Full-automatic biodegradable film blow molding and winding system and intelligent control process

Abstract

The invention relates to the technical field of blow molding and rolling of biodegradable films, in particular to a full-automatic blow molding and rolling system for biodegradable films and an intelligent control process. The device comprises a plasma activation module (for improving the surface tension of a raw material to be greater than or equal to 50mN / m), a multi-field collaborative blow molding module (integrated reinforcement learning bubble control), a quantum level detection module (with terahertz scanning precision of 0.05 mm < 2 >), a magnetic suspension winding module (with an electromagnetic coil array control distance of + / -5 [mu] m) and a digital twinborn center, wherein the plasma activation module is used for improving the surface tension of the raw material to be greater than or equal to 50mN / m; the process comprises the following steps of: reinforcement learning control of film bubble stability: outputting an air ring / inner cooling / herringbone clamp adjusting quantity in real time; stress balance control: delta sigma = K / (E (t) / r) dr + alpha delta T; carrying out digital twinning online optimization; the invention achieves the effects that the thickness uniformity (CV is less than or equal to 2.1%), the production efficiency (the linear speed is 120m / min), the energy consumption (0.68 kWh / kg) and the intelligent level (the response is less than or equal to 80ms) are advanced in the industry.

Description

Technical Field

[0001] This invention relates to the field of biodegradable film blow molding and winding technology, and in particular to a fully automated biodegradable film blow molding and winding system and intelligent control process. Background Technology

[0002] With increasingly stringent environmental policies, the market demand for biodegradable films such as PLA, PBAT, and PBS has surged. As a key link in film production, blown winding has two main technological development paths: one is the traditional mechanical winding system, and the other is the first-generation automated winding system.

[0003] Traditional mechanical winding systems, including magnetic powder brake tension control, pneumatic pressure roller contact winding, and manually set taper curves, have the following shortcomings: ① rigid mechanical contact leads to scratches on the film surface; ② fixed taper curves lead to core collapse; ③ excessive manual intervention easily leads to large quality fluctuations.

[0004] The initial automated winding system, including PID-based closed-loop tension control, ultrasonic online thickness measurement (accuracy ±3μm), and an automatic roll-changing robot, had the following shortcomings: ① The detection-execution delay caused thickness control lag, resulting in a thickness CV value greater than 8%; ② The failure to consider the time-varying viscoelastic properties of the material led to uncontrolled bubble oscillation, resulting in a film breakage rate of 23 times / 10,000 meters; ③ The unreasonable layout of the ion fan caused severe static electricity accumulation, resulting in a dust adsorption defect rate of 15%; ④ Ignoring the influence of temperature gradients resulted in uncompensated stress relaxation, leading to a core deformation rate of 38% after 7 days of storage.

[0005] To address these issues, we propose a fully automated biodegradable film blow molding and winding system and an intelligent control process. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of existing technologies by proposing a fully automated biodegradable film blow molding and winding system and intelligent control process.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A fully automated biodegradable membrane blow molding and winding system includes a plasma raw material activation module, a multi-field collaborative blow molding module, a quantum-level online detection module, a magnetic levitation winding execution module, and a digital twin control center;

[0009] The plasma raw material activation module uses atmospheric pressure cold plasma (15-20kV / 40kHz) to increase the surface tension of the raw material to ≥50mN / m;

[0010] A multi-field collaborative blow molding module integrates a reinforcement learning controller and an electromagnetic eddy current auxiliary device to suppress membrane bubble oscillation;

[0011] The quantum-level online detection module includes a quantum dot thickness sensor (0.1 μm resolution) and a terahertz defect scanner (0.05 mm resolution). 2 Detection accuracy);

[0012] The magnetic levitation winding execution module achieves zero-contact winding with a gap of 50±5μm through an electromagnetic coil array;

[0013] A digital twin control center is used to construct a five-dimensional model to achieve process self-optimization.

[0014] As a preferred embodiment of the present invention, the plasma raw material activation module adopts atmospheric pressure dielectric barrier discharge (DBD) technology to perform surface modification before the raw material melts, which can increase the surface tension of PLA from 38mN / m to 52mN / m (measured data) and solve the problem of poor biofilm adhesion.

[0015] It includes: an atmospheric pressure cold plasma generator with a power of 15-20kV and a frequency of 40±5kHz;

[0016] Nanocrystalline nucleus addition unit, containing 0.1-0.5 wt% POSS nanoparticles;

[0017] Electrostatic neutralization air curtain, ion concentration ≥1×10 6 pcs / cm 3 .

[0018] As a preferred embodiment of the present invention, the quantum-level online detection module comprises:

[0019] CdSe / ZnS quantum dot thickness sensor with a resolution of 0.1 μm;

[0020] Graphene strain sensing array, attached to the film surface;

[0021] Terahertz defect scanner, operating frequency 0.3THz.

[0022] As a preferred embodiment of the present invention, the magnetic levitation winding execution module includes:

[0023] An array of 32 electromagnetic coils, with a single coil magnetic field strength of 0.5-1.2T;

[0024] Air gap sensor, detection accuracy ±5μm;

[0025] The stress balance controller executes the following algorithm:

[0026] As a preferred embodiment of the present invention, the digital twin control center implements:

[0027] Multi-production line parameter optimization based on federated learning;

[0028] Real-time mapping of five-dimensional models (physical entities / virtual models / service data / connectivity and interaction / intelligent decision-making);

[0029] Construction of process knowledge graph (≥2000 nodes).

[0030] As a preferred embodiment of the present invention, an intelligent control process for a fully automated biodegradable film blow molding and winding system includes the following steps:

[0031] S1. The raw material undergoes plasma activation treatment to achieve a surface tension ≥50mN / m;

[0032] S2. The code for the membrane bubble stability control logic based on reinforcement learning is as follows:

[0033] action=policy_net([thickness_dev,wind_pressure,speed,te mp_grad]);

[0034] S3, Real-time feedback of quantum dot thickness (sampling rate ≥ 500Hz);

[0035] S4. Magnetic levitation winding gap control 50±5μm;

[0036] S5, parameter self-optimization driven by digital twin.

[0037] As a preferred embodiment of the present invention, in S2:

[0038] The control variables are generated using a D3QN network.

[0039] Input parameters include thickness deviation, cooling air pressure, traction speed, and temperature gradient;

[0040] The outputs are air ring opening, internal cooling flow rate, and V-clamp angle.

[0041] As a preferred embodiment of the present invention, stress balance control is performed in S4:

[0042] The material memory factor K is updated in real time by an online rheometer, and the coefficient of thermal expansion α is set according to the material type, with α value of 1.2 for PLA, 0.8 for PBAT, and 0.95 for PBS.

[0043] As a preferred embodiment of the present invention, S5 includes:

[0044] Real-time generation of process optimization dashboards;

[0045] Control commands are triggered based on confidence thresholds;

[0046] Autonomous backtracking analysis of abnormal operating conditions.

[0047] A computer-readable storage medium storing program instructions for performing the process described in any one of claims 6-9.

[0048] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0049] This invention achieves the following through four core technologies: plasma raw material activation, reinforcement learning bubble control (action = policy_net([thickness deviation, wind pressure, speed, temperature difference])), quantum-level online detection (thickness resolution 0.1μm), and magnetic levitation zero-contact winding (gap control 50±5μm):

[0050] ① Quality breakthrough: Thickness CV value ≤ 2.1% (improved by 75.9%), core eccentricity ≤ 0.1mm (improved by 66.7%), 7-day stress relaxation rate ≤ 8.5% (improved by 77.6%);

[0051] ② Efficiency leap: Linear speed ≥120m / min (increased by 50%), roll change time ≤12s (reduced by 86.7%), membrane bubble oscillation shutdown ≤1 time / shift (reduced by 95.7%);

[0052] ③ Cost optimization: Energy consumption ≤ 0.68 kWh / kg (reduced by 43.3%), raw material loss rate ≤ 2.2% (estimated annual savings of 315 tons), and premium product rate ≥ 98.5% (estimated annual revenue increase of 6.2 million yuan / line). Attached Figure Description

[0053] Figure 1 This is a logic block diagram of the magnetic levitation winding execution module in this invention;

[0054] Figure 2 This is a diagram of the reinforcement learning control architecture in the reinforcement learning-based membrane bubble stability control of this invention.

[0055] Figure 3 This is a data fusion architecture diagram of the core sensor in the quantum-level online detection module of this invention. Detailed Implementation

[0056] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0057] A fully automated biodegradable membrane blow molding and winding system includes a plasma raw material activation module, a multi-field collaborative blow molding module, a quantum-level online detection module, a magnetic levitation winding execution module, and a digital twin control center.

[0058] The plasma raw material activation module employs atmospheric pressure dielectric barrier discharge (DBD) technology to modify the surface of the raw material before melting, increasing the surface tension of PLA from 38 mN / m to 52 mN / m (measured data), thus solving the problem of poor biofilm adhesion. It includes: an atmospheric pressure cold plasma generator with a power of 15-20 kV and a frequency of 40 ± 5 kHz; a nanocrystalline nucleus addition unit containing 0.1-0.5 wt% POSS nanoparticles; and an electrostatic neutralization air curtain with an ion concentration ≥ 1 × 10⁻⁶. 6 pcs / cm 3 .

[0059] The structural composition and technical parameters of the multi-field collaborative blow molding module are shown in Table 5 below.

[0060] The quantum-level online detection module includes: a CdSe / ZnS quantum dot thickness sensor with a resolution of 0.1 μm; a graphene strain sensing array attached to the film surface; and a terahertz defect scanner with an operating frequency of 0.3 THz.

[0061] Table 6 shows the core sensors and performance of the quantum-level online detection module, and the data fusion architecture reference. Figure 3 ;

[0062] The modular collaborative workflow (taking PLA film production as an example) includes the blow molding stage, the real-time detection stage, and the closed-loop control stage;

[0063] During the blow molding stage: the electromagnetic eddy current device generates counterclockwise rotating eddies (1200 rpm) inside the bubble, and the zoned micro-mist system operates according to the set temperature gradient: upper zone: 45℃ → to prevent excessive crystallization; middle zone: 35℃ → main cooling zone; lower zone: 50℃ → to prevent shrinkage stress.

[0064] During the real-time detection phase: the quantum dot sensor detected an anomaly in thickness (coordinates X: 350 mm, Y: 1200 mm); the terahertz scanner confirmed the presence of a 0.08 mm thickness at this location. 2 Bubble defects; alarm signals are generated after data fusion;

[0065] The closed-loop control logic during the closed-loop control phase is as follows:

[0066] #Code snippet for reinforcement learning controller decision-making

[0067] state = [thickness_dev = 0.15, #thickness deviation 15%]

[0068] wind_pressure = 12.3, #wind pressure 12.3 kPa

[0069] speed = 95, # Linear velocity 95 m / min

[0070] [temp_grad=7.8] # Temperature gradient 7.8℃

[0071] action = policy_net(state) # Output control variable

[0072] #Execution: Air ring nozzle #47 opening +18%, central cooling flow -15%.

[0073] The electromagnetic coils of the magnetic levitation winding execution module are arranged in a Halbach array, which increases the magnetic field strength by 40%. The block diagram of its control logic is as follows: Figure 1 As shown; the magnetic levitation winding execution module includes: a 32-group electromagnetic coil array, with a single coil magnetic field strength of 0.5-1.2T; an air gap sensor with a detection accuracy of ±5μm; and a stress balance controller. The execution algorithm is: Please refer to Tables 1 and 2 below.

[0074] Digital twin control center implementation: multi-production line parameter optimization based on federated learning; real-time mapping of five-dimensional models (physical entity / virtual model / service data / connection interaction / intelligent decision-making); construction of process knowledge graph (≥2000 nodes).

[0075] A smart control process for a fully automated biodegradable film blow molding and winding system includes the following steps:

[0076] S1. The raw material undergoes plasma activation treatment to achieve a surface tension ≥50mN / m;

[0077] S2. The code for the membrane bubble stability control logic based on reinforcement learning is as follows:

[0078] action=policy_net([thickness_dev,wind_pressure,speed,te mp_grad]), refer to Table 4 below;

[0079] In S2, a D3QN network is used to generate control variables;

[0080] Input parameters include thickness deviation, cooling air pressure, traction speed, and temperature gradient;

[0081] The outputs are air ring opening, internal cooling flow rate, and V-clamp angle;

[0082] More specific reinforcement learning-based membrane bubble stability control includes reinforcement learning control architectures, see reference. Figure 2 ,

[0083] S3, Real-time feedback of quantum dot thickness (sampling rate ≥ 500Hz);

[0084] S4. Magnetic levitation winding gap control 50±5μm;

[0085] In S4, stress balance control:

[0086] The material memory factor K is updated in real time via an online rheometer, and the coefficient of thermal expansion α is determined according to the material type, as shown in Table 3 below.

[0087] S5, digital twin-driven parameter self-optimization;

[0088] S5 includes: real-time generation of process optimization dashboards, control commands triggered based on confidence thresholds, and autonomous backtracking analysis of abnormal operating conditions.

[0089] In addition, the present invention also proposes a computer-readable storage medium storing program instructions for executing any of the processes of claims 6-9.

[0090]

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097]

[0098]

[0099]

[0100] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A fully automated biodegradable film blow molding and winding system, characterized in that, It includes a plasma raw material activation module, a multi-field collaborative blow molding module, a quantum-level online detection module, a magnetic levitation winding execution module, and a digital twin control center.

2. The fully automated biodegradable film blow molding and winding system according to claim 1, characterized in that, The plasma raw material activation module includes: Atmospheric pressure cold plasma generator, power 15-20kV, frequency 40±5kHz; Nanocrystalline nucleus addition unit, containing 0.1-0.5 wt% POSS nanoparticles; Electrostatic neutralization air curtain, ion concentration ≥1×10 6 pcs / cm 3 .

3. The fully automated biodegradable film blow molding and winding system according to claim 1, characterized in that, The quantum-level online detection module includes: CdSe / ZnS quantum dot thickness sensor with a resolution of 0.1 μm; Graphene strain sensing array, attached to the film surface; Terahertz defect scanner, operating frequency 0.3THz.

4. The fully automated biodegradable film blow molding and winding system according to claim 1, characterized in that, The magnetic levitation winding execution module includes: An array of 32 electromagnetic coils, with a single coil magnetic field strength of 0.5-1.2T; Air gap sensor, detection accuracy ±5μm; The stress balance controller executes the following algorithm:

5. The fully automated biodegradable film blow molding and winding system according to claim 1, characterized in that, The digital twin control center achieves: Multi-production line parameter optimization based on federated learning; Real-time mapping of five-dimensional models (physical entities / virtual models / service data / connectivity and interaction / intelligent decision-making); Construction of process knowledge graph (≥2000 nodes).

6. The intelligent control process of a fully automated biodegradable film blow molding and winding system according to any one of claims 1-5, characterized in that, Includes the following steps: S1. The raw material undergoes plasma activation treatment to achieve a surface tension ≥50mN / m; S2. The membrane bubble stability control logic based on reinforcement learning is as follows: action=policy_net([thickness_dev,wind_pressure,speed,te mp_grad]); S3, Real-time feedback of quantum dot thickness (sampling rate ≥ 500Hz); S4. Magnetic levitation winding gap control 50±5μm; S5, parameter self-optimization driven by digital twin.

7. The intelligent control process of the fully automated biodegradable film blow molding and winding system according to claim 6, characterized in that, In S2: The control variables are generated using a D3QN network. Input parameters include thickness deviation, cooling air pressure, traction speed, and temperature gradient; The outputs are air ring opening, internal cooling flow rate, and V-clamp angle.

8. The intelligent control process of a fully automated biodegradable film blow molding and winding system according to claim 6, characterized in that, Stress balance control in S4: The material memory factor K is updated in real time via an online rheometer, and the coefficient of thermal expansion α is assigned a value according to the material type.

9. The intelligent control process of a fully automated biodegradable film blow molding and winding system according to claim 6, characterized in that, S5 includes: Real-time generation of process optimization dashboards; Control commands are triggered based on confidence thresholds; Autonomous backtracking analysis of abnormal operating conditions.

10. A computer-readable storage medium storing program instructions for performing the process of any one of claims 6-9.

Images