A method for synergic control of process parameters in a PE film preparation process

Abstract

The present application relates to the technical field of industrial control system, and more particularly to a PE film preparation process parameter collaborative control method, comprising collecting multiple types of core process parameters through a sensor network deployed at full-process nodes; using a synchronization algorithm based on unified timing and a device physical position topology map to achieve spatiotemporal alignment of data in the material flow dimension; using a nonlinear dynamic correlation algorithm to construct a multidimensional coupling mapping model among extrusion speed, blow pressure and traction speed; using a multivariable collaborative control algorithm to solve adjustment priorities and generate a collaborative action vector including heating, motor and air pressure to suppress coupling conflicts; and introducing a state compensation mechanism based on a disturbance observer to implement closed-loop feedback adjustment to suppress time-varying disturbances. The present application realizes precise decoupling of strongly coupled parameters and significantly improves process stability.

Description

Technical Field

[0001] This invention relates to the field of industrial control system technology, and in particular to a method for coordinated control of process parameters in the PE film preparation process. Background Technology

[0002] Polyethylene (PE) film, as a widely used basic material, involves several highly interconnected processes in its production, including raw material proportioning, melt extrusion, blow-forming, and winding and cooling. Existing technologies struggle to effectively address the strong correlation between extrusion speed, blow-forming pressure, and traction speed in the complex coupling characteristics of PE film production. This leads to oscillations easily induced during parameter adjustments, affecting film surface quality.

[0003] Meanwhile, due to the insufficient ability of existing control systems to suppress nonlinear time-varying disturbances, the production process is difficult to maintain long-term stability when faced with uncertainties such as raw material fluctuations or mechanical wear. Summary of the Invention

[0004] The purpose of this invention is to provide a method for coordinated control of process parameters in the PE film preparation process to solve the technical problems in the background art.

[0005] This invention provides a method for coordinated control of process parameters in the preparation of PE film, comprising: S1. By deploying a sensor network at every node of the PE film production line, core process parameters are captured in real time. These core process parameters include at least the raw material ratio accuracy, the temperature gradient of the melt extrusion section, the screw speed of the extruder, the melt pressure of the die head, the blowing air pressure, the cooling air ring speed, the traction roller linear speed, and the finished product winding tension. S2. Using a sampling synchronization algorithm based on a unified timing clock, sensor data with different sampling frequencies are mapped to a unified time axis, and combined with the physical location topology of the production line equipment, the lag time of material flow in each section is calculated. The lag time is estimated based on the extrusion mass flow rate determined by the current main drive motor speed of the extruder, screw geometric parameters, and melt density. S3. Based on the associated samples in the historical production database, a multidimensional coupling mapping model between extrusion speed, blowing pressure and traction speed is constructed using a nonlinear dynamic correlation algorithm. The contribution rate of fluctuations in each parameter to film thickness deviation, tensile strength and transparency is quantified. The multidimensional coupling mapping model establishes the mapping function between process parameters and quality indicators through regression analysis. S4. Based on the real-time collected parameter deviations, the adjustment priority of each actuator is calculated using a multivariable collaborative control algorithm, and a collaborative action vector including heating current compensation, motor frequency fine-tuning and air pressure proportional adjustment is generated. The collaborative action vector includes a pulse width modulation signal that adjusts the power output cycle of the temperature control zone. S5. By comparing the online detection results of the finished thin film with the preset quality standards, a state compensation mechanism based on a disturbance observer is introduced to dynamically correct the internal parameters of the control model, so as to achieve coordinated control of process parameters in the preparation process.

[0006] In some embodiments, the S1 process specifically includes: S1.1. Utilize multiple sets of armored thermocouples deployed in each section of the extruder barrel to monitor in real time the thermodynamic characteristics of polyethylene raw material from the solid state in the feeding section, through the compression section, and finally to the molten state in the metering section; S1.2. The armored thermocouple is installed independently in each temperature control zone, and the thermoelectric potential signal is transmitted to the distributed input / output module with cold junction compensation circuit via shielded twisted pair cable. The cold junction compensation circuit maintains the conversion accuracy of the thermocouple signal when the ambient temperature fluctuates. S1.3. The melt pressure of the die head is collected by a melt pressure sensor located in the flow channel behind the filter screen. The melt pressure sensor adopts a standard current signal output and is used in conjunction with an analog-to-digital converter for digital processing. S1.4. Pressure fluctuation data is obtained by a high-frequency micro pressure transmitter installed in the air inlet pipe and inside the membrane bubble of the blown die head. The pressure fluctuation data includes at least the absolute pressure value and the rate of change of pressure over time. S1.5. The weighing hopper and weighing sensor of the feeding system are used to control the accuracy of the raw material ratio. The feeding system is equipped with multiple independent weighing hoppers, which store polyethylene granules, color masterbatch and functional additives respectively. The system calculates the descent rate of the hopper weight in real time and transmits the real-time feeding rate to the main control unit through the industrial Ethernet communication protocol.

[0007] In some embodiments, the S2 process specifically includes: S2.1. Achieve millisecond-level clock synchronization using a sampling synchronization algorithm based on a unified timing clock; S2.2. Identify the arrival time of data packets from each sensor and extract the characteristic jump points in each data stream. The characteristic jump points are composed of sudden changes in screw speed or pressure pulse points. S2.3. Calculate the relative offset between different data streams using cross-correlation analysis, and use interpolation algorithm to upsample low-frequency data to make all process parameters synchronized at a preset time resolution; S2.4. During the spatial alignment process in the material flow dimension, a mathematical model for material conveying is established. Based on the current main drive motor speed of the extruder, screw geometric parameters, and melt density, the extrusion mass flow rate of the melt per unit time is calculated in real time. S2.5. Combining the traction speed and blow-up ratio, the transit time of the film between each monitoring section is calculated. The lag time is determined by the physical length of each monitoring section on the production line, the melt density, the average cross-sectional area of ​​the material, and the extrusion mass flow rate calculated in real time. The quality deviation at the finished product end is traced back to the parameter fluctuation at the extrusion end. Under the condition of multi-layer co-extrusion, the mathematical model establishes the spatial mapping relationship after multi-layer composite by calculating the extrusion ratio of each extruder, and introduces a viscosity correction coefficient to differentiate the lag time of different layers in the flow channel.

[0008] In some embodiments, the S3 process specifically includes: S3.1. Denoise the original large data stream by using a filtering algorithm with a preset moving average window width to remove outliers; S3.2. Extract feature vectors, wherein the feature vectors include at least the mean, standard deviation, rate of change, and cross-parameter mutual information of each parameter within a preset time window, wherein the mutual information is used to quantify the degree of coupling between parameters; S3.3. Using regression analysis based on gradient boosting decision tree, establish a mapping function between process parameters and quality indicators, wherein the quality indicators include at least film thickness deviation, tensile strength and transparency; S3.4. The multidimensional coupled mapping model models the interaction between extrusion speed and die pressure, inflation pressure and film bubble stability, and traction speed and longitudinal tensile strength. A first preset coefficient is used as the sensitivity coefficient of extrusion speed to die pressure, and a second preset coefficient is used as the weight of the influence of inflation pressure on transverse thickness uniformity.

[0009] In some embodiments, the method further includes comparing the current batch data with historical data in real time to address batch differences in raw materials, using a nonlinear dynamic correlation algorithm to calculate the sensitivity coefficient of the current batch of raw materials to process fluctuations, and adjusting the gain of the temperature control loop when the sensitivity of the current raw materials to temperature exceeds a preset threshold.

[0010] 6. The method according to claim 1, wherein the specific process of S4 includes: S4.1. The adjustment priority of each actuator is calculated using a multivariable collaborative control algorithm. By calculating the inverse matrix of the gain matrix, the mutually coupled process parameters are decomposed into independent control loops, and the mutual interference between loops is eliminated by using a feedforward compensation link. S4.2. When a change in film thickness and a fluctuation in die pressure are detected, commands to adjust the extrusion speed and traction speed are issued simultaneously, and the blowing pressure is finely adjusted according to the feedback signal from the high-frequency micro pressure transmitter to maintain a constant film bubble diameter. In some embodiments, the heating current compensation is achieved through pulse width modulation technology. Based on the thermal inertia characteristics of the temperature control zone, the power output cycle is dynamically adjusted to limit the temperature fluctuation range within a preset temperature deviation range. In multi-layer co-extrusion, the collaborative control strategy is achieved through multi-machine linkage, maintaining a constant total thickness and film width while correcting the layer thickness ratio. In the case of product switching, a smooth transition algorithm is adopted to perform feedforward adjustment based on the melting enthalpy of the new raw material and control the step amplitude of motor frequency fine-tuning. For ultra-thin film preparation, the collaborative control algorithm integrates fuzzy neuron control logic, pre-calculates the influence of cooling air ring speed on film bubble tension, and simultaneously fine-tunes the traction speed and blowing pressure when adjusting the air speed to maintain the decoupled balance between parameters.

[0011] In some embodiments, the specific process of S5 includes: using a disturbance observer to estimate the load disturbance outside the system and the parameter changes inside the system, identifying the hardware performance degradation state by monitoring the abnormal deviation of the motor torque current, and adding a reverse compensation term to the control law to offset the decrease in control accuracy. Closed-loop feedback regulation includes the intervention of an online thickness measurement system. An infrared multi-point scanning thickness gauge deployed in front of the winding station completes a full-width scan at a preset cycle, converting the acquired thickness data from multiple measurement points into a feedback vector. The feedback vector triggers the adjustment of the segmented airflow of the air ring. The air ring speed control uses a frequency converter to adjust the frequency of the centrifugal fan, and multiple pneumatic regulating valves arranged in the circumferential direction of the air ring are used to adjust the uniformity of film bubble cooling, thereby controlling the haze and transparency of the film.

[0012] In some embodiments, under multi-layer co-extrusion conditions, the disturbance observer identifies the target extruder that causes the finished product thickness deviation and feeds back the online layer thickness detection data to the frequency converter of the target extruder; for the thickness fluctuation of the ultrathin film, a predictive control algorithm based on a multidimensional coupled mapping model is used to predict the impact of damper adjustment on the subsequent finished product thickness based on the film linear velocity.

[0013] In some embodiments, the collaborative control method further includes big data analysis and storage optimization processing, utilizing production records stored on a cloud server to extract the optimal combination of process parameters for a specific raw material grade through a big data predictive analysis module.

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention overcomes the limitations of traditional single closed-loop regulation in handling conflicts between extrusion speed, blowing pressure and traction speed by constructing a multi-dimensional process parameter deep correlation model. During parameter adjustment, it can calculate the optimal action vector in real time, which significantly reduces film thickness deviation and improves film strength consistency. 2. This invention introduces nonlinear compensation based on a disturbance observer, which can identify and suppress disturbances caused by raw material batch fluctuations, mechanical component wear, and changes in ambient temperature; when faced with external disturbances, the recovery time of the production process is significantly shortened, ensuring the long-term stability of the production process; 3. This invention eliminates the lag problem in the distributed sampling process by using preprocessing techniques based on timestamp synchronization and spatial topology mapping, thereby achieving matching between process data and material status; 4. This invention utilizes big data acquisition and predictive control algorithms to capture and compensate for minute disturbances occurring on the production line in advance, effectively preventing the generation of defective products due to control lag. 5. This invention reduces the frequency of ineffective actions through a collaborative control method, thereby reducing energy consumption. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart of the method for coordinated control of process parameters in the PE film preparation process of the present invention; Figure 2 This is a schematic diagram of the collaborative control strategy of the present invention. Detailed Implementation

[0017] The following will be based on embodiments of the present invention. Figures 1-2 The technical solutions in the embodiments of the present invention will be clearly and completely described together. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0018] Application Overview In the pursuit of high output and high quality in production, establishing a comprehensive industrial monitoring and parameter control system is crucial for ensuring the physical strength, transparency, and thickness consistency of films. Among these, collaborative control of process parameters based on big data acquisition is a key technological path to achieve high-quality PE film preparation. This technology utilizes a sensor network deployed at key nodes of the production line to acquire real-time process data such as temperature, pressure, flow rate, and speed, constructing a dynamic monitoring and feedback mechanism covering the entire process. Through the unified integration and analysis of large amounts of production data, the thermodynamic and hydrodynamic states during the preparation process can be mapped, providing data support for strategy optimization under complex operating conditions. Based on this, this invention provides a method for collaborative control of process parameters in the PE film preparation process, which is detailed below.

[0019] Example 1 First, execute S1, which uses a sensor network deployed at every node of the PE film production line to capture core process parameters in real time. These core process parameters include at least the accuracy of raw material proportioning, temperature gradient of the melt extrusion section, screw speed of the extruder, melt pressure of the die head, blowing air pressure, cooling air ring speed, traction roller linear speed, and finished product winding tension.

[0020] Specifically, the melt pressure in the die head is acquired using a high-temperature melt pressure sensor. This sensor is installed in the flow channel behind the filter screen, with a range set from 0 MPa to 70 MPa. The signal output uses a standard current signal of 4 mA to 20 mA. This current signal enhances the anti-interference capability in the complex electromagnetic environment generated by high-power equipment such as extruder frequency converters. The high-temperature melt pressure sensor probe uses a special alloy diaphragm capable of withstanding the impact of melt at temperatures exceeding 250 degrees Celsius. During acquisition, the pressure signal is digitized by a 24-bit analog-to-digital converter, and the sampling frequency is set to 500 Hz.

[0021] Furthermore, the monitoring points for the inflation air pressure are located in the air inlet pipe and inside the membrane bubble of the inflation die, and pressure fluctuation data are acquired in real time through a high-frequency micro-pressure transmitter. This transmitter has a range of 0 Pa to 5000 Pa and a sampling frequency set to 1000 Hz to ensure the capture of transient pulsations caused by compressor pulsation or airflow disturbances during the inflation process. The collected data includes not only absolute pressure values ​​but also the rate of pressure change over time, used to assess the stability of the membrane bubble.

[0022] Furthermore, the precise control of raw material proportioning is achieved through a feeding system equipped with four independent weighing hoppers, each storing polyethylene granules, masterbatch, antistatic agents, and other functional additives. Each weighing hopper is fitted with a high-precision cantilever beam load cell, ensuring the proportioning error is controlled within a preset error range of 0.2%. The feeding system calculates the descent rate of the hopper weight in real time and transmits this real-time feeding rate to the main control unit via the Profinet real-time industrial Ethernet communication protocol.

[0023] All collected data, including the accuracy of raw material proportions, temperature gradient of the melt extrusion section, screw speed of the extruder, melt pressure of the die head, blowing air pressure, cooling air ring speed, traction roller linear speed, and finished product winding tension, are encapsulated into data frames with timestamps, forming a raw big data stream.

[0024] Next, S2 is executed, which uses a sampling synchronization algorithm based on a unified timing clock to map sensor data with different sampling frequencies to a unified time axis. Combined with the physical location topology of the production line equipment, the lag time of material flow in each section is calculated. The lag time is estimated based on the extrusion mass flow rate determined by the current main drive motor speed of the extruder, screw geometry parameters, and melt density.

[0025] Specifically, the arrival time of data packets from each sensor is identified, and characteristic abrupt change points in each data stream are extracted, such as sudden changes in screw speed or pressure pulse points. The relative offset between different data streams is calculated using cross-correlation analysis. For the low-frequency temperature signal (10 Hz) and the high-frequency pressure signal (500 Hz), a cubic spline interpolation algorithm is used to upsample the low-frequency data, ensuring synchronization of all process parameters at a time resolution of 1 millisecond.

[0026] Furthermore, based on the current main drive motor speed of the extruder, screw geometry parameters (such as screw diameter, lead, and groove depth), and melt density, the extruded mass flow rate of the melt per unit time is calculated in real time. Combined with the traction speed and blow-up ratio, the transit time of the film between each monitoring section is estimated. The formula for calculating the lag time of the material from the extruder die to the winding station is as follows: in, L i Indicates the first on the production line i The physical length of each monitoring segment; ρ This indicates the density of polyethylene melt or solid film; S i This represents the average cross-sectional area of ​​the material segment; This represents the extrusion mass flow rate calculated in real time by the main control unit. Using this formula, the system can accurately trace quality deviations at the finished product end back to parameter fluctuations at the extrusion end, achieving spatial alignment of process data in the material flow dimension.

[0027] Next, S3 is executed. Based on the associated samples in the historical production database, a nonlinear dynamic correlation algorithm is used to construct a multidimensional coupling mapping model between extrusion speed, blowing pressure and traction speed. The contribution rate of each parameter fluctuation to film thickness deviation, tensile strength and transparency is quantified. The multidimensional coupling mapping model establishes the mapping function between process parameters and quality indicators through regression analysis.

[0028] Specifically, the construction process of the multidimensional coupled mapping model first involves denoising the original data. A filtering algorithm with a moving average window width of 50 sampling points is used to remove outliers. Then, feature vectors are extracted, including the mean, standard deviation, rate of change, and cross-parameter mutual information of each parameter within a preset time window. Regression analysis based on gradient boosting decision trees is used to establish a mapping function between process parameters and quality indicators (film thickness deviation, tensile strength, and transparency). The multidimensional coupled mapping model focuses on modeling the interactions between extrusion speed and die pressure, blow-up pressure and film bubble stability, and traction speed and longitudinal tensile strength. Specifically, the sensitivity coefficient of extrusion speed to die pressure is set to a first preset coefficient of 0.85, and the weight of the influence of blow-up pressure on transverse thickness uniformity is set to a second preset coefficient of 0.72. The model dynamically updates the weights of each node through a combination of offline training and online incremental learning to adapt to the needs of different production grades.

[0029] Then, S4 is executed. Based on the parameter deviation collected in real time, the adjustment priority of each actuator is calculated using a multivariable collaborative control algorithm. A collaborative action vector is generated, which includes heating current compensation, motor frequency fine-tuning, and air pressure proportional adjustment. The collaborative action vector includes a pulse width modulation signal that adjusts the power output cycle of the temperature control zone.

[0030] Specifically, the multivariable collaborative control algorithm first decomposes the coupled process parameters (such as the combined effect of extrusion rate and linear velocity on thickness) into independent control loops by calculating the inverse of the gain matrix, and eliminates the mutual interference between loops through a feedforward compensation stage. When a thinning of the film thickness and a drop in die pressure are detected, a linkage command is simultaneously issued to increase the extrusion speed by 2% and decrease the traction speed by 1%. At the same time, the blowing pressure is finely adjusted through feedback signals from a high-frequency micro-pressure transmitter to maintain a constant film bubble diameter.

[0031] Furthermore, the heating current compensation employs pulse width modulation technology to dynamically adjust the power output cycle based on the thermal inertia characteristics of the temperature control zones. This strictly limits temperature fluctuations to a preset temperature deviation range of ±0.5 degrees Celsius.

[0032] Finally, S5 is executed. By comparing the online detection results of the finished thin film with the preset quality standards, a state compensation mechanism based on the disturbance observer is introduced to dynamically correct the internal parameters of the control model, so as to achieve coordinated control of the process parameters in the preparation process.

[0033] Specifically, the disturbance observer is used to estimate external load disturbances and internal parameter changes in real time. By monitoring abnormal deviations in motor torque current, it identifies the degree of filter blockage. When the torque current exceeds the reference value by 5%, it is identified as hardware performance degradation, and an inverse compensation term is automatically added to the control law to offset the decrease in control accuracy caused by hardware degradation.

[0034] Furthermore, the closed-loop feedback regulation also includes the intervention of an online thickness measurement system. An infrared multi-point scanning thickness gauge deployed in front of the winding station completes a full-width scan at a preset cycle of 5 seconds per revolution. The acquired data from 512 thickness measurement points are converted into a feedback vector. The feedback vector triggers precise adjustment of the segmented airflow in the cooling air ring. The cooling air ring speed control uses a frequency converter to adjust the frequency of the centrifugal fan, keeping the airflow speed within a preset range of 5 to 25 meters per second. By using 72 pneumatic regulating valves evenly arranged in the circumference of the air ring, high-precision adjustment of the film bubble cooling uniformity is achieved, thereby controlling the haze and transparency of the film.

[0035] Example 2 This embodiment relates to the coordinated control of process parameters in the preparation of a three-layer co-extruded polyethylene film. The three-layer structure includes an inner layer, a middle layer, and an outer layer, each of which is supplied with melt by an independent extrusion system.

[0036] During S1 execution, the system equips each of the three extruders with an independent data acquisition unit. The screw speed, temperature of each section, and melt pressure of each extruder are uploaded to the central controller in real time. The multi-dimensional sensing network extends the monitoring of the layer thickness ratio, and the thickness data of each layer is analyzed in real time through an online multi-layer infrared thickness gauge.

[0037] When executing S2, the mathematical model needs to consider the fusion process of the three melts at the confluence die. By calculating the extrusion ratio of each extruder, the spatial mapping relationship of the three-layer composite is established. Since the melt viscosity of different layers is different, the model introduces a viscosity correction coefficient to compensate for the lag time of different layers in the flow channel, so as to ensure the parameter alignment accuracy of the confluence point.

[0038] To execute S3, it is necessary to calculate the impact of the first extruder speed fluctuation on the total thickness and layer thickness distribution. Using feature selection based on correlation analysis, the parameter combination that contributes the most to the film transparency is extracted, namely the ratio of the intermediate layer melt temperature to the die pressure.

[0039] When executing S4, if the online detection system detects that the inner layer thickness ratio is too low, the collaborative control algorithm not only generates a command to increase the speed of the first extruder, but also fine-tunes the speeds of the second and third extruders, and synchronously adjusts the total traction speed. This coordinated adjustment ensures that while correcting the layer thickness ratio, the total thickness and film width remain constant.

[0040] When executing S5, an adaptive decoupling algorithm is introduced to address the high coupling characteristics of three-layer co-extrusion. The disturbance observer not only monitors the motor load but also evaluates the instability of the interfaces between each layer in real time. When uneven melt flow rate is detected causing interface disturbance, the heating current compensation value of each extruder layer is automatically adjusted to optimize melt viscosity matching and suppress interlayer disturbances caused by differences in rheological properties.

[0041] Furthermore, in multi-layer co-extrusion operations, segmented adjustment of the cooling air ring is particularly critical. Due to the high overall heat content of the multi-layer film, the air speed control logic increases compensation for the heat dissipation path of the inner layer. The frequency converter adjusts the output pressure of the centrifugal fan, and through the nonlinear adjustment of 72 air valves, it compensates for the cooling differences caused by the uneven temperature distribution of the multi-layer melt, maintaining the geometric symmetry of the film bubble.

[0042] Example 3 This embodiment describes the synergistic control of the present invention in the preparation of multilayer co-extruded polyethylene barrier film, involving a five-layer co-extrusion structure.

[0043] The S1 is configured to support full parameter acquisition from five independent extruders, each equipped with an independent die melt pressure sensor and temperature control system. The data acquisition layer of the sensing network synchronously aggregates the five melt data streams through a distributed I / O module.

[0044] In S2, the system calculates the independent path delay of each layer of material before the merging die. Due to the significant difference in rheological properties between the barrier layer (such as EVOH) and the PE layer, the spatial consistency mathematical model introduces a multiphase flow slip coefficient correction.

[0045] In S3, the system uses a nonlinear dynamic correlation algorithm to calculate the potential impact of temperature on interlayer peel strength. This step is achieved by extracting feature vectors of interface defects from the historical database.

[0046] In S4, it is necessary not only to maintain the thickness ratio of each layer, but also to control the interfacial shear stress. When the extrusion amount of the main PE layer is increased, the system uses decoupling logic to simultaneously fine-tune the extrusion speed of the barrier layer and the adhesive layer, and adjusts the heating current compensation to change the melt viscosity of each layer, so as to ensure the smoothness of the interlayer interface.

[0047] The global robustness optimization in S5 introduces a disturbance observer for multi-layer structures. This observer can identify which extruder's fluctuations caused the final product thickness deviation. The closed-loop feedback regulation not only affects the total thickness but also feeds back to the frequency converters of each extruder through online layer thickness detection data.

[0048] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the scope of the invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0049] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for synergic control of process parameters in a PE film production process, characterized in that, include: S1. By deploying a sensor network at every node of the PE film production line, core process parameters are captured in real time. These core process parameters include at least the raw material ratio accuracy, the temperature gradient of the melt extrusion section, the screw speed of the extruder, the melt pressure of the die head, the blowing air pressure, the cooling air ring speed, the traction roller linear speed, and the finished product winding tension. S2. Using a sampling synchronization algorithm based on a unified timing clock, sensor data with different sampling frequencies are mapped to a unified time axis, and combined with the physical location topology of the production line equipment, the lag time of material flow in each section is calculated. The lag time is estimated based on the extrusion mass flow rate determined by the current main drive motor speed of the extruder, screw geometric parameters, and melt density. S3. Based on the associated samples in the historical production database, a multidimensional coupling mapping model between extrusion speed, blowing pressure and traction speed is constructed using a nonlinear dynamic correlation algorithm. The contribution rate of fluctuations in each parameter to film thickness deviation, tensile strength and transparency is quantified. The multidimensional coupling mapping model establishes the mapping function between process parameters and quality indicators through regression analysis. S4. Based on the real-time collected parameter deviations, the adjustment priority of each actuator is calculated using a multivariable collaborative control algorithm, and a collaborative action vector including heating current compensation, motor frequency fine-tuning and air pressure proportional adjustment is generated. The collaborative action vector includes a pulse width modulation signal that adjusts the power output cycle of the temperature control zone. S5. By comparing the online detection results of the finished thin film with the preset quality standards, a state compensation mechanism based on a disturbance observer is introduced to dynamically correct the internal parameters of the control model, so as to achieve coordinated control of process parameters in the preparation process.

2. The method of claim 1, wherein, The specific process of S1 includes: S1.

1. Utilize multiple sets of armored thermocouples deployed in each section of the extruder barrel to monitor in real time the thermodynamic characteristics of polyethylene raw material from the solid state in the feeding section, through the compression section, and finally to the molten state in the metering section; S1.

2. The armored thermocouple is installed independently in each temperature control zone, and the thermoelectric potential signal is transmitted to the distributed input / output module with cold junction compensation circuit via shielded twisted pair cable. The cold junction compensation circuit maintains the conversion accuracy of the thermocouple signal when the ambient temperature fluctuates. S1.

3. The melt pressure of the die head is collected by a melt pressure sensor located in the flow channel behind the filter screen. The melt pressure sensor adopts a standard current signal output and is used in conjunction with an analog-to-digital converter for digital processing. S1.

4. Pressure fluctuation data is obtained by a high-frequency micro pressure transmitter installed in the air inlet pipe and inside the membrane bubble of the blown die head. The pressure fluctuation data includes at least the absolute pressure value and the rate of change of pressure over time. S1.

5. The weighing hopper and weighing sensor of the feeding system are used to control the accuracy of the raw material ratio. The feeding system is equipped with multiple independent weighing hoppers, which store polyethylene granules, color masterbatch and functional additives respectively. The system calculates the descent rate of the hopper weight in real time and transmits the real-time feeding rate to the main control unit through the industrial Ethernet communication protocol.

3. The method according to claim 1, characterized in that, The specific process of S2 includes: S2.

1. Achieve millisecond-level clock synchronization using a sampling synchronization algorithm based on a unified timing clock; S2.

2. Identify the arrival time of data packets from each sensor and extract the characteristic jump points in each data stream. The characteristic jump points are composed of sudden changes in screw speed or pressure pulse points. S2.

3. Calculate the relative offset between different data streams using cross-correlation analysis, and use interpolation algorithm to upsample low-frequency data to make all process parameters synchronized at a preset time resolution; S2.

4. During the spatial alignment process in the material flow dimension, a mathematical model for material conveying is established. Based on the current main drive motor speed of the extruder, screw geometric parameters, and melt density, the extrusion mass flow rate of the melt per unit time is calculated in real time. S2.

5. Combining the traction speed and blow-up ratio, the transit time of the film between each monitoring section is calculated. The lag time is determined by the physical length of each monitoring section on the production line, the melt density, the average cross-sectional area of ​​the material, and the extrusion mass flow rate calculated in real time. The quality deviation at the finished product end is traced back to the parameter fluctuation at the extrusion end. Under the condition of multi-layer co-extrusion, the mathematical model establishes the spatial mapping relationship after multi-layer composite by calculating the extrusion ratio of each extruder, and introduces a viscosity correction coefficient to differentiate the lag time of different layers in the flow channel.

4. The method according to claim 1, characterized in that, The specific process of S3 includes: S3.

1. Denoise the original large data stream by using a filtering algorithm with a preset moving average window width to remove outliers; S3.

2. Extract feature vectors, wherein the feature vectors include at least the mean, standard deviation, rate of change, and cross-parameter mutual information of each parameter within a preset time window, wherein the mutual information is used to quantify the degree of coupling between parameters; S3.

3. Using regression analysis based on gradient boosting decision tree, establish a mapping function between process parameters and quality indicators, wherein the quality indicators include at least film thickness deviation, tensile strength and transparency; S3.

4. The multidimensional coupled mapping model models the interaction between extrusion speed and die pressure, inflation pressure and film bubble stability, and traction speed and longitudinal tensile strength. A first preset coefficient is used as the sensitivity coefficient of extrusion speed to die pressure, and a second preset coefficient is used as the weight of the influence of inflation pressure on transverse thickness uniformity.

5. The method according to claim 4, characterized in that, It also includes real-time comparison of current batch data with historical data to address batch differences in raw materials, using a nonlinear dynamic correlation algorithm to calculate the sensitivity coefficient of the current batch of raw materials to process fluctuations, and adjusting the gain of the temperature control loop when the sensitivity of the current raw materials to temperature exceeds a preset threshold.

6. The method according to claim 1, characterized in that, The specific process of S4 includes: S4.

1. The adjustment priority of each actuator is calculated using a multivariable collaborative control algorithm. By calculating the inverse matrix of the gain matrix, the mutually coupled process parameters are decomposed into independent control loops, and the mutual interference between loops is eliminated by using a feedforward compensation link. S4.

2. When a change in film thickness and a fluctuation in die pressure are detected, commands to adjust the extrusion speed and traction speed are issued simultaneously, and the blowing pressure is finely adjusted according to the feedback signal from the high-frequency micro pressure transmitter to maintain a constant film bubble diameter.

7. The method according to claim 6, characterized in that, The heating current compensation is achieved through pulse width modulation technology. Based on the thermal inertia characteristics of the temperature control zone, the power output cycle is dynamically adjusted to limit the temperature fluctuation range within the preset temperature deviation range. In multi-layer co-extrusion, the collaborative control strategy is achieved through multi-machine linkage, which maintains the total thickness and film width constant while correcting the layer thickness ratio. In the case of product switching, a smooth transition algorithm is adopted to make feedforward adjustments based on the melting enthalpy of the new raw material and control the step amplitude of the motor frequency fine-tuning. For the preparation of ultra-thin films, the collaborative control algorithm integrates fuzzy neuron control logic, pre-calculates the influence of cooling air ring speed on film bubble tension, and simultaneously fine-tunes the traction speed and blowing pressure when adjusting the air speed to maintain the decoupled balance between parameters.

8. The method according to claim 1, characterized in that, The specific process of S5 includes: using a disturbance observer to estimate the load disturbance outside the system and the parameter changes inside the system, identifying the hardware performance degradation state by monitoring the abnormal deviation of the motor torque current, and adding a reverse compensation term to the control law to offset the decrease in control accuracy. Closed-loop feedback regulation includes the intervention of an online thickness measurement system. An infrared multi-point scanning thickness gauge deployed in front of the winding station completes a full-width scan at a preset cycle, converting the acquired thickness data from multiple measurement points into a feedback vector. The feedback vector triggers the adjustment of the segmented airflow of the air ring. The air ring speed control uses a frequency converter to adjust the frequency of the centrifugal fan, and multiple pneumatic regulating valves arranged in the circumferential direction of the air ring are used to adjust the uniformity of film bubble cooling, thereby controlling the haze and transparency of the film.

9. The method according to claim 8, characterized in that, In multi-layer co-extrusion, the disturbance observer identifies the target extruder that causes the finished product thickness deviation and feeds back the online layer thickness detection data to the frequency converter of the target extruder. For the thickness fluctuation of ultra-thin films, a predictive control algorithm based on a multi-dimensional coupled mapping model is used to predict the impact of damper adjustment on the subsequent finished product thickness based on the film linear velocity.

10. The method according to claim 1, characterized in that, The collaborative control method also includes big data analysis and storage optimization processing, which utilizes production records stored on a cloud server to extract the optimal combination of process parameters for a specific raw material grade through a big data prediction and analysis module.

Images