Intelligent control system and method for dehydration filter press based on styrene butadiene rubber processing

Abstract

The invention relates to the technical field of intelligent control of rubber processing equipment, in particular to an intelligent control system and method for a dehydration filter press based on styrene butadiene rubber processing, and aims to solve the problems that the rheological property of the styrene butadiene rubber is obviously influenced by brand, temperature, pressure and batch difference in the filter pressing process at present; an online dynamic modeling mechanism fusing multi-source information is lacked, and accurate and real-time characterization of viscosity and key rheological parameters is difficult to realize; according to the invention, the filter pressing rheological characteristic decision-making module fuses butadiene styrene rubber marks, real-time working conditions and similar historical batch data, through table look-up interpolation, Cross-WLF online fitting and transfer learning cooperation, high-precision dynamic prediction of rubber material viscosity and key rheological parameters is realized, multi-source estimation is adaptively fused under steady-state and dynamic working conditions, and high-precision dynamic prediction of rubber material viscosity and key rheological parameters is realized. Long-term stable operation is ensured, and accurate, robust and interpretable rheological decision support is provided for the filter pressing process.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent control of rubber processing equipment, more particularly to a dehydration filter press intelligent control system and method based on styrene-butadiene rubber processing. BACKGROUND

[0002] In the production of styrene-butadiene rubber by wet process, dehydration and filter pressing is a key process, which directly affects the water content, quality stability and energy consumption of the product. However, the existing filter press relies on fixed parameters or manual experience for control, which is difficult to adapt to the fluctuations of wet rubber particle water content, viscoelasticity and particle size distribution, and is prone to cause insufficient dehydration or excessive squeezing, resulting in unstable product quality, filter cloth blockage and low energy efficiency. Although some equipment introduces PLC or simple feedback control, it lacks real-time sensing and dynamic optimization capability for key process states such as filter liquid flow, pressure difference and cake water content during the dehydration process of styrene-butadiene rubber, and has not formed an intelligent closed-loop control facing actual working conditions, which restricts the improvement of dehydration efficiency and intelligent level.

[0003] The patent application with publication number CN118698181A discloses a filter press control system and method, which includes a pressing plate control unit and a pulling plate control unit. The pressing plate control unit includes a pressing plate compression start button, a first pressure detection element, a pressing plate compression electromagnetic valve and an oil pump motor, which are connected by hard wiring for controlling the pressing plate to perform compression action. The pulling plate control unit includes a pulling plate start button, a pulling plate start relay and a time relay, which are connected by hard wiring for controlling the pulling plate to start or stop. However, the above-mentioned reference patent adopts hard-wired method to construct the pressing plate and pulling plate control units, and adjusts the self-holding and stopping logic of the pulling plate circuit through the time relay, which realizes automatic control, facilitates fault intuitive positioning and rapid processing, and improves system reliability. However, the rheological properties of current styrene-butadiene rubber during filter pressing are significantly affected by grade, temperature, pressure and batch differences, lack online dynamic modeling mechanism integrating multi-source information, and are difficult to accurately and real-timely represent viscosity and key rheological parameters. Moreover, the filter pressing process control lacks dynamic decision-making capability linked with the rheological state of the rubber, and cannot adaptively adjust the pressure and filtration period according to the actual dehydration process, which easily leads to water content control deviation or energy waste. At the same time, the equipment execution layer lacks closed-loop feedback and self-correction mechanism based on multi-dimensional working conditions, which is difficult to guarantee the accuracy of control instructions and the stability of long-term operation of the system when facing working condition disturbance, equipment performance degradation or abnormal operation.

[0004] Therefore, the present application proposes a dehydration filter press intelligent control system and method based on styrene-butadiene rubber processing to solve the above problems. SUMMARY

[0005] The application aims to provide a dehydration filter press intelligent control system and method based on styrene-butadiene rubber processing, solve the problem that the rheological properties of current styrene-butadiene rubber are significantly affected by brand, temperature, pressure and batch difference in the filter pressing process, lack online dynamic modeling mechanism integrating multi-source information, it is difficult to realize accurate and real-time characterization of viscosity and key rheological parameters, and the filter pressing process control lacks dynamic decision-making ability linked with the rheological state of the rubber compound, cannot adaptively adjust the pressure and filtration cycle according to the actual dehydration process, and is prone to cause water content control deviation or energy waste; at the same time, the equipment execution layer lacks a closed-loop feedback and self-correction mechanism based on multi-dimensional working conditions, and it is difficult to ensure the accuracy of the control command and the stability of the long-term operation of the system when facing working condition disturbance, equipment performance degradation or abnormal operation.

[0006] The application achieves the above-mentioned purpose through the following technical solutions. A dehydration filter press intelligent control system based on styrene-butadiene rubber processing is integrated into a styrene-butadiene rubber intelligent production platform and comprises: A data acquisition and preprocessing module is used to acquire filter pressing working condition data in the dehydration filter pressing process of the styrene-butadiene rubber, pre-process the acquired filter pressing working condition data, and generate a pre-processed filter pressing working condition data sequence. A filter pressing fault early warning module is used to identify potential fault modes of key components of the dehydration filter press based on real-time and historical data in the pre-processed filter pressing working condition data sequence and issue a warning. A filter pressing rheological property decision module is used to predict the viscosity and key rheological parameters of the rubber compound in the filter pressing process based on the brand of the styrene-butadiene rubber, real-time data in the pre-processed filter pressing working condition data sequence and similar historical batch data, and adopt a method combining rheological property table lookup interpolation, Cross-WLF equation online parameter fitting and transfer learning. A filter pressing dehydration quality regulation module is used to dynamically adjust the pressure and filtration cycle of the filter pressing process based on target moisture standards, rheological property prediction results of the rubber compound and pre-processed real-time working condition data, integrate particle swarm optimization and adaptive mechanisms. A filter pressing execution strategy regulation module is used to adopt a reverse design method to deduce the action sequence of the equipment based on the target pressure and target filtration cycle generated by dynamic optimization, integrate the pre-processed real-time working condition data to construct a multi-dimensional closed loop, and dynamically generate control instructions of the hydraulic system and driving mechanism.

[0007] As a preferred embodiment of the application, the process in which the filter pressing fault early warning module identifies potential fault modes of key components of the dehydration filter press based on real-time and historical data in the pre-processed filter pressing working condition data sequence and issues a warning comprises: Obtain the working pressure of the main cylinder in the current filter press cycle and the running time of the current filter press stage. Determine the start and end times of the pressure holding stage based on the current running time of the filter press stage. Calculate the difference between the starting and ending values ​​of the working pressure of the main cylinder in the pressure holding stage as the pressure decay value of the pressure holding stage. Determine if the pressure decay value of the pressure holding stage is not less than the absolute threshold of pressure decay, the pressure decay deviation is not less than the pressure decay deviation threshold, and the clamping time deviation is not greater than the upper limit of the clamping time deviation. If these conditions are met, output the current filter press cycle number, the name of the key component (main hydraulic cylinder), and the fault mode name (hydraulic system internal leakage). Obtain the pressure plate displacement, main motor current, and current filter pressing stage running time of the current filter pressing cycle. Sum the main motor current value of the unloading stage to obtain the unloading stage current integral value. Determine if the unloading stage current integral value is not less than the unloading current integral absolute threshold and the unloading current integral deviation is not less than the unloading current integral deviation threshold. If these conditions are met, output the current filter pressing cycle number, the key component name (pulling plate mechanism), and the fault mode name (pulling plate mechanism jamming).

[0008] In a preferred embodiment of the present invention, the filter press rheological property decision module, based on the styrene-butadiene rubber grade, real-time data from the pre-processed filter press operating condition data sequence, and similar historical batch data, employs a collaborative mechanism of rheological property lookup interpolation and Cross-WLF online fitting, including: Obtain the current styrene-butadiene rubber grade, use the reference viscosity at the corresponding reference temperature and the temperature-pressure-viscosity three-dimensional lookup table to obtain the rubber inlet temperature, main cylinder working pressure, and online viscosity sensor measured viscosity time series for the current filter press cycle. For each time point, obtain the rubber inlet temperature and main cylinder working pressure, locate the smallest rectangular area containing the corresponding temperature and pressure points in the lookup table, calculate the temperature change rate and pressure change rate. If both are less than their respective preset thresholds, perform bilinear interpolation; otherwise, fit a local quadratic surface based on the four corner points, and calculate the function value of the surface at the corresponding temperature and pressure points as the initial viscosity value. When entering the parameter update stage, the Cross-WLF rheological parameters of the same grade and those that have completed the pressure filtration task are obtained from the historical batch library. The sample buffer and counter are initialized. For each time point, when the counter equals the number of samples required for the preset parameter update, the predicted viscosity of the model is divided into the same interval to obtain the predicted probability distribution. The Williams-Landel-Ferry parameters are optimized through the objective function, and the updated Williams-Landel-Ferr parameters are output. Using the latest Williams-Landel-Ferry parameters, combined with the rubber inlet temperature at each time point, the Cross-WLF model is substituted to calculate and output the final viscosity estimate.

[0009] In a preferred embodiment of the present invention, the process of the filter press rheological characteristic decision module employing a migration calibration mechanism based on historical batches includes: Obtain the styrene-butadiene rubber grade and process path identifier of the current filtration cycle, obtain the online viscosity sensor measured viscosity time series and the number of accumulated valid samples for the current filtration cycle. If the number of valid samples is less than the preset minimum number of samples, perform migration calibration; otherwise, terminate the process. Select the batch with the same styrene-butadiene rubber grade, consistent process path and completed filtration task from the historical batch library as the migration source, and obtain the final converged Cross-WLF parameters of the migration source batch. Obtain the operating condition similarity corresponding to the migration source, use the Cross-WLF parameters of the migration source as the current initial parameters, optimize the current Cross-WLF parameters through the corresponding objective function, and output the update result.

[0010] In a preferred embodiment of the present invention, the pressure filtration and dewatering quality control module dynamically adjusts the pressure and filtration cycle of the pressure filtration process based on the target moisture standard, the predicted rheological properties of the rubber compound, and pre-processed real-time operating data, integrating particle swarm optimization and adaptive mechanisms. The process includes: Obtain the target moisture standard, initial flow resistance, and dehydration rate decay coefficient. Obtain the main cylinder working pressure, filtrate instantaneous flow rate, pressure plate displacement, current filtration stage running time, and previous cycle filtrate instantaneous flow rate. Calculate the current flow rate change as the difference between the current and previous cycle filtrate instantaneous flow rates. Look up the target dehydration trajectory table to obtain the target flow rate change corresponding to the current filtration stage. Based on the comparison between the current flow rate change and the target flow rate change, if it is less, update the main cylinder working pressure and pressure holding stage running time according to the pressure increase and time extension rules. If it is greater, update according to the pressure decrease and time shortening rules. If they are equal, keep the main cylinder working pressure and the running time of each stage unchanged. The system acquires the main motor current and hydraulic system power. If either the main motor current or the hydraulic system power exceeds the limit, the main cylinder working pressure and the pressure holding stage running time are updated according to the safety pressure reduction and safety delay rules. Particle swarm optimization is run to generate candidate parameter combinations and merge them with the adaptive adjustment results to form a candidate scheme. If the flow rate change deviates from the target by more than the allowable deviation for two consecutive control cycles, the current plan is terminated, particle swarm optimization is reinitialized, new parameters are generated, and execution is switched immediately. The main cylinder working pressure, the pressing stage running time, and the pressure holding stage running time are output.

[0011] In a preferred embodiment of the present invention, the process by which the pressure filtration execution strategy control module derives the equipment action sequence based on dynamically optimized target pressure and target filtration cycle using a reverse design method includes: Obtain the target pressure and target filtration cycle, obtain the current time and the start and end times of each filtration stage, determine the current filtration stage based on the current time and stage start and end times, obtain the final pressure and final time of the current filtration stage, if the target pressure is greater than the maximum allowable working pressure of the equipment, then the target pressure is equal to the maximum allowable working pressure of the equipment, obtain the control cycle number, if the control cycle number is not the first cycle and the target pressure is less than the minimum effective filtration pressure, then the target pressure is equal to the target pressure of the previous control cycle, if the control cycle number is the first cycle and the target pressure is less than the minimum effective filtration pressure, then the target pressure is equal to the minimum effective filtration pressure; The system acquires chamber pressure, filtrate flow rate, and hydraulic oil temperature; it also acquires the duration of the current filtration stage, speed penalty parameters, and acceleration penalty parameters. Using the target pressure and target filtration cycle as boundary conditions, it reverse-engineers the equipment action sequence from the final pressure and final time. The equipment action sequence is generated by solving an optimization problem and is then output.

[0012] In a preferred embodiment of the present invention, the process by which the pressure filtration execution strategy control module integrates preprocessed real-time operating data to construct a multi-dimensional closed-loop correction includes: Obtain the current chamber pressure and the expected pressure. If the absolute deviation between the two is greater than the pressure tracking deviation limit, increase the feed valve opening by the feed valve opening increment. Obtain the filtrate flow rate for the current and the two most recent historical control cycles. If the filtrate flow rate decreases by more than the filtrate flow rate decrease limit for two consecutive cycles, and the main frequency energy of the feed pressure fluctuation is greater than the energy rise limit, increase the duration of the current stage by the stage extension time. If the current pressure filtration stage is equal to the pressure boosting stage, multiply the feed valve opening by the pressure boosting opening attenuation factor. Get the hydraulic oil temperature and its upper limit. If the hydraulic oil temperature is greater than the upper limit, get the hydraulic oil viscosity and its upper limit. If the hydraulic oil viscosity is greater than the upper limit, multiply the target pressure by the target pressure attenuation factor, set the auxiliary cooling pump operation flag to enabled, and output the feed valve opening, target pressure, auxiliary cooling pump operation flag, and current stage duration.

[0013] A smart control method for a dewatering filter press based on styrene-butadiene rubber processing includes the following steps: Step 1: Collect filter press operation data during the dehydration and filter press process of styrene-butadiene rubber, preprocess the collected filter press operation data, and generate a preprocessed filter press operation data sequence. Step 2: Based on the real-time and historical data in the pre-processed filter press operating condition data sequence, identify potential failure modes of key components of the dewatering filter press and issue early warnings; Step 3: Based on the real-time data and similar historical batch data in the pre-processed filter press condition data sequence of styrene-butadiene rubber grade, the viscosity of the rubber compound and its key rheological parameters during the filter press process are predicted by a combination of rheological property lookup table interpolation, online parameter fitting of the Cross-WLF equation and transfer learning. Step 4: Based on the target moisture standard, the predicted rheological properties of the rubber compound, and the pre-processed real-time operating data, the pressure and filtration cycle of the filter press process are dynamically adjusted by integrating particle swarm optimization and adaptive mechanisms. Step 5: Based on the target pressure and target filtration cycle generated by dynamic optimization, the reverse design method is used to derive the equipment action sequence, and the pre-processed real-time working condition data is integrated to construct a multi-dimensional closed loop, dynamically generating control commands for the hydraulic system and drive mechanism.

[0014] Compared with the prior art, the advantages of this invention are: (1) In this invention, the styrene-butadiene rubber grade, real-time working conditions and similar historical batch data are integrated by the pressure filter rheological property decision module. Through table lookup interpolation, Cross-WLF online fitting and transfer learning collaboration, high-precision dynamic prediction of rubber viscosity and key rheological parameters is achieved. Under steady-state and dynamic working conditions, multi-source estimation is adaptively integrated. Combined with distribution matching optimization and working condition similarity-driven transfer calibration, the generalization ability and convergence speed of the model are improved. It also has a misalignment self-check and automatic refit mechanism to ensure long-term stable operation and provide accurate, robust and interpretable rheological decision support for the pressure filter process. (2) In this invention, the pressure filtration execution strategy control module generates the equipment action sequence accurately based on the dynamic optimization target by using the reverse design method, and integrates multi-dimensional real-time working condition data such as chamber pressure, filtrate flow rate and hydraulic oil temperature to build a closed-loop correction mechanism, thereby realizing high-precision adaptive control of the feed valve, pressing pump and drive mechanism. While ensuring stable process execution, it also has safety protection, deviation response and energy efficiency optimization capabilities, significantly improving the system robustness, control accuracy and operation safety. Attached Figure Description

[0015] Figure 1 This is a system block diagram of Embodiment 1 and Embodiment 2 of the present invention; Figure 2 This is a flowchart of the steps in the intelligent control method for the dewatering filter press of the present invention. Detailed Implementation

[0016] 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. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0017] Example 1: As Figure 1 As shown, the present invention proposes an intelligent control system for a dewatering filter press based on styrene-butadiene rubber processing, integrated into a styrene-butadiene rubber intelligent production platform, comprising: The data acquisition and preprocessing module is used to collect the filtration conditions data during the dehydration and filtration process of styrene-butadiene rubber, including the working pressure of the main cylinder, the displacement of the pressure plate, the inlet temperature of the rubber material, the instantaneous flow rate of the filtrate, the running time of the current filtration stage, the main motor current and the power of the hydraulic system. The module preprocesses the collected filtration conditions data, including data cleaning, digital filtering, time synchronization alignment and unit range calibration, and generates a preprocessed filtration conditions data sequence. By uniformly collecting key pressure filter operating parameters and implementing data cleaning, digital filtering, time synchronization, and range calibration, sensor noise, timing offset, and dimensional differences are effectively eliminated, improving data consistency and timeliness. This provides a highly reliable, highly synchronized, and standardized input basis for subsequent control strategies, enhancing system robustness and control accuracy.

[0018] The filter press fault early warning module is used to identify potential fault modes of key components of the dewatering filter press and issue early warnings based on real-time and historical data in the pre-processed filter press operating condition data sequence. The filter press fault early warning module identifies potential fault modes of key components of the dewatering filter press and issues early warnings based on real-time and historical data from the pre-processed filter press operating condition data sequence. The process includes: Obtain the working pressure of the main cylinder and the running time of the current filter press cycle. Determine the start and end times of the pressure holding stage based on the current running time of the filter press stage. Calculate the difference between the starting and ending values ​​of the working pressure of the main cylinder during the pressure holding stage as the pressure attenuation value during the pressure holding stage. Calculate the reference pressure attenuation value based on the pressure attenuation values ​​of the pressure holding stage in the historical normal filter press cycle set. The difference between the current value and the reference value is the pressure attenuation deviation. Determine the duration of the pressing stage. The difference between the current value and the average duration of the pressing stage in the historical normal cycle is the pressing time deviation. Determine that the pressure attenuation value of the pressure holding stage is not less than the absolute threshold of pressure attenuation, the pressure attenuation deviation is not less than the pressure attenuation deviation threshold, and the pressing time deviation is not greater than the upper limit of the pressing time deviation. If these conditions are met, output the current filter press cycle number, the name of the key component (main hydraulic cylinder), and the fault mode name (hydraulic system internal leakage). Obtain the pressure plate displacement, main motor current, and current filtration stage running time of the current filtration cycle. Determine the unloading stage time interval based on the current filtration stage running time. Check whether the pressure plate displacement in the unloading stage has a continuous number of time points that are not less than the minimum number of displacement stagnation points and the displacement values ​​are equal. Sum the main motor current values ​​in the unloading stage to obtain the unloading stage current integral value. Calculate the benchmark unloading current integral value based on the unloading stage current integral value of the historical normal filtration cycle set. The difference between the current value and the benchmark value is the unloading current integral deviation. Determine whether the unloading stage current integral value is not less than the unloading current integral absolute threshold and the unloading current integral deviation is not less than the unloading current integral deviation threshold. If these conditions are met, output the current filtration cycle number, the key component name (pulling plate mechanism), and the fault mode name (pulling plate mechanism jamming). Obtain the master cylinder working pressure and current filtration stage running time for the current and several previous filtration cycles (the number is determined by the degradation observation period). For each cycle, determine the duration of the pressing stage based on the current filtration stage running time. The difference between this duration and the average duration of the pressing stage in historical normal cycles is the pressing time deviation. Extract the master cylinder working pressure value at the end of the pressing stage in each cycle. Calculate the benchmark pressing end pressure value based on the pressing stage end pressure value of historical normal cycles. Determine that the pressing time deviation of all cycles is not less than the lower limit of the pressing time deviation, and that the pressing stage end pressure value of all cycles is not greater than the difference between the benchmark pressing end pressure value and the pressing end pressure decay threshold. If these conditions are met, output the current filtration cycle number, the name of the key component (filter plate sealing assembly), and the fault mode name (filter plate sealing ring aging). By integrating real-time and historical operating data through the filter press fault early warning module, early faults such as internal leakage in the main hydraulic cylinder, jamming of the plate pulling mechanism, and aging of the filter plate seal ring can be accurately identified. This enables targeted early warning and predictive maintenance of key components, effectively improving equipment reliability, extending service life, reducing operation and maintenance costs, and ensuring production safety and continuity.

[0019] The filter press rheological property decision module is used to predict the viscosity and key rheological parameters of the rubber compound during the filter press process based on real-time data from the pre-processed filter press condition data sequence and similar historical batch data of styrene-butadiene rubber grade, using a combination of rheological property lookup table interpolation, online parameter fitting of the Cross-WLF equation and transfer learning. The decision-making module for filter press rheological properties, based on styrene-butadiene rubber grade, real-time data from the pre-processed filter press operating condition data sequence, and similar historical batch data, employs a collaborative mechanism of rheological property lookup interpolation and Cross-WLF online fitting, including the following steps: Obtain the current styrene-butadiene rubber grade, use the reference viscosity at the corresponding reference temperature and the temperature-pressure-viscosity three-dimensional lookup table. The lookup table is a regular grid. The input is temperature and pressure, and the output is viscosity. Obtain the rubber material inlet temperature, main cylinder working pressure, and online viscosity sensor measured viscosity time series for the current filter press cycle. For each time point, obtain the rubber material inlet temperature and main cylinder working pressure. Locate the smallest rectangular area containing the corresponding temperature and pressure points in the lookup table. The area consists of four corner points, each corner point corresponding to a known viscosity value. Calculate the rate of temperature change and the rate of pressure change. If both are less than their respective preset thresholds, perform bilinear interpolation: first, perform linear interpolation along the temperature direction for the two pressure edges, and then perform linear interpolation along the pressure direction to obtain the initial viscosity value. Otherwise, fit a local quadratic surface based on the four corner points and calculate the function value of the surface at the corresponding temperature and pressure points as the initial viscosity value. Upon entering the parameter update phase, the rheological parameters of Cross-WLF with the same grade and completed filtration tasks are retrieved from the historical batch database. The sample cache and counter are initialized. For each time point, the measured viscosity, inlet temperature of the compound, and initial viscosity value are stored in the cache, and the counter is incremented. When the counter equals the number of samples required for the preset parameter update, the following operations are performed: the measured viscosity is divided into fixed value intervals, the frequencies are statistically analyzed, and normalized to obtain the measured probability distribution. The model-predicted viscosity is divided into the same intervals to obtain the predicted probability distribution. The Williams-Landel-Ferry parameters are optimized using the objective function. ,in Indicated by Williams-Landel-Ferry parameters and The total loss value of the independent variable. This represents the viscosity of the rubber compound measured by an online viscosity sensor at the i-th time point. Indicates the grade of styrene-butadiene rubber at the reference temperature. The reference viscosity is as follows. This represents the actual melt temperature inside the filter press chamber at the i-th time point. Indicates the preset reference temperature. This represents the first parameter to be optimized in the Williams-Landel-Ferry equation, used to describe the sensitivity of viscosity to temperature changes. This represents the second parameter to be optimized in the Williams-Landel-Ferry equation, used to describe the shift characteristics of the reference temperature. This indicates the number of accumulated samples required to update the preset parameters. This represents the KL divergence, used to measure the difference in the overall distribution pattern between the measured viscosity sequence and the model-predicted viscosity sequence. This represents a probability distribution constructed from a fixed number of continuous measured viscosity values ​​divided into fixed intervals. This represents the probability distribution constructed by dividing the viscosity values ​​predicted by the model at the same time point into the same intervals. Indicates the preset distribution matching weight coefficients; Output the updated Williams-Landel-Ferry parameters, clear the cache, reset the counter, use the latest Williams-Landel-Ferry parameters, combine the rubber inlet temperature at each time point, substitute into the Cross-WLF model to calculate and output the final viscosity estimate. The process of the filter press rheological property decision module using a migration calibration mechanism based on historical batches includes: Obtain the styrene-butadiene rubber grade and process path identifier of the current filtration cycle, obtain the online viscosity sensor measured viscosity time series and the number of accumulated valid samples for the current filtration cycle. If the number of valid samples is less than the preset minimum number of samples, perform migration calibration; otherwise, terminate the process. Select the batch with the same styrene-butadiene rubber grade, consistent process path and completed filtration task from the historical batch library as the migration source, and obtain the final converged Cross-WLF parameters of the migration source batch. Obtain the operating condition similarity corresponding to the migration source. If the operating condition similarity is greater than or equal to the preset similarity threshold, the migration regularization coefficient takes the weight of similar operating conditions; otherwise, the migration regularization coefficient takes the weight of dissimilar operating conditions. Use the Cross-WLF parameters of the migration source as the current initial parameters and optimize them through the corresponding objective function. ,in This indicates the current model parameter vector. The total loss value of the independent variable. This represents the Cross-WLF model's prediction of the viscosity at time point i. This represents the measured viscosity at the i-th time point. This indicates the number of valid samples currently accumulated in the real-time data sequence for filter press operation. This represents the j-th adjustable parameter of the current Cross-WLF model. This represents the j-th parameter of the Cross-WLF model for the migration source batch. This represents the migration regularization coefficient. The operating condition similarity is calculated using three features: temperature curve shape, pressure rise rate, and peak duration. The current Cross-WLF parameters are optimized, and the updated results are output. The process by which the pressure filtration rheological property decision module predicts the viscosity and key rheological parameters of the compound during pressure filtration includes: The system acquires the inlet temperature of the rubber compound, the working pressure of the main cylinder, the initial viscosity value obtained by interpolation from a lookup table, and the viscosity value predicted by the Cross-WLF model. It also acquires the rate of temperature change and the rate of pressure change. If both the rate of temperature and pressure change are less than their respective preset thresholds, the interpolation result and the Cross-WLF predicted value are weighted and fused according to preset steady-state and dynamic weights to obtain the real-time viscosity. Otherwise, the real-time viscosity is obtained by weighting and fusion according to preset dynamic and steady-state weights. The output value of the Cross-WLF model at zero shear rate is acquired as the zero-shear viscosity. Finally, the viscosity and shear rate sequences for the preset medium-to-high shear rate range are acquired, and the natural logarithm of both is taken. Calculate the absolute value of its slope as the shear thinning index. Obtain the host current, die pressure, and extrusion flow rate. If the real-time viscosity decrease, the host current increase is greater than or equal to the preset current change threshold, the extrusion flow rate decrease is greater than or equal to the preset flow rate change threshold, and the duration of all three is greater than or equal to the preset duration threshold, then the Cross-WLF model is determined to be inaccurate, triggering Cross-WLF parameter refitting. Set the similarity threshold of the historical batch retrieval as the refit-specific similarity threshold. Record the event time, the signal used, the judgment result, and the processing action in the batch log, and output the real-time viscosity, zero shear viscosity, and shear thinning index. The pressure filtration rheological property decision module integrates styrene-butadiene rubber grade information, real-time operating condition data, and similar historical batch data. Through table lookup interpolation, online fitting of the Cross-WLF equation, and a transfer learning collaborative mechanism, it achieves high-precision dynamic prediction of rubber viscosity and key rheological parameters (such as zero-shear viscosity and shear thinning index). This module adaptively weights and fuses multi-source viscosity estimates under steady-state and dynamic operating conditions, and introduces distribution matching optimization based on KL divergence and a transfer calibration strategy driven by operating condition similarity to effectively improve the model's generalization ability and parameter convergence speed. At the same time, it has a misalignment self-checking and automatic refitting mechanism to ensure long-term operational stability, providing accurate, robust, and interpretable rheological decision support for pressure filtration process control.

[0020] The pressure filter dewatering quality control module is used to dynamically adjust the pressure and filtration cycle of the pressure filter process based on the target moisture standard, the predicted results of the rheological properties of the rubber compound and the pre-processed real-time working condition data, and by integrating particle swarm optimization and adaptive mechanisms. The pressure filter dewatering quality control module, based on the target moisture standard, predicted rheological properties of the colloid, and pre-processed real-time operating data, integrates particle swarm optimization and adaptive mechanisms to dynamically adjust the pressure and filtration cycle of the pressure filter process. This process includes: Obtain the target moisture standard, initial flow resistance, and dehydration rate decay coefficient. Obtain the main cylinder working pressure, filtrate instantaneous flow rate, pressure plate displacement, current filtration stage running time, and previous cycle filtrate instantaneous flow rate. Calculate the current flow rate change as the difference between the current and previous cycle filtrate instantaneous flow rates. Look up the target dehydration trajectory table to obtain the target flow rate change corresponding to the current filtration stage. Based on the comparison between the current flow rate change and the target flow rate change, if it is less, update the main cylinder working pressure and pressure holding stage running time according to the pressure increase and time extension rules. If it is greater, update according to the pressure decrease and time shortening rules. If they are equal, keep the main cylinder working pressure and the running time of each stage unchanged. The system acquires the main motor current and hydraulic system power. If either the main motor current or the hydraulic system power exceeds the limit, the main cylinder working pressure and the pressure holding stage running time are updated according to the safety pressure reduction and safety delay rules. Particle swarm optimization is run to generate candidate parameter combinations and merge them with the adaptive adjustment results to form a candidate scheme. If the flow rate change deviates from the target by more than the allowable deviation for two consecutive control cycles, the current plan is terminated, particle swarm optimization is reinitialized, new parameters are generated, and execution is immediately switched. The main cylinder working pressure, pressing stage running time, and pressure holding stage running time are output. The filter press dewatering quality control module integrates target moisture content, rheological prediction, and real-time operating conditions. Through adaptive feedback and particle swarm optimization, it dynamically adjusts pressure and filtration cycle, accurately tracks the dewatering trajectory, and automatically and safely reduces load or reconfigures control strategies when exceeding limits or exhibiting continuous deviations, achieving efficient, stable, and safe intelligent closed-loop control.

[0021] Example 2: The technical solution of this embodiment of the invention differs from that of Example 1 in that: like Figure 1 As shown, the pressure filtration execution strategy control module is used to derive the equipment action sequence based on the target pressure and target filtration cycle generated by dynamic optimization, and to construct a multi-dimensional closed loop by integrating pre-processed real-time working data, and dynamically generate control commands for the hydraulic system and drive mechanism. The process by which the pressure filtration execution strategy control module derives the equipment action sequence based on dynamically optimized target pressure and target filtration cycle using a reverse design method includes: Obtain the target pressure and target filtration cycle, obtain the current time and the start and end times of each filtration stage, determine the current filtration stage based on the current time and stage start and end times, obtain the final pressure and final time of the current filtration stage, if the target pressure is greater than the maximum allowable working pressure of the equipment, then the target pressure is equal to the maximum allowable working pressure of the equipment, obtain the control cycle number, if the control cycle number is not the first cycle and the target pressure is less than the minimum effective filtration pressure, then the target pressure is equal to the target pressure of the previous control cycle, if the control cycle number is the first cycle and the target pressure is less than the minimum effective filtration pressure, then the target pressure is equal to the minimum effective filtration pressure; The system acquires chamber pressure, filtrate flow rate, and hydraulic oil temperature; it also acquires the duration, speed penalty parameters, and acceleration penalty parameters of the current filtration stage. Using the target pressure and target filtration cycle as boundary conditions, it reverse-engineers the equipment action sequence from the final pressure and final time. The equipment action sequence consists of action commands from multiple control cycles, each command including the feed valve opening degree and the press pump operation flag. The equipment action sequence is generated by solving an optimization problem. ,in For the sequence of equipment actions, The chamber pressure and filtrate flow rate are measured after the i-th action is performed. The stage target state is defined by the target pressure and the target filtering cycle. This is the action command for the i-th control cycle, which includes the feed valve opening degree and the press pump operation flag. This is a speed penalty parameter, the value of which is defined by the control performance configuration file. This is an acceleration penalty parameter; its value is defined in the control performance configuration file. For the target filtering cycle, This represents the duration of the current pressure filtration stage, with a value defined by the process formulation, and outputs the sequence of equipment actions. The process by which the pressure filtration execution strategy control module integrates preprocessed real-time operating data to construct a multi-dimensional closed-loop correction includes: Obtain the current chamber pressure and the expected pressure. If the absolute deviation between the two is greater than the pressure tracking deviation limit, increase the feed valve opening by the feed valve opening increment. Obtain the filtrate flow rate for the current and the two most recent historical control cycles. If the filtrate flow rate decreases by more than the filtrate flow rate decrease limit for two consecutive cycles, and the main frequency energy of the feed pressure fluctuation is greater than the energy rise limit, increase the duration of the current stage by the stage extension time. If the current pressure filtration stage is equal to the pressure boosting stage, multiply the feed valve opening by the pressure boosting opening attenuation factor. Get the hydraulic oil temperature and its upper limit. If the hydraulic oil temperature is greater than the upper limit, get the hydraulic oil viscosity and its upper limit. If the hydraulic oil viscosity is greater than the upper limit, multiply the target pressure by the target pressure attenuation factor, set the auxiliary cooling pump operation flag to enabled, and set upper and lower limits for the feed valve opening, chamber pressure, and filtrate flow rate. If the limits are exceeded, set the corresponding limit values. Output the feed valve opening, target pressure, auxiliary cooling pump operation flag, and current stage duration. The process by which the pressure filter execution strategy control module dynamically generates control commands for the hydraulic system and drive mechanism includes: Obtain the corrected chamber pressure and chamber pressure upper limit. If the chamber pressure is greater than the upper limit, set it as the upper limit. Obtain the corrected target filtrate flow rate and filtrate flow rate upper limit. If the target filtrate flow rate is greater than the upper limit, set it as the upper limit. Obtain the hydraulic oil temperature and hydraulic oil temperature inactivation limit. If the hydraulic oil temperature is greater than the inactivation limit, obtain the corrected feed valve opening and keep it unchanged. At the same time, obtain the press pump operation flag and set it to stop. Otherwise, obtain the corrected feed valve opening and press pump operation flag. Obtain the corrected target speed. Generate and output feed valve control commands, press pump control commands, and servo motor control commands. The pressure filtration execution strategy control module is based on dynamic optimization objectives and uses reverse design to accurately derive the equipment action sequence. It also integrates multi-dimensional real-time operating data such as chamber pressure, filtrate flow rate, and hydraulic oil temperature to construct a closed-loop correction mechanism, achieving high-precision and adaptive control of the feed valve, pressing pump, and drive mechanism. This module has safety protection (such as overpressure limiting and high-temperature load reduction), dynamic response (deviation trigger parameter adjustment), and energy efficiency optimization capabilities. While ensuring stable process execution, it significantly improves system robustness, control accuracy, and operational safety.

[0022] Example 3: The technical solution of this embodiment of the invention differs from that of Example 1 and Example 2 in that: like Figure 2 As shown, an intelligent control method for a dewatering filter press based on styrene-butadiene rubber processing includes the following steps: Step 1: Collect filter press operation data during the dehydration and filter press process of styrene-butadiene rubber, preprocess the collected filter press operation data, and generate a preprocessed filter press operation data sequence. Step 2: Based on the real-time and historical data in the pre-processed filter press operating condition data sequence, identify potential failure modes of key components of the dewatering filter press and issue early warnings; Step 3: Based on the real-time data and similar historical batch data in the pre-processed filter press condition data sequence of styrene-butadiene rubber grade, the viscosity of the rubber compound and its key rheological parameters during the filter press process are predicted by a combination of rheological property lookup table interpolation, online parameter fitting of the Cross-WLF equation and transfer learning. Step 4: Based on the target moisture standard, the predicted rheological properties of the rubber compound, and the pre-processed real-time operating data, the pressure and filtration cycle of the filter press process are dynamically adjusted by integrating particle swarm optimization and adaptive mechanisms. Step 5: Based on the target pressure and target filtration cycle generated by dynamic optimization, the reverse design method is used to derive the equipment action sequence, and the pre-processed real-time working condition data is integrated to construct a multi-dimensional closed loop, dynamically generating control commands for the hydraulic system and drive mechanism; This intelligent control method for dewatering filter presses based on styrene-butadiene rubber (SBR) processing achieves closed-loop intelligent control from perception, diagnosis, prediction to optimization and execution through full-process data-driven and multi-module collaboration. Step one ensures high-quality data input; step two provides early fault warnings for key components, improving equipment reliability; step three integrates lookup table interpolation, Cross-WLF online fitting, and transfer learning to accurately predict the rheological properties of the rubber compound; step four dynamically optimizes pressure and cycle based on target moisture content and rheological state to ensure dewatering quality; and step five generates high-precision, safe, and reliable equipment control commands through reverse engineering and multi-dimensional closed-loop correction. The overall method combines high precision, strong adaptability, self-learning capability, and operational safety, significantly improving dewatering efficiency, product quality stability, and intelligent operation and maintenance.

[0023] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and its improved concept, should be covered within the scope of protection of the present invention.

Claims

1. An intelligent control system for a dewatering filter press based on styrene-butadiene rubber processing, integrated into a styrene-butadiene rubber intelligent production platform, characterized in that, include: The data acquisition and preprocessing module is used to collect the filter press operation data during the dehydration and filter press process of styrene-butadiene rubber, preprocess the collected filter press operation data, and generate a preprocessed filter press operation data sequence. The filter press fault early warning module is used to identify potential fault modes of key components of the dewatering filter press and issue early warnings based on real-time and historical data in the pre-processed filter press operating condition data sequence. The filter press rheological property decision module is used to predict the viscosity and key rheological parameters of the rubber compound during the filter press process based on real-time data from the pre-processed filter press condition data sequence and similar historical batch data of styrene-butadiene rubber grade, using a combination of rheological property lookup table interpolation, online parameter fitting of the Cross-WLF equation and transfer learning. The pressure filter dewatering quality control module is used to dynamically adjust the pressure and filtration cycle of the pressure filter process based on the target moisture standard, the predicted results of the rheological properties of the rubber compound and the pre-processed real-time working condition data, and by integrating particle swarm optimization and adaptive mechanisms. The pressure filtration execution strategy control module is used to derive the equipment action sequence based on the target pressure and target filtration cycle generated by dynamic optimization, and to build a multi-dimensional closed loop by integrating pre-processed real-time operating data, and dynamically generate control commands for the hydraulic system and drive mechanism.

2. The intelligent control system for a dewatering filter press based on styrene-butadiene rubber processing according to claim 1, characterized in that, The process by which the filter press fault early warning module identifies potential fault modes of key components of the dewatering filter press and issues early warnings based on real-time and historical data from the pre-processed filter press operating condition data sequence includes: Obtain the working pressure of the main cylinder in the current filter press cycle and the running time of the current filter press stage. Determine the start and end times of the pressure holding stage based on the current running time of the filter press stage. Calculate the difference between the starting and ending values ​​of the working pressure of the main cylinder in the pressure holding stage as the pressure decay value of the pressure holding stage. Determine if the pressure decay value of the pressure holding stage is not less than the absolute threshold of pressure decay, the pressure decay deviation is not less than the pressure decay deviation threshold, and the clamping time deviation is not greater than the upper limit of the clamping time deviation. If these conditions are met, output the current filter press cycle number, the name of the key component (main hydraulic cylinder), and the fault mode name (hydraulic system internal leakage). Obtain the pressure plate displacement, main motor current, and current filter pressing stage running time of the current filter pressing cycle. Sum the main motor current value of the unloading stage to obtain the unloading stage current integral value. Determine if the unloading stage current integral value is not less than the unloading current integral absolute threshold and the unloading current integral deviation is not less than the unloading current integral deviation threshold. If these conditions are met, output the current filter pressing cycle number, the key component name (pulling plate mechanism), and the fault mode name (pulling plate mechanism jamming).

3. The intelligent control system for a dewatering filter press based on styrene-butadiene rubber processing according to claim 1, characterized in that, The filter press rheological property decision module, based on styrene-butadiene rubber grade, real-time data from the pre-processed filter press operating condition data sequence, and similar historical batch data, employs a collaborative mechanism of rheological property lookup interpolation and Cross-WLF online fitting, including the following steps: Obtain the current styrene-butadiene rubber grade, use the reference viscosity at the corresponding reference temperature and the temperature-pressure-viscosity three-dimensional lookup table to obtain the rubber inlet temperature, main cylinder working pressure, and online viscosity sensor measured viscosity time series for the current filter press cycle. For each time point, obtain the rubber inlet temperature and main cylinder working pressure, locate the smallest rectangular area containing the corresponding temperature and pressure points in the lookup table, calculate the temperature change rate and pressure change rate. If both are less than their respective preset thresholds, perform bilinear interpolation; otherwise, fit a local quadratic surface based on the four corner points, and calculate the function value of the surface at the corresponding temperature and pressure points as the initial viscosity value. When entering the parameter update stage, the Cross-WLF rheological parameters of the same grade and those that have completed the pressure filtration task are obtained from the historical batch library. The sample buffer and counter are initialized. For each time point, when the counter equals the number of samples required for the preset parameter update, the predicted viscosity of the model is divided into the same interval to obtain the predicted probability distribution. The Williams-Landel-Ferry parameters are optimized through the objective function, and the updated Williams-Landel-Ferr parameters are output. Using the latest Williams-Landel-Ferry parameters, combined with the rubber inlet temperature at each time point, the Cross-WLF model is substituted to calculate and output the final viscosity estimate.

4. The intelligent control system for a dewatering filter press based on styrene-butadiene rubber processing according to claim 3, characterized in that, The process of the filter press rheological property decision module using a migration calibration mechanism based on historical batches includes: Obtain the styrene-butadiene rubber grade and process path identifier of the current filtration cycle, obtain the online viscosity sensor measured viscosity time series and the number of accumulated valid samples for the current filtration cycle. If the number of valid samples is less than the preset minimum number of samples, perform migration calibration; otherwise, terminate the process. Select the batch with the same styrene-butadiene rubber grade, consistent process path and completed filtration task from the historical batch library as the migration source, and obtain the final converged Cross-WLF parameters of the migration source batch. Obtain the operating condition similarity corresponding to the migration source, use the Cross-WLF parameters of the migration source as the current initial parameters, optimize the current Cross-WLF parameters through the corresponding objective function, and output the update result.

5. The intelligent control system for a dewatering filter press based on styrene-butadiene rubber processing according to claim 4, characterized in that, The process by which the pressure filtration rheological property decision module predicts the viscosity and key rheological parameters of the filtration compound during the pressure filtration process includes: The system acquires the inlet temperature of the rubber compound, the working pressure of the main cylinder, the initial viscosity value obtained by interpolation from a lookup table, and the viscosity value predicted by the Cross-WLF model. It also acquires the rate of temperature change and the rate of pressure change. If both the rate of temperature change and the rate of pressure change are less than their respective preset thresholds, the interpolation result and the Cross-WLF prediction value are weighted and fused according to preset steady-state weights and dynamic weights to obtain the real-time viscosity. Otherwise, the real-time viscosity is obtained by weighting and fusion according to preset dynamic weights and steady-state weights. The system acquires the output value of the Cross-WLF model when the shear rate is zero as the zero-shear viscosity. It acquires the viscosity and shear rate sequence in the preset medium-high shear rate range, takes the natural logarithm of both, and calculates the absolute value of their slopes as the shear thinning index. The system outputs the real-time viscosity, the zero-shear viscosity, and the shear thinning index.

6. The intelligent control system for a dewatering filter press based on styrene-butadiene rubber processing according to claim 1, characterized in that, The pressure filtration and dewatering quality control module, based on the target moisture standard, predicted rheological properties of the rubber compound, and pre-processed real-time operating data, integrates particle swarm optimization and adaptive mechanisms to dynamically adjust the pressure and filtration cycle of the pressure filtration process. The process includes: Obtain the target moisture standard, initial flow resistance, and dehydration rate decay coefficient. Obtain the main cylinder working pressure, filtrate instantaneous flow rate, pressure plate displacement, current filtration stage running time, and previous cycle filtrate instantaneous flow rate. Calculate the current flow rate change as the difference between the current and previous cycle filtrate instantaneous flow rates. Look up the target dehydration trajectory table to obtain the target flow rate change corresponding to the current filtration stage. Based on the comparison between the current flow rate change and the target flow rate change, if it is less, update the main cylinder working pressure and pressure holding stage running time according to the pressure increase and time extension rules. If it is greater, update according to the pressure decrease and time shortening rules. If they are equal, keep the main cylinder working pressure and the running time of each stage unchanged. The system acquires the main motor current and hydraulic system power. If either the main motor current or the hydraulic system power exceeds the limit, the main cylinder working pressure and the pressure holding stage running time are updated according to the safety pressure reduction and safety delay rules. Particle swarm optimization is run to generate candidate parameter combinations and merge them with the adaptive adjustment results to form a candidate scheme. If the flow rate change deviates from the target by more than the allowable deviation for two consecutive control cycles, the current plan is terminated, particle swarm optimization is reinitialized, new parameters are generated, and execution is switched immediately. The main cylinder working pressure, the pressing stage running time, and the pressure holding stage running time are output.

7. The intelligent control system for a dewatering filter press based on styrene-butadiene rubber processing according to claim 1, characterized in that, The process by which the pressure filtration execution strategy control module derives the equipment action sequence based on dynamically optimized target pressure and target filtration cycle and using a reverse design method includes: Obtain the target pressure and target filtration cycle, obtain the current time and the start and end times of each filtration stage, determine the current filtration stage based on the current time and stage start and end times, obtain the final pressure and final time of the current filtration stage, if the target pressure is greater than the maximum allowable working pressure of the equipment, then the target pressure is equal to the maximum allowable working pressure of the equipment, obtain the control cycle number, if the control cycle number is not the first cycle and the target pressure is less than the minimum effective filtration pressure, then the target pressure is equal to the target pressure of the previous control cycle, if the control cycle number is the first cycle and the target pressure is less than the minimum effective filtration pressure, then the target pressure is equal to the minimum effective filtration pressure; The system acquires chamber pressure, filtrate flow rate, and hydraulic oil temperature; it also acquires the duration of the current filtration stage, speed penalty parameters, and acceleration penalty parameters. Using the target pressure and target filtration cycle as boundary conditions, it reverse-engineers the equipment action sequence from the final pressure and final time. The equipment action sequence is generated by solving an optimization problem and is then output.

8. The intelligent control system for a dewatering filter press based on styrene-butadiene rubber processing according to claim 7, characterized in that, The process by which the pressure filtration execution strategy control module integrates preprocessed real-time operating data to construct a multi-dimensional closed-loop correction includes: Obtain the current chamber pressure and the expected pressure. If the absolute deviation between the two is greater than the pressure tracking deviation limit, increase the feed valve opening by the feed valve opening increment. Obtain the filtrate flow rate for the current and the two most recent historical control cycles. If the filtrate flow rate decreases by more than the filtrate flow rate decrease limit for two consecutive cycles, and the main frequency energy of the feed pressure fluctuation is greater than the energy rise limit, increase the duration of the current stage by the stage extension time. If the current pressure filtration stage is equal to the pressure boosting stage, multiply the feed valve opening by the pressure boosting opening attenuation factor. Get the hydraulic oil temperature and its upper limit. If the hydraulic oil temperature is greater than the upper limit, get the hydraulic oil viscosity and its upper limit. If the hydraulic oil viscosity is greater than the upper limit, multiply the target pressure by the target pressure attenuation factor, set the auxiliary cooling pump operation flag to enabled, and output the feed valve opening, target pressure, auxiliary cooling pump operation flag, and current stage duration.

9. The intelligent control system for a dewatering filter press based on styrene-butadiene rubber processing according to claim 8, characterized in that, The process by which the pressure filtration execution strategy control module dynamically generates control commands for the hydraulic system and drive mechanism includes: Obtain the corrected chamber pressure and chamber pressure upper limit. If the chamber pressure is greater than the upper limit, set it as the upper limit. Obtain the corrected target filtrate flow rate and filtrate flow rate upper limit. If the target filtrate flow rate is greater than the upper limit, set it as the upper limit. Obtain the hydraulic oil temperature and hydraulic oil temperature inactivation limit. If the hydraulic oil temperature is greater than the inactivation limit, obtain the corrected feed valve opening and keep it unchanged. At the same time, obtain the press pump operation flag and set it to stop. Otherwise, obtain the corrected feed valve opening and press pump operation flag. Obtain the corrected target speed. Generate and output feed valve control commands, press pump control commands, and servo motor control commands.

10. A method for intelligent control of a dewatering filter press based on styrene-butadiene rubber processing, applied to an intelligent control system for a dewatering filter press based on styrene-butadiene rubber processing as described in any one of claims 1-9, characterized in that, Includes the following steps: Step 1: Collect filter press operation data during the dehydration and filter press process of styrene-butadiene rubber, preprocess the collected filter press operation data, and generate a preprocessed filter press operation data sequence. Step 2: Based on the real-time and historical data in the pre-processed filter press operating condition data sequence, identify potential failure modes of key components of the dewatering filter press and issue early warnings; Step 3: Based on the real-time data and similar historical batch data in the pre-processed filter press condition data sequence of styrene-butadiene rubber grade, the viscosity of the rubber compound and its key rheological parameters during the filter press process are predicted by a combination of rheological property lookup table interpolation, online parameter fitting of the Cross-WLF equation and transfer learning. Step 4: Based on the target moisture standard, the predicted rheological properties of the rubber compound, and the pre-processed real-time operating data, the pressure and filtration cycle of the filter press process are dynamically adjusted by integrating particle swarm optimization and adaptive mechanisms. Step 5: Based on the target pressure and target filtration cycle generated by dynamic optimization, the reverse design method is used to derive the equipment action sequence, and the pre-processed real-time working condition data is integrated to construct a multi-dimensional closed loop, dynamically generating control commands for the hydraulic system and drive mechanism.

Images