Degradable express bag blowing film thickness intelligent control method and system
By using non-Newtonian flow decoupling calculations and mutual compensation of thermodynamic energy cost models, the problem of film bubble tearing caused by thermal period control interference during the blown film production of biodegradable express bags was solved, thereby improving the stability and thickness accuracy of film production.
Patent Information
- Application Number
- CN202611113053.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-25
- Publication Date
- 2026-11-10
Smart Images

Figure CN123060902A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent manufacturing technology for polymer films, and in particular to an intelligent control method and system for the thickness of blown film for biodegradable express delivery bags. Background Technology
[0002] In the continuous blown film production process of fully biodegradable express delivery bags, the production line utilizes a midstream cooling air ring and a downstream traction device to control the film bubble formation and film thickness. Existing thickness closed-loop control systems mostly employ an industrial control architecture with independent feedback from multiple actuators. When dealing with biodegradable polymer raw materials exhibiting low melt strength and high thermosensitivity, the material releases latent heat of crystallization at the solid-liquid phase transition boundary. Existing independent feedback control systems do not incorporate thermodynamic hardware interlocking constraints into their underlying logic. When the circumferential region of the film bubble is in a thermosensitive and vulnerable period of intense exothermic crystallization, existing control systems forcibly apply excessive pneumatic cooling air compensation to the vulnerable region to correct local thickness or height deviations. The abrupt change in airflow kinetic energy and the mechanical traction tensile stress blindly superimpose at the phase boundary plane, triggering severe cross-physical domain control interference, completely disrupting the rheological stress balance of the biodegradable film bubble, ultimately causing cold oscillations and even large-scale physical tearing of the film. Summary of the Invention
[0003] To overcome the above shortcomings, this invention provides a method and system for intelligent control of the thickness of blown film for biodegradable express delivery bags. It aims to improve the problem that traditional blown film control mostly relies on independent feedback from the actuator, which causes the biodegradable film bubble to be easily torn by cold oscillation because it does not avoid control interference during the heat-sensitive period of crystallization.
[0004] In a first aspect, the present invention provides the following technical solution: a method for intelligent control of the thickness of blown film for biodegradable express delivery bags, comprising the following steps: S1. Obtain the real-time melt temperature and real-time internal pressure of the die head, perform non-Newtonian flow decoupling calculation based on the real-time melt temperature and real-time internal pressure of the die head, and output the reference air valve control voltage and reference traction line speed. S2. Acquire infrared thermal distribution image data, and based on the infrared thermal distribution image data, extract the actual frost line height and the axial temperature gradient of each circumferential sector at the actual frost line height. S3. Calculate the displacement error based on the actual frost line height and the target frost line height, input the displacement error and the axial temperature gradient into the thermodynamic energy cost model to solve, and output the thermodynamic cooperative allocation factor. S4. Based on the thermodynamic cooperative distribution factor and the complementary value of the thermodynamic cooperative distribution factor, perform mutual exclusion compensation calculation on the reference air valve control voltage and the reference traction line speed, and generate the actual air valve control voltage and the delayed traction line speed respectively by combining the stress transmission delay constant. Use the actual air valve control voltage to drive the local proportional valve, and use the delayed traction line speed to drive the traction device. S5. Obtain thickness deviation data, generate an updated target frost line height based on the thickness deviation data, and feed the updated target frost line height back into the thermodynamic energy cost model.
[0005] By adopting the above technical solution, the axial temperature gradient is introduced into the cost model to calculate the allocation factor and execute the mutual exclusion compensation of the air valve and traction, thereby realizing the physical isolation of cooling and stretching. This improves the problem that traditional blown film control mostly uses independent feedback of the actuator, which does not avoid the control interference during the crystallization heat sensitive period, resulting in the biodegradable film bubble being easily torn by cold oscillation.
[0006] Secondly, the present invention provides the following technical solution: an intelligent control system for the thickness of blown film in a biodegradable express delivery bag, comprising the following modules: The rheological decoupling control module is used to acquire the real-time melt temperature and real-time internal pressure of the die head, perform non-Newtonian rheological decoupling calculations based on the real-time melt temperature and real-time internal pressure of the die head, and output the reference air valve control voltage and reference traction linear velocity. The image feature extraction module is used to acquire infrared thermal distribution image data, and based on the infrared thermal distribution image data, extract the actual frost line height and the axial temperature gradient of each circumferential sector at the actual frost line height. The cost model optimization module is used to calculate the displacement error based on the actual frost line height and the target frost line height, input the displacement error and the axial temperature gradient into the thermodynamic energy cost model for solution, and output the thermodynamic cooperative allocation factor. The mutual exclusion compensation and drive module is used to perform mutual exclusion compensation calculations on the reference air valve control voltage and the reference traction line speed based on the complementary values of the thermodynamic cooperative distribution factor and the thermodynamic cooperative distribution factor. It generates the actual air valve control voltage and the delayed traction line speed by combining the stress transmission delay constant. The actual air valve control voltage is used to drive the local proportional valve, and the delayed traction line speed is used to drive the traction device. The macroscopic deviation feedback module is used to acquire thickness deviation data, generate an updated target frost line height based on the thickness deviation data, and feed the updated target frost line height back into the thermodynamic energy cost model.
[0007] The present invention has the following beneficial effects: 1. In this invention, by introducing the axial temperature gradient into the cost model to calculate the allocation factor, the mutual exclusion compensation between the air valve and the traction is achieved, thereby realizing the physical isolation between cooling and stretching. This improves the problem that traditional blown film control mostly uses independent feedback of the actuator, which, due to the failure to avoid control interference during the crystallization heat sensitive period, causes the biodegradable film bubble to be easily torn by cold oscillation.
[0008] 2. In this invention, non-Newtonian rheological decoupling calculations are performed based on the real-time melt temperature and internal pressure of the die head, thereby outputting a reference control quantity to smooth out source fluctuations. This improves the problem that traditional extrusion control mostly uses constant speed operation, which fails to compensate for the drastic fluctuations in melt viscosity, resulting in unstable extrusion volume and excessive overall film thickness.
[0009] 3. In this invention, by combining the stress transmission delay constant to generate a delayed traction linear velocity driving traction device, the phase misalignment of the action caused by the spatial span is eliminated, thereby improving the problem that traditional traction control mostly adopts instantaneous synchronous speed regulation, which causes the traction action to lead to membrane tensile resonance due to ignoring the inherent physical lag of tensile stress transmission.
[0010] 4. In this invention, by updating the target frost line height based on thickness deviation data and feeding it back to the cost model, a macroscopic repair mechanism to eliminate long-term drift is established, thereby improving the problem that traditional membrane bubble control mostly adopts fixed height static tracking, which, due to being detached from the guidance of terminal film formation quality, causes the equipment to experience long-term thickness trend deviations. Attached Figure Description
[0011] Figure 1 This is a flowchart of an intelligent control method for blown film thickness of biodegradable express bags proposed in this invention; Figure 2 This is a flowchart of the non-Newtonian flow-decoupled computation parallel process for an intelligent control method of blown film thickness for biodegradable express bags proposed in this invention. Figure 3 This is a flowchart of the thermodynamic energy cost model optimization process for an intelligent control method for blown film thickness of biodegradable express bags proposed in this invention. Figure 4 This is a flowchart illustrating the mutual exclusion compensation and hardware interlocking drive of a method for intelligent control of blown film thickness in biodegradable express bags proposed in this invention. Figure 5 This is an architecture diagram of an intelligent control system for the thickness of blown film in a biodegradable express delivery bag, as proposed in this invention. Detailed Implementation
[0012] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0013] Example 1: In the first embodiment of the present invention, the present invention provides a method for intelligent control of the blown film thickness of biodegradable express delivery bags, such as... Figures 1-2 As shown, it includes the following steps: S1. Obtain the real-time melt temperature and real-time internal pressure of the die head, perform non-Newtonian flow decoupling calculation based on the real-time melt temperature and real-time internal pressure of the die head, and output the reference air valve control voltage and reference traction line speed. Furthermore, in S1, the real-time melt temperature and real-time internal pressure of the die head are obtained. Based on the real-time melt temperature and real-time internal pressure of the die head, non-Newtonian flow decoupling calculations are performed, and the reference air valve control voltage and reference traction linear velocity are output, including: A physical state sampling action is performed using a high-frequency thermocouple and a high-frequency melt pressure sensor to obtain the real-time melt temperature and real-time internal pressure of the die head, respectively. Perform time series alignment and filtering on the real-time melt temperature and real-time internal pressure of the die head; Calculate the pressure difference between the real-time internal pressure of the mold head and the preset target internal pressure of the mold head, and multiply the pressure difference with the preset pressure and speed coupling gain coefficient to generate the first intermediate product; The numerical quotient is obtained by dividing the preset viscous flow activation temperature constant by the real-time melt temperature at the die head; Take the natural logarithm of the negative logarithm of the quotient to generate the second intermediate product; Multiply the first intermediate product by the second intermediate product to generate the pressure pulse speed adjustment amount; The reference screw speed is generated by subtracting the pressure pulse speed adjustment amount from the rated operating speed. Multiply the reference screw speed by the aerodynamic mapping constant and the motion mapping constant respectively to generate the reference air valve control voltage and the reference traction linear velocity.
[0014] Specifically, the control system directly deploys high-frequency thermocouples and high-frequency melt pressure sensors at the end of the extruder die channel. The biodegradable melt directly contacts the sensor probes exposed inside the channel during extrusion. The high-frequency thermocouples and melt pressure sensors perform physical state sampling according to a set high-frequency sampling clock cycle, acquiring analog temperature and pressure signals respectively. The control system's underlying data acquisition board receives the analog temperature and pressure signals. The board's built-in analog-to-digital converter performs discretization sampling and quantization encoding on the input signals.
[0015] After completing the analog-to-digital conversion, the system generates digitized real-time melt temperature and real-time internal pressure of the die head. The rheological decoupling control module retrieves the real-time melt temperature and real-time internal pressure of the die head stored in the memory register. Due to physical differences in the hardware sampling frequencies and bus communication protocols of the two types of sensors, the controller performs time series alignment on the input variables based on the timestamps embedded in the data packets.
[0016] The dual-channel data strings, after timing matching, are imported into a first-order low-pass filter algorithm structure for filtering. This filters out power frequency interference and bus glitches caused by the servo motor's operation, outputting smoothed, clean state variables. The control system establishes a multi-rate collaborative underlying task timing. The underlying algorithm unit performs a subtraction operation to calculate the pressure difference between the smoothed real-time die head internal pressure and the pre-programmed target die head internal pressure. The target die head internal pressure is not set as a fixed value; the system uses a lookup table established from previous steady-state extrusion process experiments to match and extract the target die head internal pressure based on the currently produced biodegradable material grade and membrane bubble diameter specifications.
[0017] The system multiplies the pressure difference by a preset pressure-speed coupling gain coefficient, outputting the first intermediate product. An independent processing thread retrieves the viscous flow activation temperature constant of the corresponding biodegradable raw material from the material property database. The preset viscous flow activation temperature constant is calculated based on multi-stage isothermal shear test data of the corresponding material in a capillary rheometer, characterizing the sensitivity level of the material's melt index to temperature fluctuations. The system uses the preset viscous flow activation temperature constant divided by the smoothed real-time melt temperature of the die to obtain a dimensionless numerical quotient. The processing unit calls the underlying mathematical function library to directly calculate the exponent of the natural logarithm of the negative quotient, outputting the second intermediate product characterizing the non-Newtonian rheological decay characteristics of the melt. The control bus performs a hard-wired multiplication operation on the first and second intermediate products, outputting the pressure pulse speed adjustment amount. The system's built-in process formula specifies the rated operating speed. The rheological decoupling control module subtracts the pressure pulse speed adjustment amount from the rated operating speed to generate a reference screw speed that resists viscous flow state variations. The rheological decoupling control module multiplies the reference screw speed by the aerodynamic mapping constant and the motion mapping constant, respectively, to generate the reference air valve control voltage and the reference traction linear velocity.
[0018] The underlying mathematical model for the decoupled computation of non-Newtonian flow and the full-line reference mapping is represented by the following set of equations: ; ; ; in This represents the reference screw speed generated after rheological decoupling correction. This indicates the rated operating speed set in the extruder frequency converter for the current product specifications. The pressure-speed coupling gain coefficient represents the relationship between pressure and spindle speed. This indicates the real-time internal pressure of the die head after alignment and filtering. This indicates the target internal pressure of the die head that the process formulation requires to be maintained. It represents the viscous flow activation temperature constant, which is specifically used to characterize the thermo-rheological properties of a particular biodegradable material. This indicates the real-time melt temperature of the die head after alignment and filtering. This represents the reference valve control voltage generated based on the rotational speed and flow rate mapping. This represents the aerodynamic mapping constant corresponding to a unit extrusion speed. This represents the baseline traction linear velocity generated based on the rotational speed-flow mapping. This represents the traction motion mapping constant corresponding to a unit extrusion speed. The calculated reference air valve control voltage and reference traction linear velocity are pushed to the input port buffer queue of the mutual exclusion compensation and drive module.
[0019] S2. Obtain infrared thermal distribution image data, and based on the infrared thermal distribution image data, extract the actual frost line height and the axial temperature gradient of each circumferential sector at the actual frost line height. Furthermore, in S2, infrared thermal distribution image data is acquired. Based on the infrared thermal distribution image data, the actual frost line height and the axial temperature gradient of each circumferential sector at the actual frost line height are extracted, including: The infrared thermal imaging sensor is used to perform a panoramic scan in the infrared band to obtain the raw thermal radiation data matrix. The background thermal noise of the raw thermal radiation data matrix is filtered out using the workshop ambient temperature reference data to generate infrared thermal distribution image data. Image spatial segmentation processing is performed on the infrared thermal distribution image data to extract the solid-liquid phase transition boundary contour on the surface of the membrane bubble; The spatial mean is calculated based on the height coordinate data of the solid-liquid phase transition boundary profile on the vertical spatial axis to generate the actual frost line height. Divide the membrane bubble into a predetermined number of circumferential sectors along the circumferential direction, and map the solid-liquid phase transition boundary contour into each circumferential sector. At the vertical spatial position corresponding to the actual frost line height, the temperature values inside each circumferential sector are extracted, and the temperature values are subjected to first-order spatial differentiation along the axial spatial coordinate axis to generate the axial temperature gradient.
[0020] Specifically, the system activates the infrared thermal imaging sensor located midway through the production line. The infrared thermal imaging sensor is mounted above the cooling fan ring at a fixed pitch angle and object distance. The sensor performs a panoramic infrared scan of the outer wall of the membrane bubble undergoing the blowing and cooling process. The sensor's focal plane array receives the mixed thermal radiation energy emitted from the membrane bubble surface and its surrounding environment, outputting a raw thermal radiation data matrix. The control system reads the workshop ambient temperature reference data. The image feature extraction module combines the workshop ambient temperature reference data with background thermal noise filtering processing of the raw thermal radiation data matrix. The background thermal noise filtering processing follows the following underlying mathematical subtraction model: ; in This represents the points in the two-dimensional space of the image after filtering out background thermal noise. OK Infrared thermal distribution image data generated at the column. This represents the radiation intensity value within the original thermal radiation data matrix. The equivalent infrared emissivity constant represents the background environment of the workshop. This represents the Stefan Boltzmann constant. This represents the baseline data for the workshop ambient temperature.
[0021] The visual processing unit within the image feature extraction module invokes a threshold segmentation algorithm to perform image spatial segmentation processing on the input infrared thermal distribution image data. The processor marks pixel regions with temperatures above the threshold as viscous liquid phases and pixel regions with temperatures below the threshold as highly elastic solid phases. An edge detection operator traverses the binarization matrix to extract the solid-liquid phase transition boundary contours on the membrane bubble surface. The computation unit reads all height coordinate data of the solid-liquid phase transition boundary contours on the vertical spatial axis. The microprocessor performs mathematical summation and arithmetic mean processing on the discrete height coordinate data to generate the actual frost line height. The physical conversion of the actual frost line height follows the following discrete mean mathematical model: ; in This represents the actual height scalar of the frost-white line output after arithmetic averaging. This represents the total number of pixels contained in the solid-liquid phase transition boundary contour extracted by the edge detection operator. Indicates the first The physical measurement value of the height of each pixel in the absolute vertical spatial coordinate system.
[0022] An infrared thermal imaging sensor is mounted above the cooling fan ring at a fixed pitch angle and object distance. The visual calibration program extracts the feature system of the calibration board, establishes a scale conversion matrix from two-dimensional pixel coordinates to three-dimensional physical space coordinates, and simultaneously loads a distortion correction algorithm based on lens surface features to perform linearization correction on the original pixel coordinates. The conversion matrix maps pixel heights to actual spatial dimensions with physical dimensions. The image feature extraction module, based on the interior of the cooling fan ring, divides the ring into a predetermined number of circumferential sectors along the circumference of the membrane bubble, according to the absolute physical arrangement of the local proportional valves within the cooling fan ring. The module performs spatial geometric projection mapping to accurately place the solid-liquid phase transition boundary contour into each circumferential sector. For each divided circumferential sector, the data acquisition thread extracts the temperature values of adjacent pixels in the vertical direction at the vertical spatial position corresponding to the actual frost line height. The digital signal processor calls a first-order spatial difference algorithm to perform first-order spatial differentiation on the temperature values along the axial spatial coordinate axis, generating an axial temperature gradient. The axial temperature gradient is extracted based on the following partial derivative model: ; in This indicates the first partition of the system. The axial temperature gradient is calculated independently for each circumferential sector. Indicates the first The internal temperature values of each circumferential sector are along the axial spatial coordinate axis. The discrete temperature field distribution function constitutes this. This represents the independent variable of the absolute vertical spatial coordinate axis set by the system. The spatial scalar representing the actual frost-white line height output from the pre-calculation serves as the absolute height constraint for the partial derivative operation here. The image feature extraction module generates a feature data package containing the actual frost-white line height and temperature gradients along each axis, and pushes the feature data package to the cost model optimization module.
[0023] It is important to note that while traditional blown film control systems utilize infrared thermal imaging sensors to obtain the frost line height, they merely treat the actual frost line height as a macroscopic, one-dimensional geometric scalar, directly inputting it into a single proportional-integral-derivative feedback loop to adjust the main fan speed. This completely severs the physical coupling between the frost line height and the local crystallization dynamics of the polymer material. In stark contrast, the control method in this application does not involve isolated closed-loop tracking of the actual frost line height. This application innovatively uses the actual frost line height as a height anchor point in a three-dimensional spatial coordinate system, accurately extracting the axial temperature gradient of each circumferential sector at that height. In subsequent applications, this application does not use the axial temperature gradient as a direct control driver, but rather deeply embeds it into the thermodynamic energy cost model as a thermodynamic penalty variable that triggers a physical interlocking mechanism. By anchoring the actual frost line height in three-dimensional space with the axial temperature gradient and using the axial temperature gradient as a constraint penalty term in the energy cost model, the control domain interference defect caused by isolated extraction of the frost line height and simple aerodynamic feedback in the prior art is improved, and the problem of physical tearing of the membrane bubble caused by excessive cold wind compensation during the thermosensitive vulnerable period of the biodegradable material is avoided.
[0024] like Figure 3 As shown, S3 calculates the displacement error based on the actual frost line height and the target frost line height, inputs the displacement error and axial temperature gradient into the thermodynamic energy cost model for solution, and outputs the thermodynamic cooperative allocation factor. Furthermore, in S3, the displacement error and axial temperature gradient are input into the thermodynamic energy cost model for solution, and the thermodynamic cooperative allocation factor is output, including: Construct a system total energy cost loss function that includes displacement cost penalty term, aerodynamic cooling compensation work penalty term and traction motion phase shift work penalty term; The aerodynamic cooling compensation work penalty term is set as an exponential positive correlation penalty function with the axial temperature gradient as the independent variable; The first-order partial derivative of the total energy cost loss function of the system is calculated with respect to the thermodynamic cooperative allocation factor variable to be solved, and the partial derivative equation is obtained. Set the calculated result of the partial derivative equation to zero, solve the algebraic solution of the extreme point of the partial derivative equation in the continuous real number domain, and extract the algebraic solution of the extreme point as the thermodynamic co-distribution factor.
[0025] In S3, the aerodynamic cooling compensation work penalty term is set as an exponential positive correlation penalty function with the axial temperature gradient as the independent variable, including: The axial temperature gradient is multiplied by a preset latent heat of crystallization buffer constant to generate an exponential independent variable value. The thermodynamic amplification factor is generated by raising the natural logarithm of the independent exponential variable to the power of the natural logarithm. Multiply the preset basic aerodynamic weight by the thermodynamic amplification factor to generate the thermal suppression penalty weight; The thermal suppression penalty weight is multiplied by the square of the thermodynamic co-allocation factor to be solved to generate the aerodynamic cooling compensation work penalty term.
[0026] Specifically, the cost model minimization optimization module reads the actual frost-white line height extracted by the preceding unit from the underlying bus interface. The system synchronously retrieves the target frost-white line height parameter set on the operation panel. The processor performs a subtraction operation on the two height data points, using the actual frost-white line height minus the target frost-white line height to calculate and generate the displacement error.
[0027] The calculation process for the aforementioned height displacement error is defined using the following mathematical equation: ; in This represents the displacement error generated after scalar subtraction. This represents the actual frost line height scalar input from the front-end vision processing node. This indicates the value of the target frost line height command that the system is currently executing.
[0028] Displacement error generated by arithmetic unit It is temporarily stored in memory, waiting to be called in the subsequent assembly process of the extreme value function.
[0029] The optimization module's other input data bus continuously receives multiple independent axial temperature gradients concurrently transmitted from each circumferential sector. In terms of physical execution logic, the more intense the exothermic crystallization in a region, the larger its corresponding axial temperature gradient value. For the discrete gradient vectors input from multiple sectors, the system executes a maximum pooling fusion strategy. The processor iterates through all local axial temperature gradient values within the current cycle, extracting the axial temperature gradient with the largest absolute value as the sole independent variable representing the thermodynamic extremum of the entire bubble. This fusion strategy relies on a conservative safety design principle dominated by the most dangerous bubble-breaking region, ensuring that global energy allocation prioritizes responses to the extreme sectors facing the risk of tearing. The algorithm core retrieves the pre-programmed latent heat buffer constant for crystallization and multiplies the extracted most dangerous axial temperature gradient with this latent heat buffer constant. The processor outputs an exponential independent variable value as input to the operator. The mathematical operation unit calculates the natural logarithm of this exponential independent variable value to generate a thermodynamic amplification factor. The arithmetic unit multiplies the system's fixed basic aerodynamic weights with the extracted thermodynamic amplification factor to generate a dynamically changing thermosensitive suppression penalty weight.
[0030] The exponential mapping rule for calculating the thermal inhibition penalty weights mentioned above is defined by the following mathematical expression: ; in This represents the calculated thermal inhibition penalty weight variable. This represents the basic aerodynamic weight constant based on the factory mechanical power calibration of the wind ring fan. This represents the latent heat of crystallization buffer constant specific to polymers. This represents the local axial temperature gradient scalar extracted by image recognition.
[0031] After generating dynamic weight variables, the algorithm enters the assembly stage of the total cost function. The system multiplies the displacement error tolerance constant with the square of the displacement error calculated in the previous step to construct a displacement cost penalty term. The arithmetic unit multiplies the thermal suppression penalty weight obtained from the previous step with the square of the thermodynamic cooperative allocation factor to be solved to construct a pneumatic cooling compensation work penalty term. The digital logic circuit subtracts the thermodynamic cooperative allocation factor to be solved from the numerical value to obtain the mutual exclusion complement of the allocation factor. The algorithm multiplies the set mechanical tension base weight with the square of the mutual exclusion complement of the allocation factor to construct a traction motion phase shift work penalty term.
[0032] The processor performs algebraic addition on the calculated displacement cost penalty, aerodynamic cooling compensation work penalty, and traction motion phase shift work penalty to construct a system total energy cost loss function describing the macroscopic power dissipation of the system. The mathematical form of this cost function is defined as follows: ; in This represents the total energy cost loss function value of the completed system. This represents the displacement error tolerance constant set by the controller for the liquid-solid phase boundary offset. This represents the displacement error scalar obtained through scalar subtraction. This represents the thermal inhibition penalty weight carrying the exponential scaling characteristic. This represents the thermodynamic co-allocation factor variable to be solved in the continuous real number domain. This represents the mechanical tensile basis weight set to overcome the tensile yield strength of the film.
[0033] The cost model minimization optimization module loads the compiled system total energy cost loss function into the calculus processor. The algorithm performs first-order partial derivative calculations on the only control variable in the function, namely the thermodynamic cooperative allocation factor to be solved. The system removes constant terms to generate partial derivative equations in a first-order algebraic form. The controller forces the substitution of the calculation results equal to zero into the partial derivative equations. Since the system total energy cost loss function is a convex function with respect to the cooperative allocation factor, the arithmetic unit does not need to perform complex iterative optimization, but directly outputs the algebraic analytical solution of the thermodynamic cooperative allocation factor.
[0034] The underlying mathematical model of the algebraic analytical solution output by the optimization algorithm is described as follows: ; in This indicates that the optimization module directly calculates the extracted target thermodynamic co-allocation factor scalar based on the partial derivative equation. This represents a pre-defined mechanical tensile basis weight constant that overcomes the tensile yield strength of the film. This represents the basic aerodynamic weight constant based on the factory mechanical power calibration of the wind ring fan. This represents the latent heat of crystallization buffer constant specific to polymers. This represents the local axial temperature gradient scalar extracted by image recognition.
[0035] The optimization module calls the aforementioned algebraic analytical solution model within each control cycle to directly calculate and extract the target thermodynamic cooperative allocation factor. This thermodynamic cooperative allocation factor is output to the read port of the mutual exclusion compensation and delay alignment module via high-speed shared memory, serving as the allocation benchmark for the underlying servo device and pneumatic proportional valve to perform correction work.
[0036] like Figure 4 As shown, S4, based on the complementary values of the thermodynamic cooperative distribution factor and the thermodynamic cooperative distribution factor, performs mutual exclusion compensation calculation on the reference air valve control voltage and the reference traction line speed, and generates the actual air valve control voltage and the delayed traction line speed by combining the stress transmission delay constant. The actual air valve control voltage is used to drive the local proportional valve, and the delayed traction line speed is used to drive the traction device. Furthermore, in S4, based on the complementary values of the thermodynamic co-allocation factor and the thermodynamic co-allocation factor, mutual exclusion compensation calculations are performed on the reference valve control voltage and the reference traction linear velocity. Combined with the stress transmission delay constant, the actual valve control voltage and the delayed traction linear velocity are generated, including: By subtracting the thermodynamic co-allocation factor from the numerical value, a complementary value of the thermodynamic co-allocation factor is generated. The thermodynamic co-allocation factor is multiplied by the preset aerodynamic coupling compensation gain coefficient to generate the first correction coefficient; Multiply the first correction factor by the displacement error to generate the aerodynamic voltage compensation increment; The reference valve control voltage is added to the pneumatic voltage compensation increment, and voltage amplitude saturation limiting is performed to generate the actual valve control voltage.
[0037] In S4, based on the complementary values of the thermodynamic co-allocation factor and the thermodynamic co-allocation factor, mutual exclusion compensation calculations are performed on the reference damper control voltage and the reference traction linear velocity. The actual damper control voltage and delayed traction linear velocity are generated by combining the stress transmission delay constant. This also includes: The complementary value of the thermodynamic co-allocation factor is multiplied by the preset kinematic follow-up compensation coefficient to generate the second correction coefficient; Perform a first-order difference derivative operation on the actual height of the frost line with respect to the time dimension to generate the vertical drift velocity value of the frost line. Multiply the second correction factor by the vertical drift velocity of the frost-white line to generate the traction speed compensation reduction; Obtain the pre-stored physical space span length of the membrane bubble, and calculate the stress transmission delay constant by dividing the physical space span length of the membrane bubble by the reference traction linear velocity. The reference traction line speed is subjected to a time delay processing based on the stress transmission delay constant, and the traction speed compensation reduction is subtracted after the delay processing to generate the delayed traction line speed.
[0038] In S4, the local proportional valve is driven by the actual air valve control voltage, and the traction device is driven by the delayed traction linear velocity, including: For each circumferential sector, the axial temperature gradient is compared with the preset critical fracture temperature gradient threshold in real time. In response to a hardware status flag indicating that the axial temperature gradient is greater than the critical temperature gradient threshold for fracture, a hardware interlock interrupt instruction is triggered. Based on the hardware interlock interrupt instruction, the compensation increment of the actual air valve control voltage is forcibly cleared to zero, the reference voltage is output to the local proportional valve, and the delayed traction linear speed is simultaneously sent to the traction device to perform the tensioning action.
[0039] Specifically, the mutual exclusion compensation and delay alignment module reads the thermodynamic cooperative allocation factor output by the cost model minimization optimization module from shared memory. The data receiving port of this module continuously receives the reference air valve control voltage and reference traction linear velocity from the rheodynamic decoupling module. The computational core internally divides the computation into two independent threads: pneumatic computation and servo motion computation.
[0040] The pneumatic calculation thread's built-in subtractor subtracts the read thermodynamic co-allocation factor from the numerical value to generate a complementary value for the thermodynamic co-allocation factor. The multiplier multiplies the thermodynamic co-allocation factor with the preset aerodynamic coupling compensation gain coefficient in memory to output the first correction coefficient. The system retrieves the displacement error from the previous calculation cache. The first correction coefficient and the displacement error are multiplied in the hardware multiplier to generate the aerodynamic voltage compensation increment. The adder adds the reference damper control voltage to the aerodynamic voltage compensation increment. The control board reads the physical range boundary of the underlying digital-to-analog converter chip and constructs a nonlinear truncation operator based on this electrical hardware boundary. The microprocessor unit uses this truncation operator to perform voltage amplitude saturation upper and lower limit processing on the added level signal, stripping out the out-of-bounds pulses that could cause electrical overload, and outputting the actual damper control voltage that conforms to the industrial equipment safety execution specifications.
[0041] The mathematical model of the control law for the aforementioned aero-voltage compensated track is expressed as follows: ; in This represents the actual valve control voltage output to the bus after amplitude limiting. This represents the algebraic truncation operator that performs voltage amplitude saturation limiting. This indicates the reference valve control voltage obtained from the underlying bus. This represents the scalar value of the thermodynamic co-allocation factor read. This indicates the hardware aerodynamic coupling compensation gain coefficient that has been retrieved. This indicates the pre-calculated displacement error of the memory block.
[0042] The servo motion calculation thread multiplies the complementary value of the extracted thermodynamic co-allocation factor with the preset kinematic following compensation coefficient to output the second correction coefficient. The calculus processor extracts the actual frost line height matrix of the current period and the previous sampling period. The system performs a first-order difference derivative operation in the time dimension for this height sequence to extract the vertical drift velocity value of the frost line at the phase interface. The multiplication unit multiplies the second correction coefficient with the vertical drift velocity value of the frost line to output the traction speed compensation reduction.
[0043] The motion controller reads the pre-stored physical span length of the membrane bubble in the system configuration table. This length represents the physical geometric distance between the cooling air ring nozzle and the top traction roller. The divider calculates the stress transmission delay constant by dividing the membrane bubble physical span length by the reference traction line speed. A data queue register constructs a first-in-first-out data queue for storing historical speeds. The system determines the physical storage depth of this data queue based on the quotient of the maximum theoretical span of the production line and the minimum design line speed. During the cold start phase of the system's initial power-on, the control bus pre-fills all storage units in this queue with the current reference traction line speed as the initial filling state, avoiding severe stalling and shaking of the traction motor due to the retrieval of empty queue data at startup. In steady-state operation, the queue mechanism strictly applies physical delay processing to the input reference traction line speed according to the currently calculated stress transmission delay constant. The arithmetic logic unit subtracts the extracted traction speed compensation reduction from the historical reference traction line speed after the wake-up delay processing. The arithmetic unit finally generates the delayed traction line speed.
[0044] The aforementioned mathematical model for compensation and time delay of the servo traction track is expressed as follows: ; ; in This indicates the delayed traction line speed generated by the current time period calculation. This represents the historical baseline traction line speed that was awakened from the queue buffer after a time delay. This represents the stress transmission delay constant of a solid-state thin film obtained through division. This represents the scalar value of the thermodynamic co-allocation factor read. This represents the kinematic following compensation coefficient written into the servo drive configuration parameter table. This represents the vertical drift velocity of the frost-white line obtained by first-order differentiation. This indicates the length of the pre-stored physical space span of the membrane bubble. This represents the current cycle reference traction line speed, which is used as the divisor input for time calculation.
[0045] After completing the dual-track parallel operation, the watchdog program inside the mutual exclusion compensation and drive module intercepts the axial temperature gradient corresponding to each circumferential sector in the feature matrix. A fracture critical temperature gradient threshold, representing the physical limit of melt fracture for a specific biodegradable material, is pre-written into the system flash memory. This threshold is not a fixed wildcard value, but is calculated based on non-isothermal tensile fracture experimental data of the corresponding polymer material in a thermomechanical analyzer. The comparator circuit establishes real-time independent polling logic for each circumferential sector, comparing the axial temperature gradient with the fracture critical temperature gradient threshold in real time. If the comparison mechanism determines that the axial temperature gradient of a specific circumferential sector is greater than the fracture critical temperature gradient threshold, a hardware interlock interrupt instruction is triggered. The hardware interlock interrupt and control quantity allocation logic follows the following piecewise control function model: ; in This indicates that the underlying control bus ultimately communicates with the first... The actual physical drive level of the local proportional valve corresponding to each circumferential sector. This represents the reference voltage value without any pneumatic voltage compensation increment. Indicates that for the first The axial temperature gradient is extracted in real time from each circumferential sector. This represents the critical temperature gradient threshold constant for writing to the system's flash memory. This represents the actual valve control voltage generated by the preceding mutual exclusion compensation module, which includes the feedforward compensation increment.
[0046] The underlying control bus, based on a piecewise control function model, cuts off the compensation update channel for the corresponding local proportional valve in the affected circumferential sector. An electrical isolation mechanism ensures that the remaining circumferential sectors, which have not triggered alarms, continue to receive the actual valve control voltage, including dynamic feedforward increments. The microcontroller unit executes a local register reset, forcibly clearing the compensation increment portion within the actual valve control voltage allocated to the alarm-triggered area. The control board's digital-to-analog converter chip locks the reference voltage and sends it to the local proportional valve. The control bus synchronously sends the delayed traction line speed to the servo driver of the traction device. The servo motor drives the traction roller to perform a physical stretching action. The film completes stretching and correction in a solid-state environment without additional cold airflow interference. The execution phase achieves a physical layer control closed loop that physically shields against airflow interference and switches to single-point stretching response during periods of extreme thermal sensitivity.
[0047] S5. Obtain thickness deviation data, generate an updated target frost line height based on the thickness deviation data, and feed the updated target frost line height back into the thermodynamic energy cost model; Furthermore, in S5, an updated target frost line height is generated based on the thickness deviation data, including: The thickness deviation data is processed by moving average filtering to remove high-frequency measurement noise. The filtered thickness deviation data is input into the discrete proportional-integral feedback controller, and the error integral accumulation calculation is performed within the set time period to generate the height correction bias. The height correction bias is summed with the current time-period baseline value of the target frost line height to generate the updated target frost line height.
[0048] Specifically, the macroscopic thickness deviation feedback module reads array measurement signals from the capacitive thickness detector deployed on the rotary scanning frame behind the traction roller. The underlying calculation unit executes an inverse kinematic mapping algorithm based on the current mechanical deflection angle parameter of the rotary scanning probe and subtracts the dwell time of the film traction rise. The system reconstructs the thickness deviation data from the one-dimensional sequence thickness measurement and maps it one-to-one to each circumferential sector physically divided by the lower air ring. Since the mechanical vibration of the traction roller and probe movement introduce transient jump interference, the controller routes the reconstructed thickness deviation data to the data cleaning thread. The moving average filter built into the cleaning thread performs data sequence translation and mean calculation logic within a set window length. The preset moving average filter window length constant strictly matches the mechanical detection cycle of a single complete rotation of the thickness scanning probe. The system uses the translation and mean calculation operation of the matching cycle to remove probe eccentricity and high-frequency environmental measurement noise from the original thickness data.
[0049] The filtered, purified thickness deviation data is fed to a discrete proportional-integral feedback controller. The proportional unit within the controller performs linear amplification calculations on the input data. The integral unit establishes an accumulator with a macroscopic time period as its boundary. The accumulator performs error integration and accumulation on the historical input errors within the time period. The adder collects the output values from the proportional and integral operations, performs a combined calculation, and outputs a height correction bias to compensate for steady-state drift.
[0050] The optimization model's data registers are granted read access to the macroscopic feedback module. The microprocessor extracts the current time-period baseline value of the target frost-white line height. The arithmetic logic unit adds the extracted height correction bias to the current time-period baseline value of the target frost-white line height. The system overwrites the old data and generates the updated target frost-white line height.
[0051] The above outer-loop filtering and feedback correction logic is mathematically defined based on the following discrete state-space equations: ; ; ; in Indicates the first The pure thickness deviation data output after each sampling period is processed by moving average filtering. This represents the default length constant of the moving average filter window preset by the controller. This represents the original thickness deviation data obtained by the bottom detection loop in the historical acquisition sequence. This represents the height correction bias of the output calculated by the discrete proportional-integral feedback controller. This represents the proportional gain coefficient of the outer loop controller. This represents the discrete integral gain coefficient of the outer loop controller. This represents the updated target frost line height, generated through feedback calculations and applied in the next time period. This represents the current time period base value of the target frost line height that the optimization module is currently executing.
[0052] The macroscopic thickness deviation feedback module writes the updated target frosting line height to the designated register address of the module to which the thermodynamic energy cost model belongs via the direct memory access mechanism of the system's internal industrial bus. The minimization optimization model automatically calls the updated target parameters to calculate the displacement error in the next computation cycle. The underlying algorithm thus completes a global closed loop, tracing from the final film formation quality back to the midstream thermodynamic control parameter constraints.
[0053] Example 2: In the large-scale continuous blown film production of fully biodegradable express delivery bags, the production line relies on the extruder die head to melt the material, which undergoes phase change after being blown by the midstream variable frequency cooling air ring, and is then pulled into rolls by the top servo device. However, due to the non-Newtonian rheological properties of biodegradable materials, which have extremely low melt strength and extremely high thermal sensitivity, actual production faces severe problems of underlying control interference and physical film breakage. External temperature disturbances can easily induce rheological pulses inside the die head, leading to instability in the source extrusion volume. The material releases latent heat of crystallization violently at the cooling phase change boundary, resulting in an extremely uneven local axial temperature gradient along the circumference of the film bubble. If the control system uses independent feedback correction from multiple actuators, and forcibly superimposes pneumatic cooling air compensation in the extremely sensitive and fragile area of crystallization exotherm, it will instantly induce cold oscillation and physical tearing of the film bubble. There is a physical spatial span of several meters between the midstream air ring and the downstream traction device. The stress wave transmission of mechanical tensile compensation has an inherent pure time delay. Blindly superimposing adjustments without a phase alignment mechanism can easily cause tensile resonance, leading to film bubble rupture. To address the above problems, this invention provides an intelligent control system for the thickness of blown film in biodegradable express delivery bags, such as... Figure 5 As shown. The specific implementation process of this system is as follows: The system relies on an industrial control computer and a low-level logic controller to construct a physical hardware and software operating environment. The rheological decoupling control module is equipped with an analog-to-digital converter and a floating-point arithmetic unit to acquire the real-time melt temperature and internal pressure of the die head. Based on these real-time melt temperature and internal pressure, it performs non-Newtonian rheological decoupling calculations and outputs the reference air valve control voltage and reference traction line speed. The image feature extraction module calls the graphics processing unit to execute a machine vision edge segmentation algorithm, acquires infrared thermal distribution image data, and extracts the actual frost line height and the axial temperature gradient of each circumferential sector at the actual frost line height. The cost model optimization module imports displacement error and axial temperature gradient in an independent software computation thread, performs partial derivative calculations of the thermodynamic energy cost model, and outputs a thermodynamic cooperative allocation factor. The mutual exclusion compensation and drive module acquires the thermodynamic cooperative allocation factor and its complementary values, performs mutual exclusion weighted calculations through an internal arithmetic logic unit, and combines this with the stress transmission delay constant to generate the actual air valve control voltage and delayed traction line speed, respectively. It then converts digital instructions into analog level signals and communication messages to directly drive the local proportional valve and traction device. The macro-deviation feedback module establishes a low-frequency polling task, generates an updated target frost line height based on the acquired thickness deviation data, and feeds the updated target frost line height back to the cost model optimization module to complete the global macro-control closed loop that eliminates steady-state drift.
[0054] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for intelligent control of blown film thickness in biodegradable express delivery bags, characterized in that, Includes the following steps: S1. Obtain the real-time melt temperature and real-time internal pressure of the die head, perform non-Newtonian flow decoupling calculation based on the real-time melt temperature and real-time internal pressure of the die head, and output the reference air valve control voltage and reference traction line speed. S2. Acquire infrared thermal distribution image data, and based on the infrared thermal distribution image data, extract the actual frost line height and the axial temperature gradient of each circumferential sector at the actual frost line height. S3. Calculate the displacement error based on the actual frost line height and the target frost line height, input the displacement error and the axial temperature gradient into the thermodynamic energy cost model to solve, and output the thermodynamic cooperative allocation factor. S4. Based on the thermodynamic cooperative distribution factor and the complementary value of the thermodynamic cooperative distribution factor, perform mutual exclusion compensation calculation on the reference air valve control voltage and the reference traction line speed, and generate the actual air valve control voltage and the delayed traction line speed respectively by combining the stress transmission delay constant. Use the actual air valve control voltage to drive the local proportional valve, and use the delayed traction line speed to drive the traction device. S5. Obtain thickness deviation data, generate an updated target frost line height based on the thickness deviation data, and feed the updated target frost line height back into the thermodynamic energy cost model.
2. The intelligent control method for blown film thickness of biodegradable express delivery bags according to claim 1, characterized in that, In S1, the process of acquiring the real-time melt temperature and real-time internal pressure of the die head, performing non-Newtonian flow decoupling calculations based on the real-time melt temperature and real-time internal pressure of the die head, and outputting the reference air valve control voltage and reference traction linear velocity includes: A physical state sampling action is performed using a high-frequency thermocouple and a high-frequency melt pressure sensor to obtain the real-time melt temperature and real-time internal pressure of the die head, respectively. Time series alignment and filtering are performed on the real-time melt temperature and real-time internal pressure of the die head. Calculate the pressure difference between the real-time internal pressure of the mold head and the preset target internal pressure of the mold head, and multiply the pressure difference by the preset pressure and speed coupling gain coefficient to generate the first intermediate product; The numerical quotient is obtained by dividing the preset viscous flow activation temperature constant by the real-time melt temperature of the die head; The negative number of the numerical quotient is raised to the power of its natural logarithm to generate the second intermediate product; Multiply the first intermediate product by the second intermediate product to generate the pressure pulse speed adjustment amount; The reference screw speed is generated by subtracting the pressure pulse speed adjustment amount from the rated operating speed. The reference screw speed is multiplied by the aerodynamic mapping constant and the motion mapping constant to generate the reference air valve control voltage and the reference traction linear velocity, respectively.
3. The intelligent control method for blown film thickness of biodegradable express delivery bags according to claim 1, characterized in that, In S2, acquiring infrared thermal distribution image data, and based on the infrared thermal distribution image data, extracting the actual frost line height and the axial temperature gradient of each circumferential sector at the actual frost line height, includes: An infrared thermal imaging sensor is used to perform an infrared band panoramic scan to obtain the original thermal radiation data matrix. The original thermal radiation data matrix is then processed to remove background thermal noise using the workshop ambient temperature reference data to generate the infrared thermal distribution image data. The infrared thermal distribution image data is subjected to image spatial segmentation processing to extract the solid-liquid phase transition boundary contour on the surface of the membrane bubble; The spatial mean is calculated based on the height coordinate data of the solid-liquid phase transition boundary contour on the vertical spatial axis to generate the actual frost line height. Divide the membrane bubble into a predetermined number of circumferential sectors along its circumferential direction, and map the solid-liquid phase transition boundary contour into each of the circumferential sectors; At the vertical spatial position corresponding to the actual frost line height, the temperature value inside each of the circumferential sectors is extracted, and the temperature value is subjected to a first-order spatial derivative operation along the axial spatial coordinate axis to generate the axial temperature gradient.
4. The intelligent control method for blown film thickness of biodegradable express delivery bags according to claim 1, characterized in that, In S3, the step of inputting the displacement error and the axial temperature gradient into the thermodynamic energy cost model for solution, and outputting the thermodynamic cooperative allocation factor, includes: Construct a system total energy cost loss function that includes displacement cost penalty term, aerodynamic cooling compensation work penalty term and traction motion phase shift work penalty term; The aerodynamic cooling compensation work penalty term is set as an exponential positive correlation penalty function with the axial temperature gradient as the independent variable; The first-order partial derivative of the total energy cost loss function of the system is calculated with respect to the thermodynamic cooperative allocation factor variable to be solved, and the partial derivative equation is obtained. Set the calculation result of the partial derivative equation to zero, solve the algebraic solution of the extreme point of the partial derivative equation in the continuous real number domain, and extract the algebraic solution of the extreme point as the thermodynamic co-allocation factor.
5. The intelligent control method for blown film thickness of biodegradable express delivery bags according to claim 4, characterized in that, In S3, setting the aerodynamic cooling compensation work penalty term as an exponential positive correlation penalty function with the axial temperature gradient as the independent variable includes: The axial temperature gradient is multiplied by a preset latent heat of crystallization buffer constant to generate an exponential independent variable value; The thermodynamic amplification factor is generated by exponentially raising the natural logarithm of the independent exponential variable. The preset basic aerodynamic weight is multiplied by the thermodynamic amplification factor to generate the thermal suppression penalty weight; The aerodynamic cooling compensation work penalty term is generated by multiplying the thermal suppression penalty weight by the square of the thermodynamic co-allocation factor to be solved.
6. The intelligent control method for blown film thickness of biodegradable express delivery bags according to claim 1, characterized in that, In S4, the mutual exclusion compensation calculation is performed on the reference damper control voltage and the reference traction line velocity based on the complementary values of the thermodynamic cooperative allocation factor and the thermodynamic cooperative allocation factor, and the actual damper control voltage and delayed traction line velocity are generated by combining the stress transmission delay constant, including: The complementary value of the thermodynamic co-allocation factor is generated by subtracting the value from the value. The thermodynamic cooperative allocation factor is multiplied by the preset aerodynamic coupling compensation gain coefficient to generate the first correction coefficient; Multiply the first correction coefficient by the displacement error to generate the aerodynamic voltage compensation increment; The reference valve control voltage is added to the pneumatic voltage compensation increment, and voltage amplitude saturation limiting processing is performed to generate the actual valve control voltage.
7. The intelligent control method for blown film thickness of biodegradable express delivery bags according to claim 6, characterized in that, In S4, the step of performing mutual exclusion compensation calculations on the reference damper control voltage and the reference traction line speed based on the complementary values of the thermodynamic cooperative distribution factor and the thermodynamic cooperative distribution factor, and generating the actual damper control voltage and delayed traction line speed by combining the stress transmission delay constant, further includes: The complementary value of the thermodynamic cooperative allocation factor is multiplied by the preset kinematic following compensation coefficient to generate the second correction coefficient; Perform a first-order difference derivative operation on the actual height of the frost-white line with respect to the time dimension to generate the vertical drift velocity value of the frost-white line. Multiply the second correction coefficient by the vertical drift velocity of the frost-white line to generate the traction speed compensation reduction; Obtain the pre-stored physical space span length of the membrane bubble, and calculate the stress transmission delay constant by dividing the physical space span length of the membrane bubble by the reference traction linear velocity. The reference traction line speed is subjected to a time delay processing based on the stress transmission delay constant, and the traction speed compensation reduction is subtracted after the delay processing to generate the delayed traction line speed.
8. The intelligent control method for blown film thickness of biodegradable express delivery bags according to claim 1, characterized in that, In S4, the step of using the actual air valve control voltage to drive the local proportional valve and using the delayed traction linear velocity to drive the traction device includes: For each of the circumferential sectors, the axial temperature gradient is compared with the preset critical fracture temperature gradient threshold in real time. In response to a hardware status indicator that the axial temperature gradient is greater than the critical fracture temperature gradient threshold, a hardware interlock interrupt command is triggered. Based on the hardware interlock interrupt instruction, the compensation increment of the actual air valve control voltage is forcibly cleared to zero, the reference voltage is output to the local proportional valve, and the delayed traction linear speed is simultaneously sent to the traction device to perform the stretching action.
9. The intelligent control method for blown film thickness of a biodegradable express delivery bag according to claim 1, characterized in that, In S5, generating the updated target frost line height based on the thickness deviation data includes: The thickness deviation data is subjected to moving average filtering to remove high-frequency measurement noise; The filtered thickness deviation data is input into a discrete proportional-integral feedback controller, and error integral accumulation is performed within a set time period to generate a height correction bias. The height correction bias is summed with the current time period base value of the target frost line height to generate the updated target frost line height.
10. A smart control system for the thickness of blown film in biodegradable express delivery bags, characterized in that, A method for intelligent control of blown film thickness in a biodegradable express delivery bag according to any one of claims 1-9, comprising the following modules: The rheological decoupling control module is used to acquire the real-time melt temperature and real-time internal pressure of the die head, perform non-Newtonian rheological decoupling calculations based on the real-time melt temperature and real-time internal pressure of the die head, and output the reference air valve control voltage and reference traction linear velocity. The image feature extraction module is used to acquire infrared thermal distribution image data, and based on the infrared thermal distribution image data, extract the actual frost line height and the axial temperature gradient of each circumferential sector at the actual frost line height. The cost model optimization module is used to calculate the displacement error based on the actual frost line height and the target frost line height, input the displacement error and the axial temperature gradient into the thermodynamic energy cost model for solution, and output the thermodynamic cooperative allocation factor. The mutual exclusion compensation and drive module is used to perform mutual exclusion compensation calculations on the reference air valve control voltage and the reference traction line speed based on the complementary values of the thermodynamic cooperative distribution factor and the thermodynamic cooperative distribution factor. It generates the actual air valve control voltage and the delayed traction line speed by combining the stress transmission delay constant. The actual air valve control voltage is used to drive the local proportional valve, and the delayed traction line speed is used to drive the traction device. The macroscopic deviation feedback module is used to acquire thickness deviation data, generate an updated target frost line height based on the thickness deviation data, and feed the updated target frost line height back into the thermodynamic energy cost model.
