Intelligent closed-loop control system for passivation solution in hot galvanizing process

Abstract

The invention discloses an intelligent closed-loop control system for a passivation solution in a hot galvanizing process, particularly relates to the technical field of industrial process automatic control, and is used for solving the problem of large parameter fluctuation caused by the fact that an existing passivation solution control mode lags behind intermittent production takt change. A signal monitoring module identifies a production cycle and calculates an estimated compensation value, an entropy weight generation module generates a weight distribution coefficient based on symbol sequence information entropy, and a data fusion module fuses the estimated compensation value and real-time monitoring data to generate a cooperation instruction. The trajectory prediction module simulates and generates a virtual change trajectory and a dynamic envelope line thereof according to the instruction, the envelope monitoring module monitors the running state of the real trajectory in the envelope line in real time, and the execution early warning module executes the instruction or generates an early warning signal according to the monitoring result. According to the system, prospective prediction and self-adaptive control of parameters of the passivation solution are realized, and the control precision and stability are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of industrial process automatic control technology, and more specifically, to an intelligent closed-loop control system for passivation solution in hot-dip galvanizing process. Background Technology

[0002] In the hot-dip galvanizing process, passivation is a crucial step in achieving the final corrosion resistance of the product. This step involves immersing the galvanized parts in a passivation bath with a specific chemical composition, where a chemical reaction occurs to form a protective film. Currently, the industry generally relies on operators' experience to perform intermittent manual checks and adjustments of key parameters of the passivation bath (such as level and concentration), or on simple automatic control methods based on single-point sensor feedback. These methods treat the passivation bath as a static or continuous reaction environment for control.

[0003] However, hot-dip galvanizing production is essentially a typical intermittent operation. The liquid level and the consumption of effective components in the bath are directly driven by the immersion and removal of batches of workpieces, which subject the bath state to periodic, step-like, and severe disturbances. Existing control strategies can only respond slowly after disturbances occur and cause parameter deviations, and their control effect is severely lagging behind the actual dynamic changes in the process. This inherent time delay causes the passivation solution parameters to fluctuate significantly, making it difficult to stabilize within the optimal range required by the process. This directly leads to problems such as uneven film quality, reduced product yield, and increased unnecessary consumption of chemical raw materials. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides an intelligent closed-loop control system for passivation solution in hot-dip galvanizing process to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: The intelligent closed-loop control system for the passivation solution in the hot-dip galvanizing process includes: The signal monitoring module is used to identify the production cycle and calculate the estimated compensation value of the passivation fluid consumption based on the preset fixed compensation amount. The entropy weight generation module is used to obtain a symbol sequence based on the symbolization of the production cycle, calculate the information entropy value of the symbol sequence, and generate weight allocation coefficients through a preset mapping relationship. The data fusion module is used to collect monitoring data on the liquid level and concentration in the passivation tank, and to generate coordinated commands for controlling the replenishment pump and the replenishment pump based on the weight allocation coefficient and the estimated compensation value fused with the monitoring data. The trajectory prediction module is used to simulate and calculate the virtual change trajectory of the key parameters of the tank liquid after expected execution based on the collaborative instructions, and generate a dynamic envelope formed by its upper and lower boundaries based on the virtual change trajectory. The envelope monitoring module is used to monitor whether the real change trajectory, which is composed of real-time collected monitoring data, continues to run within the range of the dynamic envelope line. The execution warning module is used to execute coordinated commands to drive the replenishment pump and the replenishment pump to perform quantitative replenishment operations when the actual change trajectory continues to run within the range of the dynamic envelope. Otherwise, the execution of coordinated commands is prohibited and a system abnormality warning signal is generated.

[0006] Furthermore, the production cycle time is identified, and the estimated compensation value for passivation fluid consumption is calculated based on a preset fixed compensation amount, including: Real-time monitoring of the rising and falling edge signals of the lifting hook of the elevator in the hot-dip galvanizing production line; The falling edge signal of the lifting hook of the elevator is used as the starting point of the production cycle, and the rising edge signal is used as the ending point of the production cycle. The specification markings of the current workpiece basket are read by an RFID tag installed on the workpiece basket; The corresponding fixed compensation amount is called based on the current workpiece basket specification markings; The fixed compensation amount is used as the estimated compensation value.

[0007] Furthermore, based on the symbolization of the production cycle, a symbol sequence is obtained. The information entropy value of the symbol sequence is calculated, and weight allocation coefficients are generated through a preset mapping relationship, including: Record the duration of each production cycle in multiple consecutive production cycle cycles to form a duration sequence; The duration of each production cycle in the time sequence is compared with a preset duration threshold. When the duration of the production cycle is greater than or equal to a preset duration threshold, it is marked as the first symbol; when the duration of the production cycle is less than the preset duration threshold, it is marked as the second symbol, forming a symbol sequence. Calculate the information entropy value based on the probability of different symbols appearing in the symbol sequence; Establish a pre-defined mapping table between information entropy values ​​and weight allocation coefficients; Match the calculated information entropy value with the entropy value interval in the mapping table, find the weight allocation coefficient corresponding to the entropy value interval in the mapping table, and output the matched weight allocation coefficient.

[0008] Furthermore, the pre-defined mapping relationship table between information entropy values ​​and weight allocation coefficients is established in the following way: the numerical range of information entropy values ​​is divided in advance through experiments, and each numerical range is bound to a pre-defined weight allocation coefficient to form an entropy value-weight lookup table.

[0009] Furthermore, monitoring data on the level and concentration of the passivation tank are collected. Based on the weighted allocation coefficient, the estimated compensation value and the monitoring data are fused to generate coordinated commands for controlling the replenishment pump and the replenishment pump, including: Real-time acquisition of liquid level monitoring data collected by the liquid level sensor and concentration monitoring data collected by the concentration sensor; Calculate the level deviation between the liquid level monitoring data and the preset liquid level setting, and the concentration deviation between the concentration monitoring data and the preset concentration setting, respectively. The feedforward compensation amount is obtained by weighting the estimated compensation value using weight allocation coefficients. The feedback compensation amount is obtained by subtracting the weighting factor from 1 and weighting the liquid level deviation value and the concentration deviation value. The feedforward compensation and the feedback compensation are added together to obtain the final compensation amount, which is then converted into a coordinated instruction for the fluid replenishment pump and the drug replenishment pump.

[0010] Furthermore, based on the collaborative instructions, the virtual change trajectory of the key parameters of the tank solution after expected execution is simulated and calculated. A dynamic envelope formed by the upper and lower boundaries of this virtual change trajectory is generated, including: Based on the replenishment amount and replenishment rate contained in the collaborative instruction, the virtual trajectory of the expected change in liquid level and concentration after execution is calculated through the tank liquid dynamic model. Based on the dynamic modal stability index obtained by symbolic dynamic modal analysis of the continuous production cycle duration sequence, the positive and negative tolerance values ​​are dynamically adjusted. The upper boundary trajectory is formed by superimposing dynamically adjusted positive tolerance values ​​at each sampling point of the virtual change trajectory; The lower boundary trajectory is formed by superimposing dynamically adjusted negative tolerance values ​​at each sampling point of the virtual change trajectory; The dynamic envelope is formed by the upper boundary trajectory and the lower boundary trajectory.

[0011] Furthermore, the virtual change trajectory of liquid level and concentration after expected execution is obtained by calculating the dynamic model of the tank liquid. This means that, based on the replenishment amount in the collaborative instruction and the principles of mass conservation and concentration balance, the expected change trajectory of the tank liquid volume and the expected change trajectory of the solute concentration after the replenishment operation are calculated respectively.

[0012] Furthermore, the dynamic modal stability index obtained based on symbolic dynamic modal analysis refers to the process of analyzing the state transition probability matrix of a symbol sequence, calculating the eigenvalue magnitude of its dominant dynamic mode, and using the eigenvalue magnitude as a dynamic modal stability index to quantify the stability of the production cycle. The closer the dynamic modal stability index is to 1, the smaller the tolerance range of the dynamic envelope will be; conversely, the smaller the index is to 1, the larger the tolerance range of the dynamic envelope will be.

[0013] Furthermore, monitoring whether the actual change trajectory, composed of real-time collected monitoring data, continuously operates within the dynamic envelope range includes: The real-time collected liquid level monitoring data and concentration monitoring data are combined in chronological order to form a true change trajectory; At multiple consecutive sampling time points, determine whether the liquid level and concentration values ​​in the actual change trajectory are simultaneously located between the upper and lower boundary trajectories of the dynamic envelope; When the liquid level and concentration values ​​at multiple consecutive sampling time points are simultaneously between the upper and lower boundary trajectories, it is determined that the actual change trajectory continues to run within the dynamic envelope range. When the liquid level or concentration value at any sampling time point exceeds the upper or lower boundary trajectory, it is determined that the true change trajectory does not continuously run within the dynamic envelope range.

[0014] Furthermore, when the actual trajectory of change continuously operates within the range of the dynamic envelope, the coordinated command is executed to drive the replenishment pump and the replenishment pump to perform quantitative replenishment operations; otherwise, the coordinated command is prohibited from being executed and a system abnormality warning signal is generated, including: When it is determined that the actual change trajectory continues to run within the range of the dynamic envelope, the coordinated command is converted into the start / stop control signal of the replenishment pump and the flow regulation signal of the replenishment pump. When it is determined that the actual change trajectory does not continuously run within the range of the dynamic envelope, the transmission of the coordination command to the infusion pump and the medication pump is blocked; at the same time, a system anomaly warning signal containing abnormal timestamps and abnormal parameter types is generated.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. By employing a forward-looking dynamic prediction and adaptive control mechanism, the response mode of traditional passivation fluid management has been transformed. By identifying the production cycle and calculating the estimated compensation value, it achieves advance perception of step disturbances. By generating dynamic weights through symbolic entropy value analysis of production cycle stability, it achieves adaptive matching between the control strategy and real-time operating conditions. By integrating predictive feedforward and monitoring feedback to generate coordinated instructions, it achieves precise advance compensation. This control method, based on process dynamic characteristics rather than simple parameter deviations, significantly improves the system's adaptability to intermittent production processes, enabling replenishment operations to synchronize with the production cycle and effectively avoiding control lag problems.

[0016] 2. By constructing a dynamic envelope monitoring mechanism, a comprehensive quality control system was established. Virtual change trajectories and their envelope boundaries were generated through simulation and prediction, providing a dynamic evaluation benchmark for real-time monitoring. Real-time verification of the control effect was achieved by continuously tracking the actual trajectory's operation within the envelope. Intelligent linkage between execution and early warning ensured that replenishment operations were only performed when the system was operating normally. This dynamic envelope-based monitoring and execution mechanism not only guaranteed the accuracy and reliability of replenishment operations but also formed a complete quality closed-loop control, significantly improving the control accuracy and stability of the passivation fluid parameters. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the intelligent closed-loop control system for the passivation solution in the hot-dip galvanizing process of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Example: Figure 1 A schematic diagram of the intelligent closed-loop control system for the passivation solution in the hot-dip galvanizing process of this invention is provided. The intelligent closed-loop control system for the passivation solution in the hot-dip galvanizing process includes: The signal monitoring module is used to identify the production cycle and calculate the estimated compensation value of the passivation fluid consumption based on the preset fixed compensation amount. The entropy weight generation module is used to obtain a symbol sequence based on the symbolization of the production cycle, calculate the information entropy value of the symbol sequence, and generate weight allocation coefficients through a preset mapping relationship. The data fusion module is used to collect monitoring data on the liquid level and concentration in the passivation tank, and to generate coordinated commands for controlling the replenishment pump and the replenishment pump based on the weight allocation coefficient and the estimated compensation value fused with the monitoring data. The trajectory prediction module is used to simulate and calculate the virtual change trajectory of the key parameters of the tank liquid after expected execution based on the collaborative instructions, and generate a dynamic envelope formed by its upper and lower boundaries based on the virtual change trajectory. The envelope monitoring module is used to monitor whether the real change trajectory, which is composed of real-time collected monitoring data, continues to run within the range of the dynamic envelope line. The execution warning module is used to execute coordinated commands to drive the replenishment pump and the replenishment pump to perform quantitative replenishment operations when the actual change trajectory continues to run within the range of the dynamic envelope. Otherwise, the execution of coordinated commands is prohibited and a system abnormality warning signal is generated.

[0020] This intelligent closed-loop control system for the passivation solution in the hot-dip galvanizing process identifies the production cycle and calculates the estimated compensation value in the following way: It monitors the rising and falling edge signals of the lifting hook in the hot-dip galvanizing production line in real time. Specifically, signal acquisition is achieved through a proximity switch installed at the end of the lifting hook's running track. This proximity switch uses an inductive sensor with a detection distance set to 10 mm to 15 mm. When the lifting hook passes through the proximity switch's sensing area, a change in signal level is generated. The control system records the moment the level changes from low to high as the rising edge signal and the moment it changes from high to low as the falling edge signal. The signal sampling frequency is 1000 Hz to ensure accurate capture of signal edge changes. The falling edge signal of the lifting hook is used as the starting point of the production cycle, corresponding to the time when the workpiece basket begins to be immersed in the passivation tank. The rising edge signal is used as the ending point of the production cycle, corresponding to the time when the workpiece basket completely leaves the passivation tank surface. By recording the time interval between adjacent falling and rising edge signals, the duration of a complete production cycle can be obtained, with time measurement accuracy reaching the millisecond level.

[0021] The specification markings of the current workpiece basket are read by an RFID tag installed on it. This RFID tag is a high-frequency electronic tag with an operating frequency of 13.56 MHz, which complies with the RFID standard set by the International Organization for Standardization. Before each workpiece basket is put into use, an electronic tag with a unique code is installed at its designated location. This code establishes a correspondence with the specification information of the workpiece basket. The reading device is an RFID reader / writer fixedly installed at the feed inlet of the passivation tank. The reader / writer antenna gain is 6dBi, and the transmission power is adjustable from 0.5 watts to 2 watts. When the workpiece basket moves to the feed inlet position with the elevator, the reader / writer transmits an RFID signal at a frequency of 10 times per second to activate the electronic tag and read its stored specification marking information. The reading distance is controlled within the range of 20 cm to 50 cm to ensure reading accuracy. The read specification markings are transmitted to the control system for processing via industrial Ethernet, using the Modbus TCP protocol.

[0022] The control system retrieves the corresponding fixed compensation amount based on the current workpiece basket's specification identifier. It pre-stores a data table of fixed compensation amounts for different workpiece basket specifications, determined through extensive process testing. Specifically, under standard operating conditions, the volume of passivation liquid carried out by different specifications of workpiece baskets after completing one production cycle is statistically analyzed. The mass of the carried-out liquid is measured by weighing and converted to volume units. Weighing is performed using an electronic balance with an accuracy of 0.1 grams. Each specification of workpiece basket undergoes at least 30 repeated tests, and the average value is taken as the fixed compensation amount for that specification. For example, a workpiece basket with specification identifier A corresponds to a fixed compensation amount of 1.5 liters, and a workpiece basket with specification identifier B corresponds to a fixed compensation amount of 2.2 liters. These correspondences are stored in the control system's non-volatile memory in the form of a data table. The storage format is a two-dimensional array structure, with the first column being the specification identifier code and the second column being the corresponding fixed compensation amount value. Upon receiving a specification identifier, the control system retrieves the corresponding fixed compensation amount value by looking up the table. The lookup algorithm uses a binary search method to improve retrieval efficiency.

[0023] A fixed compensation amount is used as the estimated compensation value. This estimated compensation value represents the basic amount of passivating fluid that needs to be replenished due to workpiece carry-out during the current production cycle. This value serves as the reference value for feedforward control and participates in subsequent data fusion calculations. The control system stores the acquired fixed compensation amount in designated data registers, with addresses ranging from 0x4000 to 0x40FF. Each register stores a 32-bit floating-point number for subsequent data fusion module calls, thereby achieving a preliminary estimate of passivating fluid consumption. The accuracy of the estimated compensation value is maintained to one decimal place, with the unit uniformly set to liters to ensure consistency with the dimensions of subsequent calculations. The entire identification and calculation process is automatically triggered at the beginning of each production cycle and automatically completed at the end of the cycle, ensuring real-time performance and synchronization with the production rhythm.

[0024] The intelligent closed-loop control system for the passivation solution in the hot-dip galvanizing process achieves symbolic processing and generation of weight allocation coefficients for the production cycle in the following way: the duration of each production cycle in multiple consecutive production cycles is recorded to form a duration sequence. Specifically, the duration data of the most recent 32 production cycles are recorded. These duration data are derived from the time interval measurement between the falling edge signal and the rising edge signal of the lifting hook of the elevator. The time measurement is achieved using a high-precision timer with a timing accuracy of 1 millisecond. All duration data are stored in a circular buffer in the form of 32-bit floating-point numbers. The buffer size is fixed at 32 elements. A first-in-first-out data update strategy is adopted, and the duration data of each new production cycle will overwrite the oldest data in the buffer. The duration of each production cycle in the time sequence is compared with a preset duration threshold, which is determined by statistical analysis. Specifically, the calculation method is to collect the duration data of the most recent 100 production cycle periods, remove the maximum and minimum values, take the arithmetic mean, and then multiply it by a coefficient of 0.8. For example, when the average production cycle period is 120 seconds, the preset duration threshold is set to 96 seconds. This threshold is recalculated every 24 hours according to the production situation to ensure that the threshold can adapt to changes in the production rhythm.

[0025] When the duration of a production cycle is greater than or equal to a preset duration threshold, it is marked as the first symbol, represented by the uppercase letter A. When the duration of a production cycle is less than the preset duration threshold, it is marked as the second symbol, represented by the uppercase letter B. This forms a symbol sequence. The length of the symbol sequence is consistent with the duration sequence. For example, 32 production cycle durations correspond to a symbol sequence of 32 symbols. The symbol sequence is stored in a specific memory area in the form of a string, with each symbol occupying one byte of storage space. The information entropy value is calculated based on the probability of different symbols appearing in a symbol sequence. First, the number of times each symbol appears in the symbol sequence is counted. Then, the probability of each symbol appearing is calculated. Specifically, the number of times a symbol appears is divided by the total length of the symbol sequence. For example, in a sequence of 32 symbols, if symbol A appears 20 times, then the probability of symbol A is 20 / 32 = 0.625, and the probability of symbol B is 12 / 32 = 0.375. The formula for calculating the information entropy value is: H = -Σ(pi × log2(pi)), where pi represents the probability of the i-th symbol appearing. The specific calculation process is to first take the logarithm of the probability of each symbol to the base 2, then multiply the probability by the logarithm, and finally sum the products of all symbols and take the negative number. During the calculation process, the calculation of the logarithm is implemented using an approximate algorithm to ensure that the calculation efficiency meets the real-time requirements.

[0026] A pre-defined mapping table between information entropy values ​​and weight allocation coefficients is established. This mapping table is determined through system process experiments. The experimental method involves recording multiple sets of information entropy values ​​and the corresponding weight allocation coefficients required for optimal control effects under different production conditions. The range of information entropy values ​​from 0 to 1 is divided into 10 equally spaced intervals, each with a length of 0.1. An optimal weight allocation coefficient is assigned to each entropy value interval by combining expert experience with experimental data. The range of the weight allocation coefficient is limited to 0 to 1. For example, when the information entropy value is in the interval [0, 0.1), the corresponding weight allocation coefficient is 0.9; when it is in the interval [0.1, 0.2), the corresponding weight allocation coefficient is 0.8, and so on, forming a complete entropy-weight lookup table. This lookup table is stored in non-volatile memory in the form of a two-dimensional array. The first column stores the lower limit of the entropy value interval, and the second column stores the corresponding weight allocation coefficient.

[0027] The calculated information entropy value is matched with the entropy value interval in the mapping table. An interval binary search algorithm is used to quickly locate the interval range of the current information entropy value in the entropy-weight lookup table. During the search process, the information entropy value is first compared with the lower limit of each interval in the lookup table to determine its specific interval. The weight allocation coefficient corresponding to the entropy value interval of the information entropy value is found in the mapping table, and the matched weight allocation coefficient is output. The weight allocation coefficient is a 32-bit floating-point number with precision retained to two decimal places. The output weight allocation coefficient is stored in a designated data register for subsequent data fusion module calls. The entire symbolization processing and weight calculation process is automatically executed once at the end of each production cycle to ensure that the weight allocation coefficient can reflect the changes in production cycle in real time. An exception handling mechanism is set in the calculation process. When the information entropy value exceeds the expected range, the default weight allocation coefficient of 0.5 is automatically used for processing.

[0028] The intelligent closed-loop control system for the passivation solution in the hot-dip galvanizing process collects monitoring data and generates collaborative instructions in the following ways: It acquires real-time liquid level monitoring data from a liquid level sensor and concentration monitoring data from a concentration sensor. The liquid level sensor is a hydrostatic level transmitter, installed on the side wall of the passivation tank 50 cm from the bottom, with a measurement range of 0-5 meters and an accuracy class of 0.5. Its output signal is a 4-20mA analog signal, collected by a programmable logic controller (PLC) at a sampling interval of 100 milliseconds. The acquired liquid level monitoring data is processed by a median filtering algorithm and stored in data registers D1000-D1031. The concentration sensor is an online densitometer, measuring the density of the passivation solution based on the vibrating tube principle and converting it into a concentration value. It is installed on the circulation pipe of the passivation tank, with a measurement range of 1.0-1.2 g / cm³ and an accuracy of ±0.001 g / cm³. Its output signal is also a 4-20mA analog signal, with the sampling interval synchronized with the liquid level monitoring. The collected data is stored in data registers D1100-D1131.

[0029] The liquid level deviation between the liquid level monitoring data and the preset liquid level setting, and the concentration deviation between the concentration monitoring data and the preset concentration setting are calculated separately. The preset liquid level setting is set to 80% of the effective height of the tank according to process requirements. For example, when the effective height of the tank is 2 meters, the liquid level setting is 1.6 meters. This setting is stored in data register D2000. The preset concentration setting is set to a specific value according to passivation process requirements, such as 120 g / L. This setting is stored in data register D2001. The liquid level deviation is calculated by subtracting the preset liquid level setting from the current liquid level monitoring data. The result is stored in data register D3000 as a 32-bit floating-point number, with the unit remaining in meters. The concentration deviation is calculated by subtracting the preset concentration setting from the current concentration monitoring data. The result is stored in data register D3001 as a 32-bit floating-point number, with the unit remaining in grams per liter.

[0030] The feedforward compensation amount is obtained by weighting the estimated compensation value using a weighting coefficient. The weighting coefficient is derived from the output of the aforementioned symbolization process and stored in data register D4000, with a value range of 0-1. The estimated compensation value is derived from a fixed compensation amount obtained based on the workpiece basket specification identifier and stored in data register D4001. The feedforward compensation amount is calculated by multiplying the weighting coefficient by the estimated compensation value, and the result is stored in data register D5000. For example, when the weighting coefficient is 0.8 and the estimated compensation value is 2.0 liters, the calculated feedforward compensation amount is 1.6 liters.

[0031] The feedback compensation amount is obtained by weighting the liquid level deviation and concentration deviation values ​​using 1 minus the weighting coefficient. First, the liquid level deviation and concentration deviation values ​​are multiplied by (1 - weighting coefficient) respectively to obtain the weighted liquid level compensation component and concentration compensation component. The liquid level compensation component is stored in data register D6000, and the concentration compensation component is stored in data register D6001. Then, the two components are added together to obtain the total feedback compensation amount, which is stored in data register D6002. For example, when the weighting coefficient is 0.8, the weighting coefficient is 0.2. If the liquid level deviation is -0.1 meters and the concentration deviation is -5 g / L, the liquid level deviation is first converted into a volume deviation and calculated based on the cross-sectional area of ​​the tank. For example, when the cross-sectional area of ​​the tank is 2 square meters, a liquid level deviation of -0.1 meters corresponds to a volume deviation of -200 liters. The concentration deviation is calculated based on the current total amount of tank liquid and the concentration deviation value. For example, when the total amount of tank liquid is 5000 liters, a concentration deviation of -5 g / L requires the addition of 25 kg of reagent.

[0032] The feedforward compensation and feedback compensation are added together to obtain the final compensation amount, which includes two parts: the replenishment volume and the replenishment volume. The replenishment volume is calculated based on the level-related compensation component and stored in data register D7000; the replenishment volume is calculated based on the concentration-related compensation component and stored in data register D7001. The final compensation amount is converted into coordinated instructions for the replenishment pump and the replenishment pump according to a predetermined conversion rule. The replenishment pump control instructions are converted into pump running time signals. The running time is calculated by dividing the replenishment volume by the pump's rated flow rate. For example, if the rated flow rate of the replenishment pump is 10... The pump speed is calculated to be 20 minutes when the replenishment volume is 200 liters per minute. The replenishment pump control command is converted into a pump flow rate adjustment signal. The proportional-integral-derivative control algorithm is used to adjust the pump speed to achieve precise dosing. The flow rate setpoint is the replenishment volume divided by the replenishment time. For example, if the replenishment is required to be completed within 30 minutes, the flow rate setpoint is 25 kg / 30 minutes. The coordination command is sent to the actuators of the replenishment pump and the replenishment pump through programmable logic control. The digital output signal controls the start and stop of the pump, and the analog output signal adjusts the pump speed to ensure the precise execution of the replenishment operation.

[0033] The intelligent closed-loop control system for the passivation solution in the hot-dip galvanizing process generates a dynamic envelope in the following way: Based on the replenishment amount and replenishment rate contained in the collaborative instruction, the system calculates the virtual change trajectory of the liquid level and concentration after the expected execution through the dynamic model of the bath solution. The replenishment amount comes from the final compensation amount calculated by the data fusion module, which includes two parts: the replenishment liquid amount and the replenishment chemical amount. The replenishment rate is set according to the production process requirements, for example, the rated flow rate of the replenishment pump is 10 liters / minute, and the rated flow rate of the replenishment chemical pump is 5 liters / minute. The dynamic model of the bath solution is established based on the principles of mass conservation and concentration balance. The principle of mass conservation is reflected in the fact that the change in the total volume of the bath solution after the replenishment operation is equal to the change in the volume of the replenished liquid minus... The carryover from the production process is eliminated. The concentration balance principle is reflected in the fact that the change in the total amount of solute in the tank solution after the replenishment operation is equal to the amount of solute replenished minus the amount consumed in the production process. The calculation process of the virtual change trajectory is based on the current state of the tank solution as the initial condition, which includes the current volume and concentration of the tank solution. The replenishment amount and replenishment rate are used as input parameters. The change of tank solution parameters in the future is predicted by numerical integration method. The time step is set to 1 second and the prediction duration is set to 30 minutes. The calculation results include the liquid level change trajectory and the concentration change trajectory, which are stored in the data registers D8000-D8099 and D8100-D8199 in the form of arrays, respectively.

[0034] Based on the dynamic modal stability index obtained from symbolic dynamic modal analysis of continuous production cycle duration sequences, positive and negative tolerance values ​​are dynamically adjusted. The continuous production cycle duration sequence is derived from the duration data of the most recent 32 production cycle durations recorded by the signal monitoring module. The symbolic dynamic modal analysis process includes converting the duration sequence into a symbolic sequence, constructing a state transition probability matrix, calculating the eigenvalues ​​and eigenvectors of this matrix, and selecting the eigenvalue magnitude corresponding to the dominant mode with the largest eigenvalue magnitude as the dynamic modal stability index. The dynamic modal stability index ranges from 0 to 1; the closer the index value is to 1, the more stable the production cycle. Stability; the adjustment rules for positive and negative tolerance values ​​are to establish a mapping relationship between the dynamic modal stability index and the tolerance adjustment coefficient. For example, when the dynamic modal stability index is greater than 0.9, the tolerance adjustment coefficient is set to 0.5; when the index is between 0.7 and 0.9, the tolerance adjustment coefficient is set to 1.0; and when the index is less than 0.7, the tolerance adjustment coefficient is set to 1.5. The reference tolerance value is set according to the process requirements. For example, the reference value for liquid level tolerance is ±0.05 meters, and the reference value for concentration tolerance is ±5 g / L. The positive tolerance value is equal to the reference tolerance value multiplied by the tolerance adjustment coefficient, and the negative tolerance value is also equal to the reference tolerance value multiplied by the tolerance adjustment coefficient, but with a negative sign.

[0035] When evaluating production stability by performing symbolic dynamic modal analysis on the duration sequence of continuous production cycle times, the analysis is conducted on a shift basis, with each shift lasting 8 hours, rather than analyzing every single production cycle. The symbolic dynamic modal analysis process includes collecting the duration data of all production cycle times within a shift, comparing these duration data with a preset duration threshold and converting them into symbolic sequences, then constructing a symbolic state transition probability matrix, calculating the eigenvalues ​​of this matrix, and taking the modulus of the largest eigenvalue as the dynamic modal stability index. The tank liquid dynamic model serves as an ideal reference model and needs to be calibrated experimentally before practical application. The calibration process includes recording actual tank liquid parameter changes under standard operating conditions, comparing them with the model's predicted values, and adjusting the model parameters using the least squares method to ensure that the prediction error is less than 5%.

[0036] The upper boundary trajectory is formed by superimposing dynamically adjusted positive tolerance values ​​on each sampling point of the virtual change trajectory. The specific operation process is to traverse all sampling points of the virtual change trajectory and add the corresponding positive tolerance value to the value of each sampling point. For example, if the liquid level prediction value at time point t is L(t) and the positive tolerance is ΔL+, then the liquid level value of the upper boundary trajectory at that time point is L(t)+ΔL+. Similarly, if the concentration prediction value is C(t) and the positive tolerance is ΔC+, then the concentration value of the upper boundary trajectory at that time point is C(t)+ΔC+. The sampling point data of the upper boundary trajectory is stored in data registers D8200-D8299.

[0037] The lower boundary trajectory is formed by superimposing dynamically adjusted negative tolerance values ​​on each sampling point of the virtual change trajectory. The specific operation process is to traverse all sampling points of the virtual change trajectory and add the corresponding negative tolerance value to the value of each sampling point. For example, if the liquid level prediction value at time point t is L(t) and the negative tolerance is ΔL-, then the liquid level value of the lower boundary trajectory at that time point is L(t)+ΔL-. Similarly, if the concentration prediction value is C(t) and the negative tolerance is ΔC-, then the concentration value of the lower boundary trajectory at that time point is C(t)+ΔC-. The sampling point data of the lower boundary trajectory are stored in data registers D8300-D8399.

[0038] The dynamic envelope is formed by the upper and lower boundary trajectories. It comprises two parts: a liquid level envelope and a concentration envelope. The liquid level envelope is defined by the liquid level data in both the upper and lower boundary trajectories, and the concentration envelope is defined by the concentration data in both trajectories. The dynamic envelope is stored in a data structure containing three arrays: a time series, an upper boundary value series, and a lower boundary value series. These arrays are used by the subsequent envelope monitoring module for trajectory comparison. The dynamic envelope is updated once every production cycle to ensure timely reflection of changes in the production process. An anomaly handling mechanism is implemented during the calculation process. When the dynamic modal stability index exceeds the expected range, a default tolerance adjustment coefficient of 1.0 is automatically applied. When an abnormal value appears in the calculation result of the tank liquid dynamic model, the calculation result from the previous cycle is automatically used to replace it, ensuring stable system operation.

[0039] The dynamic model of the bath solution is established based on the principles of mass conservation and concentration balance, and its core equation is: and Where V is the tank liquid volume (m³), C is the concentration (g / L), Qin is the replenishment flow rate (m³ / s), Qout is the carry-out flow rate (m³ / s), Cin is the replenishment liquid concentration (g / L), and Cout is the carry-out liquid concentration (g / L). The model parameters were calibrated through the following experiments: First, at least 30 sets of actual carry-out data under different production cycles were recorded under steady-state conditions, and the least squares method was used for fitting. The coefficient k1 (range 0.6-0.9) is used, where A is the cross-sectional area of ​​the workpiece basket (m²) and v is the lifting speed (m / s). Subsequently, the concentration dynamic response characteristics are determined using the tracer method to determine the concentration diffusion coefficient D (typical value 1.2 × 10⁻⁶). -9 Up to 5.8×10 -9 The model requires a mixed time constant τ (range 60 to 180 s) and a relative error between the predicted and measured values ​​of less than 5%, and a coefficient of determination R² > 0.92.

[0040] The intelligent closed-loop control system for the passivation solution in the hot-dip galvanizing process monitors whether the actual change trajectory continuously operates within the dynamic envelope range in the following way: Real-time collected liquid level monitoring data and concentration monitoring data are combined in chronological order to construct the actual change trajectory. The liquid level monitoring data comes from a hydrostatic liquid level transmitter installed on the side wall of the passivation tank. This transmitter has a measurement range of 0-5 meters, an accuracy class of 0.5, and an output signal of 4-20mA analog signal. It is acquired by a programmable logic controller at a sampling interval of 100 milliseconds. The acquired liquid level monitoring data is digitally filtered and stored in data registers D9000-D9099; concentration monitoring... The data comes from an online densitometer installed on the circulation pipeline. This densitometer measures the density of the passivation liquid based on the principle of a vibrating tube and converts it into a concentration value. The measurement range is 1.0 to 1.2 g / cm³, with an accuracy of ±0.001 g / cm³. The output signal is also a 4 to 20 mA analog signal. The acquisition interval is synchronized with the liquid level monitoring. The acquired data is stored in data registers D9100-D9199. The actual change trajectory is stored in the form of a time series. Each time point corresponds to a liquid level value and a concentration value. The time series length is the monitoring data of the most recent 300 seconds, corresponding to 3000 sampling points. A circular buffer storage method is used, with new data overwriting old data.

[0041] At multiple consecutive sampling time points, it is determined whether the liquid level and concentration values ​​in the actual change trajectory are simultaneously between the upper and lower boundary trajectories of the dynamic envelope. The dynamic envelope is derived from the calculation results of the trajectory prediction module and includes the upper and lower boundary trajectories, which are stored in data registers D8200-D8299 and D8300-D8399, respectively. The judgment process adopts a point-by-point comparison algorithm. First, the liquid level monitoring value and concentration monitoring value at the current time point are obtained. Then, the upper and lower boundary values ​​at the same time are obtained from the dynamic envelope. The judgment condition for the liquid level value is that the current liquid level monitoring value is greater than or equal to the lower boundary liquid level value and less than or equal to the upper boundary liquid level value. The judgment condition for the concentration value is that the current concentration monitoring value is greater than or equal to the lower boundary concentration value and less than or equal to the upper boundary concentration value. Both conditions must be met simultaneously to determine that the sampling point is within the envelope range. The number of sampling points for continuous judgment is set according to process requirements, for example, 32 consecutive sampling points, corresponding to a 3.2-second time window. This number is stored in data register D9400 and can be adjusted according to actual production conditions.

[0042] When the liquid level and concentration values ​​at multiple consecutive sampling time points are simultaneously between the upper and lower boundary trajectories, it is determined that the actual change trajectory is continuously operating within the dynamic envelope range. At this time, the system generates a trajectory normal operation signal, which is a Boolean data stored in data register D9500. A value of 1 indicates that the trajectory is operating normally. At the same time, an execution counter is started to record the number of consecutive normal sampling points. When the counter reaches a set threshold, the subsequent instruction execution process is triggered. The threshold for the number of consecutive normal sampling points is set to 32, and this value is stored in data register D9401. When the counter value reaches 32, the system confirms that the trajectory is in a continuous normal operation state.

[0043] In actual industrial environments, due to unavoidable noise and minor fluctuations in sensor measurements, a reasonable tolerance range and judgment mechanism are set for the condition that "the liquid level and concentration values ​​are simultaneously between the upper and lower boundary trajectories of the dynamic envelope". The system uses a moving window averaging algorithm to preprocess the monitoring data, with the window size set to 8 sampling points, corresponding to a time window of 0.8 seconds, effectively smoothing measurement noise. When judging whether it is within the envelope range, a measurement tolerance band of ±0.5% is set. For example, when the monitored value exceeds the boundary value but is within the tolerance band, it is still judged as meeting the requirements. The threshold for the number of consecutive sampling time points that meet the condition is dynamically adjusted according to the characteristics of the production process. Under normal production conditions, it is set to 16 consecutive sampling points (corresponding to 1.6 seconds), and automatically relaxed to 8 consecutive sampling points (corresponding to 0.8 seconds) when production fluctuations are large.

[0044] When the liquid level or concentration value at any sampling time point exceeds the upper or lower boundary trajectory, it is determined that the actual change trajectory does not continuously run within the dynamic envelope range. At this time, the system generates a trajectory abnormality signal, which is stored in data register D9500. A value of 0 indicates a trajectory abnormality. At the same time, the execution counter is reset, and the time point and type of abnormality are recorded. Abnormality types include two types: liquid level abnormality and concentration abnormality. Liquid level abnormality is further divided into liquid level too high abnormality and liquid level too low abnormality, and concentration abnormality is divided into concentration too high abnormality and concentration too low abnormality. Abnormal information is stored in data registers D9600-D9699, including abnormal timestamp, abnormal parameter type, and abnormal deviation value. The abnormal timestamp records the specific time of the abnormality, accurate to milliseconds. The abnormal parameter type identifies whether it is a liquid level abnormality or a concentration abnormality. The abnormal deviation value records the difference between the abnormal value and the boundary value.

[0045] A data validity check mechanism is set up during the monitoring process. When abnormal values ​​are detected in the monitoring data, such as liquid level exceeding the range (less than 0 meters or greater than 5 meters) or concentration values ​​changing abruptly (change exceeding 10 g / L between adjacent sampling points), the system automatically marks the sampling point as invalid data and excludes it from the envelope judgment. At the same time, valid data from the previous moment is used to replace it to ensure the continuity of the judgment process. The monitoring results are updated every 100 milliseconds to keep synchronized with the data acquisition and ensure that the changes in the tank liquid state are reflected in real time. All judgment logic is implemented through the cyclic scanning program of the programmable logic controller, with a scanning period set to 10 milliseconds to ensure the real-time response performance of the system. The system also has a watchdog timer that automatically resets the system when the monitoring program malfunctions, ensuring the reliability of the monitoring function.

[0046] The intelligent closed-loop control system for the passivation solution in the hot-dip galvanizing process executes coordinated commands or generates early warning signals in the following ways: When it is determined that the actual change trajectory continues to operate within the dynamic envelope range, the coordinated command is converted into a start / stop control signal for the replenishment pump and a flow regulation signal for the replenishment pump. The coordinated command originates from the calculation results of the data fusion module and includes two parts: replenishment volume and replenishment volume. The replenishment volume is stored in data register D7000, and the replenishment volume is stored in data register D7001. The start / stop control signal for the replenishment pump is generated by calculating the running time based on the replenishment volume and the rated flow rate of the replenishment pump. For example, if the rated flow rate of the replenishment pump is 10 liters / minute and the replenishment volume is 200 liters, the running time is calculated... The runtime is 200 / 10 = 20 minutes. This runtime is then converted into a pulse signal and sent to the replenishment pump controller via a programmable logic controller (PLC). The output signal is a 24V DC pulse signal, with the pulse width corresponding to the runtime. The flow rate adjustment signal generation process for the replenishment pump involves calculating the flow rate setpoint based on the replenishment amount and the required replenishment time. For example, if the replenishment is required to be completed within 30 minutes and the replenishment amount is 25 kg, the flow rate setpoint is 25 / 30 = 0.83 kg / minute. Then, a 4-20mA analog signal is generated using a proportional-integral-derivative (PID) control algorithm and sent to the replenishment pump's frequency converter. The proportional coefficient is set to 0.8, the integral time to 60 seconds, and the derivative time to 10 seconds.

[0047] When the actual trajectory is determined not to continuously operate within the dynamic envelope range, the transmission of the coordinated command to the infusion pump and the medication pump is blocked. The blocking mechanism is implemented through both hardware and software. The hardware method uses a relay isolation circuit with a 24V DC coil voltage and a 10A / 250V AC contact capacity. When a blocking signal is received, the relay contacts open, cutting off the control signal path. The software method sets an enable flag in the programmable logic controller (PLC) program. This flag is stored in the data register D9500. When the trajectory is abnormal, this flag is set to 0, and all control signals are blocked. Before execution, the control signal output function checks this flag bit; if it is 0, the output operation is not performed. Simultaneously, it generates a system anomaly warning signal containing an anomaly timestamp and anomaly parameter type. The anomaly timestamp comes from the system's real-time clock module, using a DS1302 chip, accurate to the millisecond level. The anomaly parameter type comes from the anomaly information recorded by the envelope monitoring module, stored in data registers D9600-D9699, including liquid level anomalies or concentration anomalies, and specific anomaly types such as too high liquid level, too low liquid level, too high concentration, and too low concentration. The warning signal uses the standard Modbus TCP alarm protocol format, including fields such as alarm code, alarm level, alarm time, and alarm description, and is sent to the host computer monitoring system via industrial Ethernet. Simultaneously, it triggers on-site audible and visual alarm devices; the audible alarm uses a 95dB buzzer, and the visual alarm uses a red LED indicator.

[0048] The system also features a command execution confirmation mechanism. After outputting a control signal, the system verifies the command execution status through feedback signals. For example, after the replenishment pump control signal is output, the system monitors the pump's operating status feedback signal. If no feedback signal is received within a set time of 5 seconds, the command execution is deemed to have failed, and the control signal is resent or a device fault alarm is generated. After the medication pump flow regulation signal is output, the system monitors the actual flow rate using an electromagnetic flow meter with an accuracy class of 0.5 and an output signal of 4–20 mA. If the deviation between the actual flow rate and the set flow rate consistently exceeds the allowable range of ±10%, the output signal is adjusted or a regulation anomaly alarm is generated. All command execution records and alarm information are stored in a historical database in SQLite database format, including command sending time, command content, execution result, alarm information, etc., with a storage time of no less than 30 days for subsequent query and analysis. The system also features a manual / automatic switching function, which is achieved through a selection switch. In automatic mode, the system executes control commands completely automatically, while in manual mode, operators can intervene in command execution through a human-machine interface, ensuring reliable operation under various working conditions.

[0049] It is worth noting that the intelligent closed-loop control system for the passivation solution in this hot-dip galvanizing process is equipped with a dual-float redundant liquid level sensing system as a safety protection device. The two float switches are installed at different heights on the side wall of the passivation tank. The high-level float corresponds to the upper limit of the liquid level, and the low-level float corresponds to the lower limit of the liquid level. They adopt a normally open reed switch structure with a contact capacity of 5A / 250V AC. When the liquid level drops to the low-level float position, the float switch contacts close, and the magnetic drive pump is started through the intermediate relay to replenish the liquid. The magnetic drive pump has a rated flow rate of 15m³ / h and a head of 20 meters. When the liquid level rises to the high-level float position, the float switch contacts open, and the programmable logic controller collects the signal and immediately outputs a stop command to achieve hard-wired interlocked shutdown. The system uses a turbine flow meter to calibrate the replenishment volume online. The flow meter has an accuracy class of 0.5 and is equipped with a pipeline damper to eliminate fluid pulsation. The replenishment volume error is controlled within ±2% through a PID adjustment algorithm to ensure that the replenishment accuracy meets the process requirements.

[0050] The calculations involved in the embodiments are all dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0051] It should be noted that this invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0052] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wireless or wired transmission; wired transmission methods include optical fiber, twisted pair, coaxial cable, etc.; wireless transmission includes infrared, microwave, etc. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center containing one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0053] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0054] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0055] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0056] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0057] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0058] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0059] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An intelligent closed-loop control system for passivation solution in hot-dip galvanizing process, characterized in that, include: The signal monitoring module is used to identify the production cycle and calculate the estimated compensation value of the passivation fluid consumption based on the preset fixed compensation amount. The entropy weight generation module is used to obtain a symbol sequence based on the symbolization of the production cycle, calculate the information entropy value of the symbol sequence, and generate weight allocation coefficients through a preset mapping relationship. The data fusion module is used to collect monitoring data on the liquid level and concentration in the passivation tank, and to generate coordinated commands for controlling the replenishment pump and the replenishment pump based on the weight allocation coefficient and the estimated compensation value fused with the monitoring data. The trajectory prediction module is used to simulate and calculate the virtual change trajectory of the key parameters of the tank liquid after expected execution based on the collaborative instructions, and generate a dynamic envelope formed by its upper and lower boundaries based on the virtual change trajectory. The envelope monitoring module is used to monitor whether the real change trajectory, which is composed of real-time collected monitoring data, continues to run within the range of the dynamic envelope line. The execution warning module is used to execute coordinated commands to drive the replenishment pump and the replenishment pump to perform quantitative replenishment operations when the actual change trajectory continues to run within the range of the dynamic envelope. Otherwise, the execution of coordinated commands is prohibited and a system abnormality warning signal is generated.

2. The intelligent closed-loop control system for passivation solution in hot-dip galvanizing process according to claim 1, characterized in that, Identify the production cycle time and calculate the estimated compensation value for passivation fluid consumption based on a preset fixed compensation amount, including: Real-time monitoring of the rising and falling edge signals of the lifting hook of the elevator in the hot-dip galvanizing production line; The falling edge signal of the lifting hook of the elevator is used as the starting point of the production cycle, and the rising edge signal is used as the ending point of the production cycle. The specification markings of the current workpiece basket are read by an RFID tag installed on the workpiece basket; The corresponding fixed compensation amount is called based on the current workpiece basket specification markings; The fixed compensation amount is used as the estimated compensation value.

3. The intelligent closed-loop control system for passivation solution in hot-dip galvanizing process according to claim 1, characterized in that, Based on the symbolization of the production cycle, a symbol sequence is obtained. The information entropy value of the symbol sequence is calculated, and weight allocation coefficients are generated through a preset mapping relationship, including: Record the duration of each production cycle in multiple consecutive production cycle cycles to form a duration sequence; The duration of each production cycle in the time sequence is compared with a preset duration threshold. When the duration of the production cycle is greater than or equal to a preset duration threshold, it is marked as the first symbol; when the duration of the production cycle is less than the preset duration threshold, it is marked as the second symbol, forming a symbol sequence. Calculate the information entropy value based on the probability of different symbols appearing in the symbol sequence; Establish a pre-defined mapping table between information entropy values ​​and weight allocation coefficients; Match the calculated information entropy value with the entropy value interval in the mapping table, find the weight allocation coefficient corresponding to the entropy value interval in the mapping table, and output the matched weight allocation coefficient.

4. The intelligent closed-loop control system for passivation solution in hot-dip galvanizing process according to claim 3, characterized in that, The pre-defined mapping relationship table between information entropy value and weight allocation coefficient is established in the following way: the numerical range of information entropy value is divided in advance through experiments, and each numerical range is bound to a pre-set weight allocation coefficient to form an entropy value-weight lookup table.

5. The intelligent closed-loop control system for passivation solution in hot-dip galvanizing process according to claim 1, characterized in that, Collect monitoring data on the level and concentration of the passivation tank, and generate coordinated control commands for the replenishment pump and the chemical replenishment pump based on the weighted allocation coefficient, merging the estimated compensation value and the monitoring data. These commands include: Real-time acquisition of liquid level monitoring data collected by the liquid level sensor and concentration monitoring data collected by the concentration sensor; Calculate the level deviation between the liquid level monitoring data and the preset liquid level setting, and the concentration deviation between the concentration monitoring data and the preset concentration setting, respectively. The feedforward compensation amount is obtained by weighting the estimated compensation value using weight allocation coefficients. The feedback compensation amount is obtained by subtracting the weighting factor from 1 and weighting the liquid level deviation value and the concentration deviation value. The feedforward compensation and the feedback compensation are added together to obtain the final compensation amount, which is then converted into a coordinated instruction for the fluid replenishment pump and the drug replenishment pump.

6. The intelligent closed-loop control system for passivation solution in hot-dip galvanizing process according to claim 1, characterized in that, Based on the collaborative instructions, the virtual change trajectory of key parameters of the tank solution after expected execution is simulated and calculated. A dynamic envelope formed by the upper and lower boundaries of this virtual change trajectory is generated, including: Based on the replenishment amount and replenishment rate contained in the collaborative instruction, the virtual trajectory of the expected change in liquid level and concentration after execution is calculated through the tank liquid dynamic model. Based on the dynamic modal stability index obtained by symbolic dynamic modal analysis of the continuous production cycle duration sequence, the positive and negative tolerance values ​​are dynamically adjusted. The upper boundary trajectory is formed by superimposing dynamically adjusted positive tolerance values ​​at each sampling point of the virtual change trajectory; The lower boundary trajectory is formed by superimposing dynamically adjusted negative tolerance values ​​at each sampling point of the virtual change trajectory; The dynamic envelope is formed by the upper boundary trajectory and the lower boundary trajectory.

7. The intelligent closed-loop control system for passivation solution in hot-dip galvanizing process according to claim 6, characterized in that, The virtual change trajectory of liquid level and concentration after expected execution is obtained by calculating the liquid dynamic model. This means that, based on the replenishment amount in the collaborative instruction and the principles of mass conservation and concentration balance, the expected change trajectory of liquid volume and solute concentration after the replenishment operation is completed are calculated respectively.

8. The intelligent closed-loop control system for passivation solution in hot-dip galvanizing process according to claim 6, characterized in that, The dynamic modal stability index based on symbolic dynamic modal analysis refers to the dynamic modal stability index obtained by analyzing the state transition probability matrix of a symbol sequence, calculating the eigenvalue magnitude of its dominant dynamic mode, and using the eigenvalue magnitude as a dynamic modal stability index to quantify the stability of the production cycle. The closer the dynamic modal stability index is to 1, the smaller the tolerance range of the dynamic envelope will be; conversely, the smaller the index is to 1, the larger the tolerance range of the dynamic envelope will be.

9. The intelligent closed-loop control system for passivation solution in hot-dip galvanizing process according to claim 1, characterized in that, Monitoring whether the actual change trajectory, composed of real-time collected monitoring data, continuously operates within the dynamic envelope range includes: The real-time collected liquid level monitoring data and concentration monitoring data are combined in chronological order to form a true change trajectory; At multiple consecutive sampling time points, determine whether the liquid level and concentration values ​​in the actual change trajectory are simultaneously located between the upper and lower boundary trajectories of the dynamic envelope; When the liquid level and concentration values ​​at multiple consecutive sampling time points are simultaneously between the upper and lower boundary trajectories, it is determined that the actual change trajectory continues to run within the dynamic envelope range. When the liquid level or concentration value at any sampling time point exceeds the upper or lower boundary trajectory, it is determined that the true change trajectory does not continuously run within the dynamic envelope range.

10. The intelligent closed-loop control system for passivation solution in hot-dip galvanizing process according to claim 1, characterized in that, When the actual trajectory of change continues to operate within the range of the dynamic envelope, the coordinated command is executed to drive the replenishment pump and the replenishment pump to perform quantitative replenishment operations; otherwise, the coordinated command is prohibited from being executed and a system abnormality warning signal is generated, including: When it is determined that the actual change trajectory continues to run within the range of the dynamic envelope, the coordinated command is converted into the start / stop control signal of the replenishment pump and the flow regulation signal of the replenishment pump. When it is determined that the actual change trajectory does not continuously run within the range of the dynamic envelope, the transmission of the coordination command to the infusion pump and the medication pump is blocked; at the same time, a system anomaly warning signal containing abnormal timestamps and abnormal parameter types is generated.

Images