pascals, and the fan speed adjustment data of 1750 rpm; the hazardous gas concentration monitoring indicator shows a nitrogen oxide concentration of 45 ppm; and the powder booth safety indicators record the current environmental safety level of the powder booth and recent abnormal events.

[0074] The aforementioned data is transmitted to the MES system via existing data interfaces and communication protocols. Upon receiving this data, the MES system begins optimizing the production data configuration of the electrical control cabinet. First, the MES system analyzes and processes this data. For example, it analyzes data from the equipment operating status set to determine whether the burner, water pump, and fan are operating at their optimal state; it checks harmful gas concentration monitoring indicators to assess the environmental safety situation within the powder chamber; and it studies powder chamber safety indicators to determine if any potential safety risks exist.

[0075] Based on these analysis results, the MES system generates a set of process optimization strategies according to preset rules and algorithms. For example, if the fuel flow rate of the burner is found to be too high, a strategy to adjust the opening of the fuel valve may be generated; if the water pressure fluctuation is large and affects the coating quality, a corresponding water pump control strategy will be formulated; if the concentration of harmful gases in the powder chamber is close to the threshold, suggestions to strengthen ventilation or adjust the production process will be proposed.

[0076] The MES system then prioritizes the generated set of process optimization strategies based on their importance and urgency to the production process. For example, if the concentration of harmful gases in the powder room exceeds the standard, the strategy of strengthening ventilation will be given a higher priority; if the equipment operating status is only slightly deviating from the optimal state, the corresponding adjustment strategy will have a relatively lower priority.

[0077] Finally, the task priority ranking results trigger the electrical control cabinet to adjust the motion parameters and spraying pressure of the spraying robot arm. For example, if a higher priority strategy is to adjust the spraying thickness to improve product quality, the MES system will send relevant parameter adjustment instructions to the electrical control cabinet. After receiving the instructions, the electrical control cabinet will adjust the motion speed, trajectory, and other parameters of the spraying robot arm, as well as the spraying pressure, to optimize the production process.

[0078] Simultaneously, the MES system feeds back the optimized production data to the digital process panel of the spraying production line, updating the real-time process parameters. Operators can see the real-time changes in process parameters such as adjusted spraying thickness and coating uniformity on the digital panel, thereby gaining timely understanding of the optimization status of the production process and enabling further monitoring and management.

[0079] In an optional embodiment, the training process of the adaptive regulation model in this application includes:

[0080] Step S210: Extract the time series data of the burner's fuel supply timing characteristics, combustion temperature fluctuation characteristics, water pump water pressure fluctuation correlation characteristics, fan speed dynamic response characteristics, and coal chamber operation frequency from the historical production data set.

[0081] The historical production data set of the ceramic coating plant contains a large number of records regarding the operation of various equipment and data related to the powder booth. For the burner, fuel supply timing characteristics are extracted from this data. For example, by analyzing the fuel supply data at different times of day over the past month and examining its changes over time, it can be found that the fuel supply is relatively stable in the morning, while it fluctuates in the afternoon due to changes in production tasks. This pattern of fuel supply variation over time is the fuel supply timing characteristic.

[0082] Simultaneously, combustion temperature fluctuation characteristics are extracted. By analyzing historical data on combustion temperature records, the fluctuations in different production stages and environmental conditions are observed. For example, during equipment startup, the combustion temperature rises rapidly. After reaching a stable value, it may experience slight fluctuations due to factors such as fuel quality. These temperature change patterns and fluctuation ranges constitute the combustion temperature fluctuation characteristics.

[0083] For water pumps, we extract the correlation characteristics of water pressure fluctuations, study the changes in the pump's input and output water pressures in historical data, and their correlation with other factors (such as water flow rate, equipment operating time, etc.). For example, we find that when the water flow rate increases, the output water pressure will rise accordingly, but the increase is not linear and is also affected by factors such as the pump's own performance. This correlation and fluctuation between water pressure and other factors constitutes the correlation characteristics of water pressure fluctuations.

[0084] For fans, the dynamic response characteristics of fan speed are extracted to analyze the fan speed response under different operating conditions. For example, how does the fan speed adjust when the concentration of harmful gases in the powder room changes, and what is the speed and magnitude of the adjustment? Through statistical analysis of historical data, the dynamic response pattern of fan speed with different factors is derived, i.e., the dynamic response characteristics of fan speed.

[0085] Finally, time-series data on the number of actions performed by the paint booth were extracted from historical data. This included recording the number of actions performed by the paint booth in different time periods, such as the number of spraying operations and the number of robotic arm movements, and analyzing their distribution patterns over time. For example, it was found that the number of actions performed by the paint booth was relatively higher on Mondays and lower on weekends. This data on the changes in the number of paint booth actions over time constitutes the time-series data on the number of actions performed by the paint booth.

[0086] Step S211: The fuel supply timing characteristics and the combustion temperature fluctuation characteristics are fused in multiple dimensions to generate a comprehensive state characteristic of the burner.

[0087] After extracting the timing characteristics of fuel supply and the fluctuation characteristics of combustion temperature from the burner, multi-dimensional fusion is performed. From a temporal perspective, the changes in fuel supply at different times are correlated with the changes in combustion temperature at the same or similar times. For example, if the fuel supply suddenly increases within a specific time period, the trend of combustion temperature change at this time is examined to see if it rises accordingly or if there is a delayed change.

[0088] From the perspective of data characteristics, the relationship between the magnitude of changes in fuel supply and the magnitude of changes in combustion temperature is analyzed. For example, how many degrees does the combustion temperature rise when the fuel supply increases by 10%? By statistically analyzing a large amount of historical data, the quantitative relationship between them is determined.

[0089] From the perspective of influencing factors, consider the combined impact of external environmental factors (such as workshop temperature and humidity) on fuel supply and combustion temperature. For example, how do the patterns of change in fuel supply and combustion temperature differ between high-temperature and normal-temperature environments?

[0090] By integrating these different dimensions of analysis, the characteristics of fuel supply timing and combustion temperature fluctuations are fused together. For example, a weighted fusion algorithm can be used to assign weights to factors of different dimensions based on their importance. Then, the data from each dimension are integrated and calculated to generate a comprehensive burner status characteristic. This characteristic integrates information from multiple aspects, including fuel supply and combustion temperature, and more comprehensively reflects the operating status of the burner.

[0091] Step S212: Perform correlation analysis between the water pressure fluctuation correlation feature and the water flow abnormal event in the water and gas consumption data to generate water pump dynamic compensation feature.

[0092] When processing water pump-related data, the correlation characteristics of water pressure fluctuations are analyzed in conjunction with abnormal water flow events in water and gas consumption data. First, the definition and judgment criteria for abnormal water flow events are clarified. For example, when the water flow exceeds or falls below a certain percentage (e.g., 15% of the historical average), it is judged as an abnormal water flow event.

[0093] Then, for each abnormal water flow event, analyze the specific manifestations of the correlation characteristics of water pressure fluctuations at the time of its occurrence. For example, when the water flow suddenly increases, check the changes in the input and output water pressure of the water pump, and whether the relationship between water pressure and other related factors (such as water pump motor power, pipeline resistance, etc.) changes.

[0094] By analyzing the correlation characteristics between a large number of abnormal water flow events and water pressure fluctuations, the inherent connections and patterns between them were identified. For example, it was found that when the water flow abnormally increases, the output water pressure will drop in a short period of time and then gradually recover, but the speed and extent of recovery are related to factors such as the operating time of the water pump.

[0095] Based on the above analysis results, a dynamic compensation feature for the water pump is generated. This feature includes the variation law of water pump water pressure and the corresponding compensation strategy information under different abnormal water flow conditions, providing a basis for subsequent dynamic compensation and control of the water pump.

[0096] Step S213: Match the dynamic response characteristics of the rotation speed with the ventilation abnormal events in the number of powder chamber actions to generate fan coordinated control characteristics.

[0097] When processing fan data, the dynamic response characteristics of the fan speed are time-series matched with the ventilation anomaly events in the number of powder chamber actions. First, the criteria for judging ventilation anomalies are determined, such as abnormal start-up and shutdown of the ventilation system and insufficient ventilation volume.

[0098] Then, for each ventilation anomaly event, find the corresponding time point and speed change in the dynamic response characteristics of the fan speed. For example, when the ventilation system suddenly stops, observe how the fan speed drops rapidly, at what rate, and how the fan speed recovers after the ventilation system restarts.

[0099] By time-series matching and analysis of numerous ventilation anomalies and dynamic speed response characteristics, the synergistic relationships and patterns among them were identified. For example, it was found that when ventilation volume is insufficient, the fan speed will automatically increase to increase ventilation volume, but the magnitude and speed of the increase will be affected by factors such as the concentration of harmful gases in the powder room.

[0100] Based on these analysis results, a fan coordinated control feature was generated. This feature includes the response mode of fan speed under different ventilation abnormalities and its synergistic relationship with other factors (such as powder room environmental parameters), providing data support for achieving coordinated control of fan and powder room ventilation needs.

[0101] Step S214: Construct a training dataset based on the comprehensive state characteristics of the burner, the dynamic compensation characteristics of the water pump, and the coordinated control characteristics of the fan. Generate training labels through the changes in the concentration of harmful gases in the powder room safety indicators. Iterate and optimize the initial neural network model until the error between the control parameters output by the initial neural network model and the training labels is lower than a preset threshold, thereby obtaining the adaptive control model.

[0102] After processing the relevant characteristics of the burner, water pump, and fan, a training dataset is constructed based on the comprehensive state characteristics of the burner, the dynamic compensation characteristics of the water pump, and the coordinated control characteristics of the fan. The feature data from these different devices are then organized according to a preset format and structure to form a dataset suitable for training the neural network model. For example, each feature data point is standardized to have a uniform dimension and value range, and then arranged in a specific batch and order.

[0103] Simultaneously, training labels are generated based on the changes in harmful gas concentrations in the powder chamber safety indicators. The changes in harmful gas concentrations in the powder chamber safety indicators are analyzed under different production stages and equipment operating conditions, and these changes are used as the desired output, i.e., the training labels. For example, when the burner, water pump, and fan are in a specific combination of operating conditions, the harmful gas concentration in the powder chamber should vary within a certain range; this range is the corresponding training label.

[0104] Next, the initial neural network model is iteratively optimized using the constructed training dataset and training labels. In each iteration, the training data is input into the initial neural network model, and the model calculates output control parameters based on its own parameters and the algorithm. Then, the output control parameters are compared with the training labels, and the error between the two is calculated. For example, error calculation methods such as mean squared error are used to measure the difference between the model output and the expected result.

[0105] Based on the calculated error, the parameters of the initial neural network model are adjusted using optimization algorithms such as backpropagation. The purpose of the adjustment is to gradually reduce the error between the model's output control parameters and the training labels.

[0106] In each iteration, the optimization algorithm updates the neural network's weights and biases based on the error gradient information, aiming to reduce the error. For example, for the weights of a neuron, the update amount is determined based on the partial derivative of the error with respect to that weight, thus adjusting the weights in a direction that reduces the error.

[0107] Through multiple iterations, the process of inputting data, calculating errors, and adjusting parameters is repeated continuously. As the number of iterations increases, the initial neural network model's ability to fit the training data gradually improves, and the error between the output control parameters and the training labels gradually decreases.

[0108] During the iteration process, the error between the model's output adjustment parameters and the training labels is continuously monitored. A preset threshold is established, determined based on actual production needs and the required model accuracy. For example, it is set that when the error decreases to below a certain small value (such as 0.05), the model is considered to have reached an acceptable level of accuracy.

[0109] When the error between the control parameters output by the model and the training labels is lower than a preset threshold, the neural network model is considered to have been optimized and trained to obtain an adaptive control model. This adaptive control model has learned the intrinsic relationship between the operating status characteristics of the burner, water pump and fan and the safety indicators of the powder room. It can accurately generate the corresponding fuel supply control parameters of the burner, water pressure compensation parameters of the water pump and speed correction parameters of the fan according to the input set of equipment operating status, thus providing strong support for the intelligent control of the ceramic spraying production line.

[0110] In an optional embodiment, the method further includes:

[0111] Step S310: Construct a three-dimensional simulation model of the spraying production line based on the digital twin strategy. The input information of the three-dimensional simulation model includes the set of equipment operating status, powder booth safety indicators, and production data of the electrical control cabinet.

[0112] In ceramic coating production environments, to more intuitively and accurately simulate and analyze the production process, a three-dimensional simulation model can be constructed based on a digital twin strategy. First, the necessary input information for the model is collected. The equipment operating status set includes detailed operating data for equipment such as the burner, water pump, and fan. For example, the burner's fuel supply efficiency is 15 liters / hour, and the combustion temperature is 1450 degrees Celsius; the water pump's input water pressure is 25 Pascals, and its output water pressure is 38 Pascals, with water pressure fluctuation data showing a certain periodic change; the fan's speed is 1800 rpm, with a speed deviation of 30 rpm.

[0113] Regarding the safety indicators of the powder booth, the monitoring indicators for harmful gas concentration showed that the nitrogen oxide concentration was 40 ppm. The environmental safety level of the powder booth was assessed as medium safety level. Among the recent powder booth operations, there were 2 spraying interruptions and 1 instance of the robotic arm exceeding its limit.

[0114] The production data of the electrical control cabinet includes information such as coating thickness, coating uniformity, and drying time. For example, if the current coating thickness is set to 0.5 mm, actual measurements and data analysis show that the coating uniformity fluctuates within a certain range, and the average drying time is 30 minutes.

[0115] Using the above input information, a 3D simulation model is constructed using corresponding 3D modeling software and a digital twin technology platform. During the modeling process, the physical structure, spatial location, and connection relationships of each device are accurately digitally modeled. For example, 3D models of the burner, water pump, and fan are created in a virtual environment according to their actual dimensions and layout, and the pipe connections and electrical wiring routes between them are accurately simulated.

[0116] Simultaneously, dynamic information such as equipment operating status, powder chamber safety indicators, and production data are linked to the 3D model. By developing existing algorithms and programs, the 3D model can reflect changes in this information in real time. For example, when the fuel supply efficiency of the burner changes, the burner model in the 3D model will display different operating states accordingly; the size and color of the flame will be dynamically adjusted according to the combustion temperature and fuel supply. Changes in the concentration of harmful gases in the powder chamber will be presented in the 3D model through visualization methods such as color gradients.

[0117] Using the above methods, a three-dimensional simulation model that can reflect the actual production process in real time can be constructed. This model can not only intuitively display the operating status of the spraying production line, but also provide a powerful tool for subsequent simulation analysis and optimization.

[0118] Step S311: Load the updated set of equipment operating states into the three-dimensional simulation model to simulate the coordinated operation of the burner, water pump and fan, and calculate the diffusion path and concentration distribution of harmful gases in the powder room in real time.

[0119] After constructing the 3D simulation model, the updated set of equipment operating statuses is loaded into the model. For example, the updated set of equipment operating statuses shows that the burner's fuel supply efficiency has increased to 18 liters / hour, and the combustion temperature has risen to 1500 degrees Celsius; the water pump's output water pressure has stabilized at 40 Pascals, and the water flow rate has increased; the fan speed has been adjusted to 1850 rpm. Upon receiving this updated data, the 3D simulation model begins to simulate the coordinated operation of the burner, water pump, and fan. In the model, the burner model simulates the fuel combustion process based on the new fuel supply efficiency and combustion temperature data, including the flame morphology and heat dissipation. As the fuel supply increases, the flame becomes more intense, releasing more heat.

[0120] The water pump model simulates water flow in the pipes based on changes in output water pressure and flow rate. An increase in flow rate leads to a faster flow velocity, and the pressure distribution within the pipes changes accordingly. The fan model simulates airflow and ventilation effects based on new rotational speed data. An increase in rotational speed increases the fan's air volume, accelerating airflow within the powder chamber.

[0121] While the simulation equipment operates in tandem, the model calculates the diffusion paths and concentration distribution of harmful gases within the powder chamber in real time. Utilizing algorithms related to fluid mechanics and diffusion models, the calculations are performed in conjunction with factors such as the harmful gases produced by the burner, airflow within the powder chamber, and the equipment's operating status. For example, harmful gases such as nitrogen oxides produced by the burner rise with the hot airflow and diffuse within the powder chamber. The ventilation effect of the fan influences the diffusion direction and speed of these harmful gases, pushing them from high-concentration areas to low-concentration areas.

[0122] By comprehensively calculating these factors, the 3D simulation model can generate real-time diffusion path maps and concentration distribution maps of harmful gases within the powder chamber. The diffusion path map clearly shows how harmful gases flow and spread within the powder chamber after being generated near the burner. The concentration distribution map uses different colors or contour lines to represent the concentration of harmful gases in different areas of the powder chamber. Operators can intuitively understand which areas have higher concentrations of harmful gases and which have relatively lower concentrations, thus providing important information for ensuring powder chamber safety and optimizing the production process.

[0123] Step S312: If the concentration of nitrogen oxides or particulate matter in the simulation results exceeds the preset threshold, the fuel supply control parameters, water pressure compensation parameters, and speed correction parameters are regenerated, and the MES system is triggered to suspend the current production task.

[0124] After calculating the diffusion paths and concentration distribution of harmful gases in the powder booth in real time using a 3D simulation model, the nitrogen oxide and particulate matter concentrations in the simulation results are compared with preset thresholds. These preset thresholds are determined based on various factors, including relevant safety standards, production process requirements, and the actual conditions of the powder booth. For example, the preset threshold for nitrogen oxide concentration is set at 50 ppm, and the preset threshold for particulate matter concentration is set at 10 mg / m³. 3 .

[0125] If the simulation results show that the nitrogen oxide concentration exceeds 50 ppm, or the particulate matter concentration exceeds 10 mg / m³, then... 3 This indicates that the concentration of harmful gases in the powder preparation room is at an unsafe level, which may pose a threat to the production environment and the health of personnel. At this time, the system will automatically activate the corresponding countermeasures.

[0126] First, the fuel supply control parameters, water pressure compensation parameters, and speed correction parameters are regenerated. Based on the simulation results showing excessive concentrations of harmful gases and the data from the equipment operating status set, an adaptive control model or other relevant optimization algorithms are invoked to recalculate these parameters. For example, if the excessive nitrogen oxide concentration is found to be due to incomplete combustion in the burner, the fuel supply control parameters may be adjusted, the fuel valve opening increased, or the combustion temperature compensation gradient optimized to improve combustion efficiency and reduce nitrogen oxide production.

[0127] For water pumps and fans, the water pressure compensation parameters and speed correction parameters will be adjusted according to the specific situation. If it is found that insufficient ventilation in the powder room leads to the accumulation of harmful gases, the ventilation compensation duration and blade angle adjustment amount in the fan speed correction parameters may be increased to enhance the ventilation effect and reduce the concentration of harmful gases.

[0128] Simultaneously, the MES system is triggered to suspend the current production task. An instruction is sent to the MES system informing it that the concentration of harmful gases in the powder booth exceeds the standard and that production needs to be suspended to avoid further risks. Upon receiving the instruction, the MES system will immediately stop sending production instructions to the electrical control cabinet, causing all equipment on the spraying production line to enter a suspended state, preventing continued production in an unsafe environment and ensuring the safety and stability of the production process.

[0129] In an optional embodiment, the method further includes:

[0130] Step S410: Input the updated equipment operating status set into the three-dimensional simulation model to simulate the impact of changes in the fuel supply efficiency of the burner on the temperature of the coal chamber, and generate a temperature-concentration correlation curve by associating the harmful gas concentration monitoring index.

[0131] In the ceramic coating production process, to verify the accuracy and reliability of the 3D simulation model, an updated set of equipment operating states can be input into the 3D simulation model. For example, the updated set of equipment operating states shows that the fuel supply efficiency of the burner has changed over a period of time, gradually increasing from an initial 15 liters / hour to 20 liters / hour.

[0132] After receiving this data, the 3D simulation model begins to simulate the impact of changes in the burner's fuel supply efficiency on the coal chamber temperature. The model performs simulation calculations based on the physical principles of the burner and the laws of heat transfer, combined with changes in fuel supply efficiency. As fuel supply efficiency increases, the heat released by the burner increases, and this heat is transferred to the air inside the coal chamber and surrounding objects through heat conduction, convection, and radiation.

[0133] During the simulation, the model monitored temperature changes in different locations within the coal chamber in real time. For example, the temperature rose more significantly in areas near the burner, while the temperature rise was relatively smaller in well-ventilated areas further away from the burner. Simultaneously, the model correlated with hazardous gas concentration monitoring indicators. Since changes in the burner's fuel supply efficiency affect the combustion process, and consequently the generation and diffusion of hazardous gases, changes in hazardous gas concentration were monitored concurrently with temperature changes.

[0134] Through continuous simulation and data acquisition, temperature data and corresponding harmful gas concentration data at different times within the coal preparation chamber were collected and analyzed. A temperature-concentration correlation curve was generated, plotting temperature on the x-axis and harmful gas concentration on the y-axis. This curve visually illustrates the relationship between temperature and harmful gas concentration within the coal preparation chamber as the burner's fuel supply efficiency changes. For example, it may be observed that as temperature increases, the harmful gas concentration also tends to rise, but the rate of increase is not constant and depends on various factors such as burner combustion efficiency and ventilation conditions. Generating this temperature-concentration correlation curve allows for a deeper understanding of the comprehensive impact of burner operation on the coal preparation chamber environment, providing crucial data support for optimizing production processes and ensuring coal preparation chamber safety.

[0135] Step S411: Simulate the effect of water pressure compensation parameters of the water pump on coating uniformity, and generate water pressure-coating quality evaluation results by combining the drying time in the production data;

[0136] In the 3D simulation model, the effect of the water pump's water pressure compensation parameters on the coating uniformity is further simulated. For example, the water pump's water pressure compensation parameters have been adjusted, increasing the output water pressure from 35 Pascals to 40 Pascals.

[0137] Based on the working principle of water pumps and knowledge of spraying processes, the model simulates the coating process on a workpiece surface under different water pressure conditions. Changes in water pressure affect the speed and pressure distribution of the coating ejected from the spray gun, thus affecting the deposition of the coating on the workpiece surface. When the water pressure increases, the coating ejection speed increases, and the coverage area and deposition thickness on the workpiece surface may change.

[0138] During the simulation, the model assesses changes in coating uniformity by performing detailed calculations and analyses of coating deposition on the workpiece surface. For example, coating uniformity is quantified by determining the differences in coating thickness at different locations on the workpiece surface. Simultaneously, analysis is conducted using drying time information from production data, revealing that drying time is influenced by various factors, among which water pressure has a certain correlation with coating distribution and drying speed on the workpiece surface.

[0139] For example, production data records the variation in coating drying time under different water pressure conditions. When simulating the impact of water pressure changes on coating uniformity, the drying time factor is taken into account. For instance, it was found that as water pressure increases, coating uniformity improves, but drying time may increase. By comprehensively analyzing the impact of water pressure changes on coating uniformity and drying time, a water pressure-coating quality assessment result is generated. This assessment result can be presented in report form, including quantitative indicators of coating uniformity under different water pressure compensation parameters, changes in drying time, and a comprehensive evaluation of the overall coating quality. Such assessment results can help production personnel understand the specific impact of water pump pressure compensation parameters on coating quality, thereby optimizing water pump control strategies and improving product quality.

[0140] Step S412: Simulate the ventilation efficiency of the dust chamber based on the fan speed correction parameter, and generate ventilation-dust diffusion optimization index by combining particulate matter concentration;

[0141] A 3D simulation model was used to simulate the ventilation efficiency of the powder chamber based on fan speed correction parameters. For example, the fan speed correction parameters adjusted the fan speed from 1800 rpm to 1900 rpm. Based on fluid dynamics principles and the design parameters of the ventilation system, the model simulated the airflow within the powder chamber under the new speed conditions. As the fan speed increased, the airflow volume increased, the airflow velocity within the powder chamber accelerated, and the ventilation path changed accordingly. By simulating the airflow within the powder chamber, the ventilation volume and velocity at different locations could be calculated, thereby evaluating the ventilation efficiency.

[0142] Simultaneously, analysis was conducted using particulate matter concentration data. During the powder booth production process, a certain amount of particulate matter is generated, and the diffusion and distribution of these particles are affected by ventilation conditions. When the fan speed changes, the diffusion path and concentration distribution of particulate matter within the powder booth also change.

[0143] The model calculates the dispersion of particulate matter under different ventilation conditions and, combined with preset evaluation criteria, generates ventilation-dust dispersion optimization indicators. For example, by monitoring changes in particulate matter concentration in different areas of the dust chamber, it calculates indicators such as average particulate matter concentration and standard deviation of particulate matter concentration to measure the optimization effect of ventilation on dust dispersion. If, under a new fan speed, the average particulate matter concentration in the dust chamber decreases and the concentration distribution becomes more uniform, it indicates that the ventilation-dust dispersion optimization indicators have improved. These indicators provide a quantitative basis for evaluating the ventilation effect of fans, helping production personnel determine the optimal fan speed correction parameters to effectively control dust concentration in the dust chamber, ensuring a safe production environment and product quality.

[0144] Step S413: If the temperature-concentration correlation curve, the water pressure-coating quality assessment result, or the ventilation-dust diffusion optimization index exceeds the preset range, the current control parameter is determined to be abnormal, and the adaptive control model is triggered to re-execute the dynamic parameter analysis processing.

[0145] After obtaining the temperature-concentration correlation curve, water pressure-coating quality assessment results, and ventilation-dust diffusion optimization indicators, these results were compared with preset ranges. The preset ranges were determined based on various factors, including production process standards, product quality objectives, and safety regulations.

[0146] For the temperature-concentration correlation curve, the preset range specifies the reasonable relationship between the temperature and the concentration of harmful gases in the coal preparation chamber under different burner operating conditions. For example, within the existing fuel supply efficiency range, the increase in the concentration of harmful gases with increasing temperature should be within a certain proportion. If the actual generated temperature-concentration correlation curve shows that the increase in the concentration of harmful gases with increasing temperature exceeds the preset range, this indicates that there may be a problem with the burner's operating condition, leading to abnormal emissions of harmful gases.

[0147] For the water pressure-coating quality assessment results, the preset range clearly defines the reasonable range of values ​​for coating uniformity and drying time under different water pressure conditions. For example, it specifies that within a certain range of water pressure fluctuations, coating uniformity should remain above the preset quantitative index, and drying time should be within the preset time interval. If the water pressure-coating quality assessment results show that coating uniformity is lower than the preset standard, or drying time exceeds the reasonable range, it indicates that the water pressure compensation parameters of the water pump may need to be adjusted.

[0148] For the ventilation-dust diffusion optimization index, preset ranges are set to define the control targets and diffusion uniformity requirements for particulate matter concentration in the dust chamber under different fan speeds. For example, it is specified that within a certain fan speed range, the average particulate matter concentration in the dust chamber should be below a certain threshold, and the standard deviation of the particulate matter concentration should be within a certain range. If the ventilation-dust diffusion optimization index exceeds these preset ranges, it means that the ventilation effect of the fan has not met expectations, and the speed correction parameters may be inappropriate.

[0149] If any of the above assessment results exceed the preset range, the system will determine that the current control parameters are abnormal. At this time, the system will automatically trigger the adaptive control model to re-execute the dynamic parameter analysis process. The system will re-input the current equipment operating status set, powder chamber safety indicators, and other relevant data into the adaptive control model. The model will re-analyze and calculate based on this data to generate new fuel supply control parameters for the burner, water pressure compensation parameters for the water pump, and speed correction parameters for the blower, in order to adjust the operating status of the equipment, bring the production process back to normal operation, and ensure product quality and the safety of the production environment.

[0150] In a non-limiting embodiment, the method further includes:

[0151] Step S510: Collect fuel flow feedback data, output water pressure calibration data, and speed adjustment data from the updated equipment operating status set to construct a dynamic control effect index; compare the dynamic control effect index with the preset process standard parameters layer by layer to generate residual distribution characteristics including burner control residual, water pump compensation residual, and fan speed residual;

[0152] After the real-time process parameters of the ceramic coating production line were updated, the effectiveness of the control measures was evaluated. First, key data was collected from the updated equipment operating status dataset, including fuel flow feedback data from the burner, output water pressure calibration data from the water pump, and fan speed adjustment data. For example, the collected fuel flow feedback data for the burner was 22 liters / hour, the output water pressure calibration data for the water pump was 42 Pascals, and the fan speed adjustment data was 1950 rpm.

[0153] The collected data is used to construct a dynamic control effect index. This index is a comprehensive indicator used to measure the degree to which the operating effect of the equipment after control measures matches the expected target. For example, for a burner, the deviation rate can be calculated based on the fuel flow feedback data and the preset ideal fuel flow value, serving as a measure of the burner's control effect; for a water pump, the water pressure deviation is calculated by comparing the output water pressure calibration data with the preset standard water pressure range; and for a fan, the speed deviation is calculated based on the speed adjustment data and the preset optimal speed value. These control effect indicators from different devices are integrated to form a comprehensive dynamic control effect index.

[0154] Next, the dynamic control effect indicators are compared layer by layer with the preset process standard parameters. The preset process standard parameters are determined according to the requirements of the production process and product quality standards, including the ideal fuel flow range, the standard water pressure value, and the optimal fan speed. For example, the preset ideal fuel flow range is 20-23 liters / hour, the standard water pressure value is 40-45 Pascals, and the optimal fan speed is 1900-2000 rpm.

[0155] During the comparison process, the residuals of burner control, water pump compensation, and fan speed were calculated separately. For the burner, the difference between the fuel flow feedback data and the ideal fuel flow range was calculated to obtain the burner control residual; for the water pump, the difference between the output water pressure calibration data and the standard water pressure value was calculated to obtain the water pump compensation residual; for the fan, the difference between the speed adjustment data and the optimal fan speed range was calculated to obtain the fan speed residual. Through this layer-by-layer comparison, the degree of difference between the control effect of each device and the preset standard was analyzed in detail.

[0156] These residuals are organized and analyzed to generate residual distribution characteristics that include burner control residuals, water pump compensation residuals, and fan speed residuals. This residual distribution characteristic can be presented in the form of charts or datasets to intuitively show the deviations in the control effects of each device. For example, when presented in chart form, the horizontal axis represents the device type (burner, water pump, fan), and the vertical axis represents the magnitude of the residuals. Different bar charts or line graphs are used to represent the residual situation of each device. This clearly shows which devices have good control effects and which devices still have significant room for improvement, providing a basis for subsequent optimization of the adaptive control model.

[0157] Step S511: Perform gradient compensation on the neural network weight matrix of the adaptive control model according to the residual distribution characteristics, and update the slope of the activation function of the fully connected layer of the adaptive control model; reload the updated adaptive control model into the electrical control cabinet, iteratively analyze the water pressure fluctuation data and speed deviation value collected in the next cycle, and generate incrementally optimized fuel supply control parameters and speed correction parameters.

[0158] After obtaining the residual distribution characteristics, gradient compensation is performed on the neural network weight matrix of the adaptive control model based on these characteristics. The weight matrix of the neural network determines the model's ability to process and map input data. By analyzing the residual distribution characteristics, it is possible to identify in which aspects the model's predictions deviate from the actual situation. For example, if the burner control residual is large, it indicates that the model's processing of burner-related parameters may not be accurate enough, and adjustments need to be made to the elements of the burner-related weight matrix.

[0159] Optimization algorithms such as gradient descent are used to calculate the gradient of the weight matrix based on the magnitude and direction of the residuals. The gradient represents the degree to which changes in the weights affect the model's output error. The weight matrix is ​​adjusted based on this gradient information to make the weights change in a direction that reduces the error; this is the process of gradient compensation. For example, if the gradient corresponding to a certain weight element is positive, it means that increasing that weight can reduce the error, so the value of that weight element is appropriately increased.

[0160] Simultaneously, the slope of the activation function in the fully connected layer of the adaptive control model is updated. The activation function introduces nonlinearity into the neural network; adjusting the slope can alter the model's nonlinear mapping ability, thus better fitting the data. Based on the residual distribution characteristics, the nonlinear fitting effect of the model on different equipment control scenarios is analyzed, and the slope of the activation function in the fully connected layer is adjusted accordingly. For example, if the model shows good linear fitting when processing water pump pressure data but performs poorly on some complex nonlinear changes, the slope of the activation function can be appropriately adjusted to enhance the model's ability to capture nonlinear relationships.

[0161] After updating the weight matrix and activation function slope of the adaptive control model, the updated adaptive control model is reloaded into the control cabinet. Upon receiving the updated model, the control cabinet prepares to iteratively analyze the water pressure fluctuation data and speed deviation values ​​collected in the next cycle. When new water pressure fluctuation data and speed deviation values ​​are collected in the next cycle, these data are input into the updated adaptive control model.

[0162] The model processes the data based on the updated parameters and structure, generating incrementally optimized fuel supply control parameters and speed correction parameters through a series of calculations and analyses. For example, when analyzing water pressure fluctuation data, the model considers the residuals from previous control processes, more accurately predicting water pressure change trends, thus generating more reasonable water pressure compensation parameters, which in turn affect the generation of fuel supply control parameters. For speed deviation values, the model combines new data and updated parameters to generate speed correction parameters that better meet actual needs, enabling the burner, water pump, and fan to operate more coordinatedly, further improving the control effect of the production process.

[0163] In a non-limiting embodiment, the method further includes:

[0164] Step S512: Extract the optimized production data from the MES system and synchronize it to the three-dimensional simulation model of the digital twin platform; input the updated current real-time process parameters into the three-dimensional simulation model to simulate the dynamic influence of the change in burner valve opening on the coal chamber temperature gradient, and output the simulated fuel consumption rate.

[0165] After updating the real-time process parameters of the ceramic coating production line, optimized production data is extracted from the MES system. This data covers multiple aspects, such as coating thickness, coating uniformity, drying time, and equipment operating status. For example, the optimized coating thickness is adjusted to 0.6 mm, coating uniformity is significantly improved, drying time is shortened to 25 minutes, and the operating parameters of the burner, water pump, and fan are also optimized accordingly.

[0166] This production data is synchronized to the 3D simulation model of the digital twin platform. Upon receiving this data, the 3D simulation model can reflect the latest status of the production process in real time. For example, the virtual spraying equipment in the model will simulate spraying based on new spraying thickness parameters, demonstrating the formation process of coatings of different thicknesses; improvements in coating uniformity will be presented through color consistency on the virtual workpiece surface or other visualization methods; and changes in drying time can be reflected by simulating the dynamic process of coating drying on the workpiece surface.

[0167] Next, the updated real-time process parameters are input into the 3D simulation model. These real-time process parameters include burner valve opening, water pump pressure, and fan speed. Taking burner valve opening as an example, the current burner valve opening is adjusted to 70%, and the model begins to simulate the dynamic impact of changes in burner valve opening on the coal chamber temperature gradient.

[0168] During the simulation, the model calculates based on the physical principles and heat transfer laws of the burner, combined with changes in valve opening. When the burner valve opening increases, the fuel supply increases, combustion becomes more intense, and more heat is released. This heat is transferred within the coal chamber through heat conduction, convection, and radiation, causing temperature changes at different locations within the chamber, thus creating a temperature gradient. The model monitors the temperature changes at various locations within the coal chamber in real time and generates temperature gradient data by calculating the temperature differences between different locations.

[0169] Simultaneously, the model outputs a simulated fuel consumption rate. Based on the burner valve opening, fuel supply efficiency, and relevant combustion process parameters, the model calculates the burner's fuel consumption under current operating conditions, deriving the simulated fuel consumption rate. For example, the simulation calculation shows that under the current valve opening and other operating conditions, the burner's simulated fuel consumption rate is 25 liters per hour. This simulated fuel consumption rate can provide a reference for production personnel, helping them understand the burner's energy consumption, assess the economics of the production process, and provide a basis for further optimizing the burner's operation.

[0170] Step S513: If the deviation between the simulated fuel consumption rate and the fuel flow feedback data in the updated equipment operating status set exceeds the tolerance range, the fuel supply control parameters are recalculated and the temperature compensation factor is superimposed; the compensated fuel supply control parameters are transmitted in reverse to the process digital panel of the spraying production line to cover the current real-time process parameters, and the fan coordination module is triggered to increase the ventilation compensation duration to balance the temperature fluctuation.

[0171] After obtaining the simulated fuel consumption rate, it is compared with the fuel flow feedback data in the updated equipment operating status set. The updated fuel flow feedback data in the equipment operating status set is 23 liters / hour, while the simulated fuel consumption rate is 25 liters / hour. Simultaneously, a tolerance range is preset, for example, set to ±1 liter / hour.

[0172] If the deviation between the simulated fuel consumption rate and the fuel flow feedback data exceeds the tolerance range, it indicates a significant difference between the actual fuel consumption and the simulation prediction. This may be due to an abnormality in the burner's operating status or errors in the model simulation. In this case, appropriate measures need to be taken to adjust the production process.

[0173] First, recalculate the fuel supply control parameters. Taking into account factors such as current equipment operating status, coal preparation room environmental parameters, and production process requirements, use an adaptive control model or other relevant algorithms to reanalyze and calculate the fuel supply control parameters. For example, considering factors such as burner combustion efficiency, oxygen content in the coal preparation room, and temperature changes, readjust parameters such as fuel valve opening and fuel supply time to ensure a more rational fuel supply that meets actual production needs.

[0174] During the recalculation of fuel supply control parameters, a temperature compensation factor is added. This temperature compensation factor is a parameter set to account for the impact of temperature changes on fuel supply and the combustion process. For example, when the temperature inside the coal preparation chamber is high, the fuel combustion rate may accelerate, requiring appropriate adjustments to the fuel supply to ensure combustion stability and efficiency. Based on the current temperature inside the coal preparation chamber and the relationship between temperature and fuel supply in historical production data, a suitable temperature compensation factor is determined and added to the recalculated fuel supply control parameters.

[0175] The compensated fuel supply control parameters are transmitted in reverse to the digital panel of the spraying production line. Through data transmission network and related communication protocols, the adjusted parameters are sent back to the control system of the spraying production line, overriding the current real-time process parameters. In this way, the burner control unit will adjust the fuel valve opening and other operations according to the new fuel supply control parameters, thereby achieving precise control of the burner's fuel supply.

[0176] Simultaneously, the fan coordination module is triggered to increase the ventilation compensation duration. Since adjustments to fuel supply may cause temperature fluctuations within the powder chamber, the fan coordination module, upon receiving a trigger signal, increases the ventilation compensation duration according to preset rules and algorithms to balance these temperature changes. For example, if the original ventilation compensation duration was 15 minutes, it is now adjusted to 20 minutes. By increasing the ventilation volume and duration, the airflow within the powder chamber is accelerated, allowing the heat generated by combustion to dissipate promptly, maintaining a relatively stable temperature within the powder chamber, ensuring the production process takes place in a suitable environment, and improving product quality and production efficiency.

[0177] In another non-independent embodiment, after updating the real-time process parameters in the digital panel of the spraying production line process, the method further includes: extracting the periodic fluctuation characteristics of the burner's fuel flow and the abnormal fluctuation events of the water pump's output water pressure based on the updated equipment operating status set; performing pattern matching between the periodic fluctuation characteristics of the fuel flow and the valve blockage characteristics in the historical fault database to generate a fuel valve blockage probability prediction result; aligning the abnormal fluctuation events of the output water pressure with the current fluctuation data of the water pump motor to generate a motor overload risk level; and generating a preventive maintenance instruction based on the fuel valve blockage probability prediction result and the motor overload risk level to trigger the electrical control cabinet's maintenance system to execute valve cleaning priority tasks and adjust the motor cooling time, and updating the maintenance records in the digital panel of the spraying production line process.

[0178] This application embodiment achieves global optimization control of the ceramic spraying production line process by integrating a real-time acquisition and dynamic control mechanism for multi-dimensional equipment parameters. Through the adaptive control model's collaborative analysis of burner fuel supply efficiency, water pump pressure fluctuations, and fan speed deviations, precise control parameters matching different process stages can be dynamically generated under complex operating conditions. The adaptive control model performs nonlinear parameter compensation based on the coupling relationship of equipment operating states, effectively suppressing the interference of combustion temperature fluctuations on spray uniformity. Simultaneously, real-time calibration of water pressure compensation parameters reduces the impact of pulsation effects on atomization quality. The dynamic adjustment mechanism of the fan blade angle, combined with speed correction parameters, can simultaneously optimize airflow distribution and powder deposition efficiency, ensuring the stability of coating thickness. Continuous iterative updates of fuel flow feedback data enable the equipment's self-learning capability, significantly improving the dynamic response accuracy of the control strategy. Furthermore, the synchronous integration of harmful gas concentration monitoring indicators and powder room safety indicators achieves multi-dimensional mapping between process safety parameters and MES production data, ensuring that the optimized configuration of the electrical control cabinet guarantees production continuity while proactively mitigating environmental risks. With this design, the embodiments of this application break through the limitations of traditional single-device independent control. Through parameter collaborative optimization of device groups and bidirectional data interaction with the MES system, an intelligent spraying control system with environmental adaptive characteristics is constructed, realizing comprehensive control processing with multi-parameter collaborative optimization and system-level intelligent decision-making capabilities.

[0179] See Figure 2 As shown in the figure, this figure is a schematic diagram of the basic structure of a ceramic spraying intelligent control system 200 provided in an embodiment of this application. The ceramic spraying intelligent control system 200 includes: a processor 201;

[0180] The storage device 202 stores a computer program 2020; when the computer program 2020 is executed by the processor 201, the processor 201 implements any of the above-described intelligent control methods for ceramic spraying based on parameter adaptation.

[0181] Based on the above, a readable storage medium is provided, on which a program or instructions are stored, and when the program or instructions are executed by a processor, the steps of the above method are implemented.

[0182] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems or apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and relevant parts can be referred to the method section.

[0183] In the technical solutions involved in the above embodiments of this application, whether it is performing comparison calculations of multi-dimensional features or constructing composite parameters, if there are problems caused by significant differences in the number of dimensions, units of measurement, and semantic meanings of different features, those skilled in the art, based on their professional knowledge and past practical experience, can fully understand that these differences need to be properly handled so that the calculation results are accurate and comparable, and to avoid situations such as logical confusion and unclear mathematical meaning.

[0184] In detail, when faced with features of different numbers of dimensions, in order to accurately calculate the similarity, matching degree or feature distance between different features, those skilled in the art can use a variety of strategies, such as feature selection, feature extraction, kernel function and other strategies for adaptive processing.

[0185] In order to achieve comparability alignment of feature spaces when processing the comparison of multidimensional features, those skilled in the art can use a variety of existing common technical means, including but not limited to the following existing technologies: standardization preprocessing, mapping transformation, spatial projection, etc.

[0186] In the process of constructing composite parameters (such as loss function values), different parameter terms often have different dimensions. Those skilled in the art can use existing normalization processing or adaptive weight allocation mechanisms based on distribution characteristics.

[0187] The aforementioned general techniques for solving feature matching and loss balance problems are all common knowledge in this field. These techniques have been fully verified and widely used in numerous practical applications, and those skilled in the art can skillfully and flexibly apply these methods to handle similar problems involving differences in dimensions.

[0188] The formulas and calculation processes involved in the embodiments of this application, whether used for multidimensional feature comparison or composite loss function construction, strictly adhere to the principle of dimensional correspondence. Each variable in the formula has a clear and explicit physical meaning, and its operational logic fully conforms to basic mathematical and physical logic. The calculation results are necessarily the reasonable results expected by this application. Those skilled in the art are capable of effectively solving various problems arising from the number of dimensions, dimensional differences, etc., in the multidimensional feature comparison calculation and composite loss function construction in the embodiments, based on specific data conditions and business needs, by comprehensively utilizing the aforementioned general technical means, thus ensuring the accuracy, reliability, and feasibility of the technical solution of this application.

Claims

1. A parameter-adaptive-based intelligent regulation method for ceramic spraying, characterized in that, The method comprises the following steps: acquiring a set of equipment operating states in a spraying flow line process digital panel, the set of equipment operating states including fuel supply efficiency and combustion temperature data of a combustion engine, water pressure fluctuation data of a water pump, rotational speed deviation value of a fan, water-gas consumption data, and harmful gas concentration monitoring indicators; calling a trained adaptive control model to perform dynamic parameter analysis processing on the set of equipment operating states, to generate fuel supply control parameters of the combustion engine, water pressure compensation parameters of the water pump, and rotational speed correction parameters of the fan; based on the fuel supply control parameters, dynamically adjusting the fuel valve of the combustion engine, based on the water pressure compensation parameters, dynamically calibrating the motor power of the water pump, based on the rotational speed correction parameters, dynamically adjusting the blade angle of the fan, and real-time collecting fuel flow feedback data of the combustion engine, output water pressure calibration data of the water pump, and rotational speed adjustment data of the fan, to update the set of equipment operating states; synchronizing the updated set of equipment operating states, the harmful gas concentration monitoring indicators, and the booth safety indicators to an MES system, triggering the MES system to optimize the production data of the electric control cabinet, and updating the real-time process parameters in the spraying flow line process digital panel.

2. The method of claim 1, wherein, The method comprises the following steps: collecting real-time fuel flow data and combustion temperature data of the combustion engine through a sensor network of the electric control cabinet, generating fuel supply efficiency based on the difference between the real-time fuel flow data and a preset fuel threshold, and generating combustion engine state characteristics by associating the combustion temperature data with the fuel supply efficiency; collecting the fluctuation difference between the input water pressure and the output water pressure of the water pump, and generating water pressure fluctuation data by combining the historical mean of water flow in the water-gas consumption data with the real-time water flow; acquiring the matching degree of the real-time rotational speed data of the fan and a preset rotational speed curve, determining the rotational speed deviation value based on the matching degree, and generating fan state characteristics by associating the nitrogen oxide concentration in the harmful gas concentration monitoring indicators; extracting abnormal action events and booth action times from the booth safety indicators, associating the abnormal action events with the harmful gas concentration monitoring indicators, and generating booth environment safety data; integrating the combustion engine state characteristics, the water pressure fluctuation data, the fan state characteristics, the water-gas consumption data, and the booth environment safety data into the set of equipment operating states, and synchronizing the set of equipment operating states to the spraying flow line process digital panel.

3. The method of claim 1, wherein, The training process of the adaptive control model comprises the following steps: extracting fuel supply time sequence characteristics, combustion temperature fluctuation characteristics of the combustion engine, water pressure fluctuation correlation characteristics of the water pump, rotational speed dynamic response characteristics of the fan, and time sequence data of booth action times from a historical production data set; performing multi-dimensional fusion on the fuel supply time sequence characteristics and the combustion temperature fluctuation characteristics, to generate combustion engine comprehensive state characteristics; performing correlation analysis on the water pressure fluctuation correlation characteristics and water flow abnormal events in the water-gas consumption data, to generate water pump dynamic compensation characteristics; performing time sequence matching on the rotational speed dynamic response characteristics and ventilation abnormal events in the booth action times, to generate fan cooperative control characteristics; The training data set is constructed based on the comprehensive state characteristics of the combustion machine, the dynamic compensation characteristics of the water pump, and the fan cooperative regulation characteristics, a training label is generated by a harmful gas concentration change value in the mill house safety index, an initial neural network model is iteratively optimized until an error between regulation parameters output by the initial neural network model and the training label is lower than a preset threshold, and the adaptive regulation model is obtained.

4. The method of claim 1, wherein, The trained adaptive regulation model is called to perform dynamic parameter analysis processing on the set of equipment operating states, generate fuel supply regulation parameters of the combustion machine, water pressure compensation parameters of the water pump, and speed correction parameters of the fan, including: The combustion machine regulation branch of the adaptive regulation model performs multi-scale feature extraction on the fuel supply efficiency and combustion temperature data to generate combustion fluctuation characteristics; the combustion fluctuation characteristics are input into the parallel full connection layer of the combustion machine regulation branch for nonlinear mapping to output a dynamic adjustment coefficient of the fuel valve opening degree and a combustion temperature compensation gradient, and the dynamic adjustment coefficient and the combustion temperature compensation gradient are fused to generate the fuel supply regulation parameters; The fuel supply regulation parameters and the water pressure fluctuation data are jointly input into the water pump compensation branch of the adaptive regulation model, the water pressure fluctuation data are trend decomposed by a sliding time window to separate water pressure trend items and water pressure residual items; the water pressure trend items and the water pressure residual items are input into a bidirectional long short-term memory network for pressure fluctuation prediction to generate a water pressure fluctuation prediction result; real-time water flow in the water-gas consumption data and historical mean values are input into a difference calculation unit to generate a dynamic compensation coefficient; the water pressure fluctuation prediction result and the dynamic compensation coefficient are weighted and fused to generate the water pressure compensation parameters; The water pressure compensation parameters and the speed deviation value are input into the fan cooperative module of the adaptive regulation model to construct a joint feature space of the speed deviation value and the harmful gas concentration monitoring index; topological association features of fan nodes and mill house gas concentration monitoring nodes in the joint feature space are extracted by a graph convolution network; the topological association features are input into a time sequence attention mechanism for dynamic weighting to generate a speed adjustment demand weight; the water pressure compensation parameters are calibrated for ventilation compensation duration according to the speed adjustment demand weight to generate the speed correction parameters including a blade angle adjustment amount, a speed change rate, and a ventilation compensation duration.

5. The method of claim 1, wherein, The fuel valve of the combustion machine is dynamically adjusted based on the fuel supply regulation parameters, the motor power of the water pump is dynamically calibrated based on the water pressure compensation parameters, the blade angle of the fan is dynamically adjusted based on the speed correction parameters, and fuel flow feedback data of the combustion machine, output water pressure calibration data of the water pump, and speed adjustment data of the fan are collected in real time to update the set of equipment operating states, including: The fuel supply regulation parameters are input into a control unit of the combustion machine, a dynamic adjustment coefficient in the fuel supply regulation parameters is used to generate a fuel valve opening degree instruction, a proportional electromagnetic valve of the fuel valve is driven to perform opening and closing actions, real-time fuel flow feedback data and combustion temperature data after the actions are collected, and updated fuel supply efficiency is generated; The water pressure compensation parameters and the updated fuel supply efficiency are input into a frequency converter control module of the water pump, a dynamic compensation coefficient in the water pressure compensation parameters is used to adjust a pulse width modulation signal of motor power, a water pump motor is driven to change output torque, output water pressure calibration data and real-time water flow after the actions are collected, and calibrated water pressure fluctuation data are generated in combination with water and gas consumption data; The calibrated water pressure fluctuation data and the rotational speed correction parameters are input into a servo controller of the fan, a blade angle adjustment amount and ventilation compensation time length in the rotational speed correction parameters are used to generate a blade deflection pulse signal, a blade angle execution mechanism is driven to perform positioning calibration, real-time rotational speed adjustment data and harmful gas concentration monitoring indicators after the actions are collected, and a dynamic ventilation compensation coefficient is generated; The updated fuel supply efficiency, the calibrated water pressure fluctuation data, the real-time rotational speed adjustment data, and the dynamic ventilation compensation coefficient are subjected to multi-dimensional data fusion, an updated equipment running state set containing fuel flow feedback data, output water pressure calibration data, and rotational speed adjustment data is generated, and the updated equipment running state set is synchronized to a spraying flow line process digital panel.

6. The method of claim 1, wherein, The updated equipment running state set, the harmful gas concentration monitoring indicators, and the powder room safety indicators are synchronized to an MES system, and the MES system is triggered to optimize production data of an electric control cabinet, including: Production data of the spraying flow line is extracted from the electric control cabinet, and the production data includes spraying thickness, coating uniformity, and drying time; The updated equipment running state set is associated and mapped with the production data, a process optimization strategy set is generated based on an association mapping result and an abnormal action event in the powder room safety indicators; Task priority ranking results are determined for the process optimization strategy set through the MES system, and the task priority ranking results trigger the electric control cabinet to adjust movement parameters and spraying pressure of a spraying mechanical arm.

7. The method of claim 2, wherein, The generation process of the powder room safety indicators includes: Nitrogen oxide concentration and particulate matter concentration in the powder room are collected in real time through a gas sensor, and a harmful gas concentration exceeding event is generated in combination with a preset safety threshold; Spraying interruption times, mechanical arm over-limit movement times, and ventilation system abnormal start-stop events in powder room action times are counted, and an abnormal action event set is generated; The harmful gas concentration exceeding event and the abnormal action event set are associated and analyzed, and a powder room environment safety level is generated; An alarm instruction is triggered based on the powder room environment safety level, and the powder room environment safety level and the alarm instruction are synchronized to a safety monitoring module in the spraying flow line process digital panel.

8. The method of claim 1, wherein, The method further includes: A three-dimensional simulation model of the spraying pipeline is constructed based on a digital twinning strategy, input information of the three-dimensional simulation model including a device running state set, a powder room safety index, and production data of an electric control cabinet; An updated device running state set is loaded in the three-dimensional simulation model, a linkage running process of a combustion machine, a water pump, and a fan is simulated, and a harmful gas diffusion path and a concentration distribution in the powder room are calculated in real time; If the nitrogen oxide concentration or the particulate matter concentration in the simulation result exceeds a preset threshold value, fuel supply control parameters, water pressure compensation parameters, and rotating speed correction parameters are regenerated, and the MES system is triggered to suspend a current production task.

9. The method of claim 8, wherein, The verification process of the three-dimensional simulation model includes: The updated device running state set is input into the three-dimensional simulation model, the influence of fuel supply efficiency variation of the combustion machine on the temperature of the powder room is simulated, and a temperature-concentration correlation curve is generated in association with a harmful gas concentration monitoring index; The effect of water pressure compensation parameters of the water pump on coating uniformity is simulated, and a water pressure-coating quality evaluation result is generated in combination with a drying time in the production data; The ventilation efficiency of the powder room is simulated based on rotating speed correction parameters of the fan, and a ventilation-dust diffusion optimization index is generated in combination with a particulate matter concentration; If the temperature-concentration correlation curve, the water pressure-coating quality evaluation result, or the ventilation-dust diffusion optimization index exceeds a preset range, it is determined that the current control parameters are abnormal, and the adaptive control model is triggered to perform dynamic parameter analysis processing again.

10. A ceramic spraying intelligent regulation system, characterized in that, It includes: a processor; a storage device having a computer program stored thereon, when the computer program is executed by the processor, the processor implements the parameter adaptive-based intelligent control method for ceramic spraying according to any one of claims 1-9.

Images