Process optimization method for continuous extrusion molding of wood-plastic composite material

Abstract

The invention provides a process optimization method for continuous extrusion molding of a wood-plastic composite material, and belongs to the technical field of wood-plastic composite material molding processes. A sectional type temperature control system is established, an extruder charging barrel is divided into four independent temperature control areas, the targeted optimal target temperature is set, a double-layer game model is constructed to achieve collaborative optimization of temperature stability and energy consumption, and parameter iteration adjustment is conducted through an intelligent multi-parameter monitoring platform according to a temperature uniformity index. And finally, online quality detection is carried out on the extruded product, so that the technical problems that wood fibers are easily subjected to thermal degradation at high temperature, a plastic matrix can be fully fused at enough high temperature, a processing temperature window of the wood fibers and the plastic matrix is very narrow, local overheating is easily caused, and the performance of a wood-plastic composite finished product is insufficient are solved.

Description

Technical Field

[0001] The invention belongs to the technical field of wood-plastic composite material molding technology, and in particular relates to a process optimization method for continuous extrusion molding of wood-plastic composite materials. Background Art

[0002] Wood-plastic composites (WPCs), as environmentally friendly composite materials, are widely used in architectural decoration, outdoor landscaping, automotive interiors, and other fields. They are primarily produced through continuous extrusion using twin-screw extruders. Traditional WPC extrusion processes utilize a segmented heating system, setting distinct temperature zones to melt and plasticize the wood fiber and plastic matrix. However, due to the different thermal sensitivities of wood fiber and the plastic matrix, the wood fiber begins to undergo significant thermal degradation above 160°C, while the plastic matrix typically requires temperatures above 180°C to fully melt, resulting in significant differences in their optimal processing temperatures. In current WPC extrusion production, the narrow processing temperature window for both the wood fiber and the plastic matrix makes it difficult to precisely control the temperature distribution across different zones, leading to localized overheating. When the temperature is too high, the wood fiber undergoes thermal degradation, producing volatile substances that weaken the composite's interfacial bonding and deteriorate its mechanical properties. When the temperature is too low, the plastic matrix fails to fully melt, compromising the coating and dispersion uniformity of the wood fiber. Existing control systems lack the ability to precisely regulate temperature distribution uniformity, making it difficult to ensure sufficient melting of the plastic matrix while avoiding excessive degradation of the wood fiber. That is to say, there is a technical problem in the existing technology that wood fibers are prone to thermal degradation at high temperatures, while the plastic matrix requires a sufficiently high temperature to fully melt. The processing temperature window of the two is very narrow, which can easily cause local overheating and lead to insufficient performance of the finished wood-plastic composite material. Summary of the Invention

[0003] In view of this, the present invention provides a process optimization method for continuous extrusion molding of wood-plastic composite materials, which can solve the technical problems in the prior art that wood fibers are easily thermally degraded at high temperatures, while the plastic matrix requires a sufficiently high temperature to fully melt. The processing temperature window of the two is very narrow, which easily causes local overheating and leads to insufficient performance of the finished wood-plastic composite materials.

[0004] The present invention is achieved as follows: The present invention provides a process optimization method for continuous extrusion molding of wood-plastic composite materials, comprising: pre-treating the wood fiber raw material with a silane coupling agent to modify the surface, controlling the moisture content of the wood fiber to be below 2%, and the particle size to be within the range of 40 to 80 meshes; mixing the wood fiber and the plastic matrix in a mass ratio of 30:70 to 50:50, adding a compatibilizer with a mass fraction of 2% to 4%, and melt-blending them through a twin-screw extruder; establishing a segmented temperature control system, dividing the extruder barrel into four temperature control zones according to a feeding zone, a plasticizing zone, a homogenizing zone, and an extrusion zone, and providing each zone with an independent temperature sensor and a heating element. , five process parameters including temperature control area temperature, barrel pressure, screw torque, extrusion speed and motor current intensity are collected through a multi-sensor data acquisition platform; the optimal target temperature of each temperature control area is determined based on genetic algorithm optimization, and a temperature classification control strategy is established; an upper-level game model with temperature stability as the goal and a lower-level game model with energy consumption minimization as the goal are established to achieve real-time fine-tuning of temperature control; the barrel temperature distribution and process parameter changes are monitored in real time through an intelligent multi-parameter monitoring platform, and the collection parameters are iteratively optimized according to the similarity of data features; online quality inspection of extruded wood-plastic composite materials is carried out, and the inspection items include surface finish, dimensional accuracy and mechanical properties.

[0005] Among them, the surface modification treatment with silane coupling agent refers to the chemical modification of the wood fiber surface using a silane coupling agent, through which the alkoxy groups in the silane molecules react with the hydroxyl groups on the wood fiber surface to form a stable chemical bond, thereby reducing the hydrophilicity of the wood fiber and improving the interfacial bonding strength with the plastic matrix.

[0006] Among them, the segmented temperature control system refers to dividing the extruder barrel into multiple independent temperature control areas along the axial direction. Each area is equipped with an independent temperature sensor and heating element to achieve precise control of the temperature of different areas and avoid local overheating or overcooling caused by overall temperature control.

[0007] Among them, the multi-sensor data acquisition platform refers to a comprehensive monitoring device that integrates a temperature sensor array, a pressure sensor, a torque sensor, a photoelectric encoder, and a current sensor. The temperature sensor uses a platinum resistance PT100 type with a measurement accuracy of ±0.1°C. The pressure sensor uses a strain gauge pressure transmitter with a measurement range of 0 to 50 MPa and an accuracy of ±0.25%. The acquisition frequency is once every 3 seconds.

[0008] Among them, the genetic algorithm optimizes the steps of determining the optimal target temperature of each temperature control area, specifically, the optimal target temperature of the feeding section is 150±5℃, the optimal target temperature of the plasticizing section is 170±5℃, the optimal target temperature of the homogenizing section is 190±5℃, and the optimal target temperature of the extrusion section is 180±5℃.

[0009] Among them, the temperature classification control strategy refers to dividing the temperature control into three categories according to the thermal degradation characteristics of wood fiber: a safe temperature range of 145-155°C, a suitable temperature range of 155-185°C, and a dangerous temperature range of 185-195°C. The processing efficiency and material properties are balanced by controlling the proportion of temperature classification in each area.

[0010] Among them, the fiber degradation rate corresponding to the safe temperature range is 0-2%, the fiber degradation rate corresponding to the suitable temperature range is 2-8%, and the fiber degradation rate corresponding to the dangerous temperature range is 8-15%. The proportion of the suitable temperature range is maintained at 60-80%, the safe temperature range accounts for 15-30%, and the dangerous temperature range accounts for 5-15%.

[0011] Among them, the genetic algorithm historical data acquisition method refers to collecting process parameters and product quality data through more than 500 consecutive batches of wood-plastic composite material extrusion experiments, establishing a multidimensional database containing temperature-performance relationship, temperature-energy consumption relationship, and temperature-stability relationship, using 70% of the data as a training set for genetic algorithm parameter optimization, and 30% of the data as a validation set for algorithm performance evaluation.

[0012] Among them, when the temperature deviation exceeds ±2℃, the game optimization adjustment is started. The intelligent multi-parameter monitoring platform includes a combination of a thermal imager, a contact temperature sensor, a non-contact infrared thermometer, a pressure sensor, and an online rheometer. The thermal imager uses a long-wave infrared detector with a wavelength of 8 to 14 microns and a temperature resolution of 0.1℃.

[0013] Among them, the method of iteratively optimizing acquisition parameters based on the similarity of data features refers to extending the acquisition interval to once every 10 seconds when the standard deviation of continuously acquired temperature data is less than 1°C, indicating that the temperature is stable. When the temperature data suddenly changes or the similarity coefficient is lower than 0.9, the acquisition frequency is increased to once every 1 second and the early warning mechanism is activated.

[0014] The objective function of the upper-level game model is to minimize the deviation between the actual temperature of each temperature control zone and the optimal target temperature. The inputs include the temperature of the temperature control zone, barrel pressure, screw torque, extrusion speed, and motor current intensity. The output is a temperature control strategy optimization plan, which is used to guide the energy consumption optimization calculation of the lower-level game model. The objective function of the lower-level game model is to minimize the total energy consumption of the system, including the heating energy consumption of each zone and the motor drive energy consumption. The inputs include the temperature of the temperature control zone, barrel pressure, screw torque, extrusion speed, and motor current intensity. The output is an energy consumption optimization control strategy, which is used to feedback and adjust the temperature control parameters of the upper-level game model.

[0015] Among them, the temperature uniformity index is defined as 1 minus the ratio of the temperature standard deviation to the average temperature. When the temperature uniformity index is lower than 0.85, the process parameters are adjusted. Adjusting the process parameters means iteratively adjusting the three control parameters of screw speed, heating power distribution, and cooling water flow according to the numerical changes of the temperature uniformity index.

[0016] The surface finish test refers to measuring the Ra value of the surface of the wood-plastic composite extruded product by a laser surface roughness measuring instrument, and the qualified threshold range is Ra≤3.2μm. The dimensional accuracy test refers to continuously measuring the outer diameter of the extruded product by a laser diameter measuring instrument, and the qualified threshold range is the design size ±0.1mm.

[0017] Among them, the mechanical properties test refers to measuring the tensile strength and flexural strength of wood-plastic composite materials through an online tensile testing machine. The qualified threshold range of tensile strength is ≥25MPa, and the qualified threshold range of flexural strength is ≥40MPa. Qualified products enter the subsequent cooling and shaping process, and unqualified products are returned for reprocessing.

[0018] The compatibilizer refers to an additive that can improve the interfacial bonding between incompatible polymers, thereby improving the compatibility between different materials by reducing interfacial tension or forming chemical bonds. Commonly used compatibilizers include maleic anhydride grafted polymers and silane coupling agents.

[0019] The present invention establishes a segmented temperature control system combined with a genetic algorithm optimization and a two-level game model intelligent control strategy to address the technical problem that wood fiber is susceptible to thermal degradation at high temperatures, while the plastic matrix requires a sufficiently high temperature to fully melt. The narrow processing temperature window between the two is prone to local overheating, resulting in insufficient performance of the finished wood-plastic composite material. The present invention adopts a temperature classification control strategy, dividing temperature control into three categories: a safe temperature range of 145-155°C, a suitable temperature range of 155-185°C, and a dangerous temperature range of 185-195°C. By controlling the proportion of each temperature category, the thermal stability of the wood fiber and the melting effect of the plastic matrix are balanced. The suitable temperature range accounts for 60-80%, and the dangerous temperature range accounts for 5-15%, effectively avoiding the problem of large-scale wood fiber degradation. At the same time, a multi-sensor data acquisition platform and an intelligent monitoring system achieve precise control of temperature distribution uniformity. When the temperature uniformity index falls below 0.85, the process parameters are automatically adjusted to eliminate local overheating. The present invention controls the degradation rate of wood fibers within a reasonable range while ensuring sufficient melting of the plastic matrix through precise temperature distribution control and real-time optimization and adjustment, thereby achieving the coordinated unification of the thermal stability of wood fibers and the melting effect of the plastic matrix, and solving the technical problems of local overheating and insufficient finished product performance caused by a narrow processing temperature window. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 is a flow chart of the method of the present invention.

[0021] Figure 2 This is a diagram of the temperature control optimization process in Example 2.

[0022] Figure 3 This is a diagram of the convergence process of the game model in Example 2.

[0023] Figure 4 This is a temperature uniformity index change diagram in Example 2.

[0024] Figure 5 This is a distribution diagram of the product quality test results in Example 2. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0026] like Figure 1 FIG. 1 is a flow chart of a process optimization method for continuous extrusion molding of a wood-plastic composite material provided by the present invention. The method comprises the following steps:

[0027] S01, pre-treating the wood fiber raw material, reducing the hydroxyl content on the surface of the wood fiber by surface modification with a silane coupling agent, controlling the moisture content of the wood fiber to below 2%, and controlling the particle size of the wood fiber to within the range of 40 to 80 mesh;

[0028] S02, mixing the wood fiber and the plastic matrix in a mass ratio of 30:70 to 50:50, adding a compatibilizer with a mass fraction of 2% to 4%, and melt-blending through a twin-screw extruder;

[0029] S03. Establish a segmented temperature control system, divide the extruder barrel into four temperature control zones according to the feeding zone, plasticizing zone, homogenizing zone, and extrusion zone, and set an independent temperature sensor and heating element in each zone;

[0030] S04. Collect five process parameters, including temperature in the temperature control area, barrel pressure, screw torque, extrusion speed, and motor current intensity, through a multi-sensor data acquisition platform at a frequency of once every 3 seconds;

[0031] S05. Determine the optimal target temperature for each temperature control zone based on genetic algorithm optimization. The optimal target temperature for the feeding section is 150±5°C, the optimal target temperature for the plasticizing section is 170±5°C, the optimal target temperature for the homogenizing section is 190±5°C, and the optimal target temperature for the extrusion section is 180±5°C. A temperature classification control strategy is then established.

[0032] S06. Establish an upper-level game model with temperature stability as the goal and a lower-level game model with energy consumption minimization as the goal to achieve real-time fine-tuning of temperature control. When the temperature deviation exceeds ±2°C, initiate game optimization adjustment;

[0033] S07. Real-time monitoring of barrel temperature distribution and process parameter changes is performed through an intelligent multi-parameter monitoring platform. The monitoring platform includes a thermal imager, pressure sensor, and rheometer combination device. The acquisition parameters are iteratively optimized based on the similarity of data features. When the temperature uniformity index is lower than 0.85, the process parameters are adjusted;

[0034] S08. Conduct online quality inspection on the extruded wood-plastic composite materials. The inspection items include surface finish, dimensional accuracy and mechanical properties. Qualified products enter the subsequent cooling and shaping process, and unqualified products are returned for reprocessing.

[0035] Among them, surface modification treatment with silane coupling agent refers to the chemical modification of the wood fiber surface using silane coupling agent, through which the alkoxy groups in the silane molecules react with the hydroxyl groups on the wood fiber surface to form a stable chemical bond, thereby reducing the hydrophilicity of the wood fiber and improving the interfacial bonding strength with the plastic matrix.

[0036] Among them, the segmented temperature control system refers to dividing the extruder barrel into multiple independent temperature control areas along the axial direction. Each area is equipped with an independent temperature sensor and heating element to achieve precise control of the temperature of different areas and avoid local overheating or overcooling caused by overall temperature control.

[0037] Among them, the multi-sensor data acquisition platform refers to a comprehensive monitoring device that integrates a temperature sensor array, a pressure sensor, a torque sensor, a photoelectric encoder, and a current sensor. In order to obtain high-precision process data, the temperature sensor uses a platinum resistance PT100 type with a measurement accuracy of ±0.1°C. The pressure sensor uses a strain gauge pressure transmitter with a measurement range of 0 to 50 MPa and an accuracy of ±0.25%. The temperature in the temperature control area comes from the measurement results of the temperature sensor array, the barrel pressure comes from the measurement results of the pressure sensor, the screw torque comes from the measurement results of the torque sensor, the extrusion speed comes from the measurement results of the photoelectric encoder, and the motor current intensity comes from the measurement results of the current sensor.

[0038] Among them, the temperature classification control strategy refers to dividing the temperature control into three categories according to the thermal degradation characteristics of wood fiber: the safe temperature range of 145-155°C, the suitable temperature range of 155-185°C, and the dangerous temperature range of 185-195°C. The fiber degradation rate corresponding to the safe temperature range is 0-2%, the fiber degradation rate corresponding to the suitable temperature range is 2-8%, and the fiber degradation rate corresponding to the dangerous temperature range is 8-15%. The processing efficiency and material properties are balanced by controlling the proportion of temperature classification in each area. The proportion of the suitable temperature range should be maintained at 60-80%, the safe temperature range accounts for 15-30%, and the dangerous temperature range accounts for 5-15%. When the proportion of the dangerous temperature range exceeds 20%, it will cause a large amount of wood fiber degradation, affecting the mechanical properties of the composite material. The negative impact can be reduced by increasing cooling measures and reducing the screw speed.

[0039] Among them, the genetic algorithm historical data acquisition method refers to collecting process parameters and product quality data through more than 500 consecutive batches of wood-plastic composite material extrusion experiments, recording key information such as temperature set value, actual temperature curve, product mechanical property test results, energy consumption data, etc. for each batch, and establishing a multidimensional database containing temperature-performance relationship, temperature-energy consumption relationship, and temperature-stability relationship. 70% of the data is used as a training set for genetic algorithm parameter optimization, and 30% of the data is used as a validation set for algorithm performance evaluation. The optimal target temperature combination for each temperature control area is found through the selection, crossover, and mutation operations of the genetic algorithm.

[0040] Among them, the objective function of the upper-level game model is used to minimize the deviation between the actual temperature of each temperature control area and the optimal target temperature. The temperature deviation weight coefficient is calculated based on the historical temperature control data. The preset functional relationship is selected as the sum of the squared temperature deviation terms multiplied by the weight coefficient, and then the barrel pressure fluctuation term and the temperature stability coupling exponential function are added. The input includes the temperature of the temperature control area, the barrel pressure, the screw torque, the extrusion speed, and the motor current intensity. The output is the temperature control strategy optimization plan, which is used to guide the energy consumption optimization calculation of the lower-level game model.

[0041] Among them, the objective function of the lower-level game model is used to minimize the total energy consumption of the system, including the heating energy consumption of each area and the motor drive energy consumption. According to the analysis results of historical energy consumption statistical data, the preset functional relationship is selected as the sum of the 1.5-power terms of the temperature of the temperature control area, multiplied by the corresponding energy consumption coefficient, plus the square term of the motor current intensity divided by the efficiency coefficient, and finally subtracted from the coupled logarithmic function of the barrel pressure and energy consumption optimization. The input includes the temperature of the temperature control area, barrel pressure, screw torque, extrusion speed, and motor current intensity. The output is the energy consumption optimization control strategy, which is used to feedback and adjust the temperature control parameters of the upper-level game model.

[0042] Among them, the intelligent multi-parameter monitoring platform includes a combination of thermal imagers, contact temperature sensors, non-contact infrared thermometers, pressure sensors, and online rheometers. The thermal imager uses a long-wave infrared detector with a wavelength of 8 to 14 microns and a temperature resolution of 0.1°C to detect the uniformity of temperature distribution on the barrel surface. The contact temperature sensor is embedded in the inner wall of the barrel to monitor the actual temperature of the melt. The online rheometer measures the change in melt viscosity in real time. The method of iterative optimization of acquisition parameters based on the similarity of data features is that when the standard deviation of continuously acquired temperature data is less than 1°C, it indicates that the temperature is stable. At this time, the acquisition interval is extended to once every 10 seconds. When the temperature data suddenly changes or the similarity coefficient is lower than 0.9, the acquisition frequency is increased to once every 1 second and the early warning mechanism is activated.

[0043] The temperature uniformity index is defined as 1 minus the ratio of the temperature standard deviation to the average temperature. The closer the temperature uniformity index is to 1, the more uniform the temperature distribution is. When the temperature uniformity index is lower than 0.85, it indicates obvious local overheating or overcooling.

[0044] Among them, adjusting the process parameters refers to iteratively adjusting the three control parameters of screw speed, heating power distribution, and cooling water flow according to the numerical changes of the temperature uniformity index. The adjustment function relationship is to determine the parameter adjustment range based on statistical analysis of historical data. When the temperature uniformity index is in the range of 0.8-0.85, the screw speed adjustment range is the current speed multiplied by 0.95, and the heating power adjustment range is the current power minus 5%. When the temperature uniformity index is lower than 0.8, the screw speed adjustment range is the current speed divided by 1.1, the heating power adjustment range is the current power multiplied by 0.9, and the cooling water flow adjustment range is the current flow plus 20%. The adjustment process adopts an iterative method. After each adjustment, wait 30 seconds to recalculate the temperature uniformity index until the temperature uniformity index reaches above 0.85.

[0045] Among them, surface finish detection refers to measuring the Ra value of the surface of wood-plastic composite extruded products by using a laser surface roughness measuring instrument. The qualified threshold range is Ra≤3.2μm. The surface finish test results are used to determine whether the surface quality of the product meets the use requirements.

[0046] Among them, dimensional accuracy detection refers to the continuous measurement of the outer diameter of the extruded product through a laser diameter gauge. The qualified threshold range is the design size ±0.1mm. The dimensional accuracy detection results are used to control the extruder pulling speed and mold temperature.

[0047] Among them, mechanical property testing refers to measuring the tensile strength and flexural strength of wood-plastic composite materials through an online tensile testing machine. The qualified threshold range of tensile strength is ≥25MPa, and the qualified threshold range of flexural strength is ≥40MPa. The mechanical property test results are used to evaluate the interface bonding effect between wood fiber and plastic matrix and the overall material performance.

[0048] Among them, compatibilizer refers to an additive that can improve the interfacial bonding between incompatible polymers, and improve the compatibility between different materials by reducing interfacial tension or forming chemical bonds. Commonly used compatibilizers include maleic anhydride grafted polymers and silane coupling agents.

[0049] The specific implementation of the above steps is described in detail below.

[0050] The specific implementation of step S01 is to first screen the wood fiber raw material for pretreatment. The wood fiber is graded according to the particle size requirement of 40 to 80 mesh through a vibrating screen to ensure that the fiber size is evenly distributed. The purpose of this step is to provide a uniform raw material base for the subsequent surface modification treatment. Then, a drying process is carried out. The screened wood fiber is placed in a vacuum oven and dried at a temperature of 80 to 90°C for 4 to 6 hours to reduce the moisture content to below 2%. The drying process adopts the principle of intermittent heating to avoid thermal degradation of the fiber by controlling the drying time and temperature. Then, a silane coupling agent surface modification treatment is carried out. The dried wood fiber is mixed with a silane coupling agent at a ratio of 1% to 3% of the fiber mass in a high-speed mixer. The mixing speed is controlled at 1500 to 2000 r / min and the mixing time is 15 to 20 minutes. This step uses mechanical force to promote sufficient contact between the silane molecules and the fiber surface. The mixed fibers are then heat-treated for activation at 120-130°C for 2-3 hours, causing the alkoxy groups in the silane molecules to condense with the hydroxyl groups on the wood fiber surface, forming a stable chemical bond. This reaction, based on the chemical reaction principle of silicon-oxygen bond formation, can significantly reduce the hydrophilicity of the wood fiber and improve the interfacial bonding strength with the plastic matrix. Finally, a cooling treatment is performed, allowing the treated fibers to cool naturally to ambient temperature at room temperature, and the moisture content is tested using an infrared moisture meter to ensure that the moisture content remains below 2%.

[0051] The specific implementation method of step S02 is to first accurately weigh the wood fiber and plastic matrix according to a mass ratio of 30:70 to 50:50, using an electronic scale for weighing, and the weighing accuracy is required to reach ±0.1g. The purpose of this step is to ensure the accuracy of the component ratio of the composite material. Then, a compatibilizer is added, and maleic anhydride grafted polypropylene or a silane coupling agent is used as a compatibilizer, and added at a ratio of 2% to 4% of the total mass. The addition of the compatibilizer is based on the interfacial compatibility theory and improves the compatibility by reducing the interfacial tension between different materials. Then, a pre-mixing process is carried out, and the weighed components are dry-mixed in a high-speed mixer at a mixing speed of 800 to 1200 r / min and a mixing time of 5 to 8 minutes to ensure that the components are evenly distributed. The premix is ​​then fed into a twin-screw extruder for melt blending. The twin-screw extruder adopts a co-rotating design, with a screw diameter of 40 to 65 mm and an aspect ratio of 32:1 to 40:1. This equipment is based on the principles of shear mixing and dispersion mixing, and achieves full mixing and plasticization of the materials through the mechanical action of the screw. During the melt blending process, the screw speed is controlled at 80 to 120 r / min and the extrusion volume is controlled at 50 to 80 kg / h. This speed range is based on the balance between the rheological properties of the material and the mixing effect. Finally, the extruded mixture is stretched into strips by a stranding machine, cooled in a cooling trough, and cut into uniform pellets of 3 to 5 mm by a pelletizer in preparation for subsequent molding.

[0052] The specific implementation method of step S03 is to first design the regional division of the extruder barrel, and divide the barrel into four independent temperature control areas along the axial direction according to the material processing flow: feeding section, plasticizing section, homogenizing section, and extrusion section. The length of each area is 150 to 200 mm. This division method is based on the state change law of the material during the extrusion process. Then, an independent temperature sensor is installed in each temperature control area, using a platinum resistance PT100 temperature sensor with a measurement accuracy of ±0.1°C and a response time of less than 5 seconds. The sensor is embedded in the barrel wall at a depth of 2 / 3 of the barrel wall thickness to ensure that the internal temperature of the barrel can be accurately detected. Then, an independent heating element is configured in each area, using a resistance heater as the main heating element. The heating power is configured according to 2 to 4kW per area, and the heater is evenly distributed around the circumference of the barrel. This configuration ensures uniform temperature distribution based on the principle of heat conduction. A temperature controller system was then established, with each zone equipped with an independent PID temperature controller, achieving a control accuracy of ±1°C. PID parameters were adjusted based on the thermal inertia characteristics of each zone, with the proportional coefficient set to 0.8-1.2, the integral time set to 60-120 seconds, and the differential time set to 15-30 seconds. A cooling system was then installed, with cooling pipes configured in each temperature-controlled zone. The cooling water flow rate was adjustable from 2 to 8 L / min. The cooling system, based on the principle of forced convection heat transfer, achieves rapid cooling by adjusting the cooling water flow rate. Finally, a temperature monitoring network was established, connecting the temperature sensors in each zone to a central control system to enable real-time acquisition, display, and recording of temperature data. The control system uses the Industrial Ethernet communication protocol, with a data transmission rate of 100 Mbps.

[0053] The specific implementation of step S04 is to first establish a multi-sensor hardware platform that integrates multiple types of sensors such as temperature sensor arrays, pressure sensors, torque sensors, photoelectric encoders, and current sensors. This platform is based on multi-sensor information fusion technology and can simultaneously collect multiple process parameters. Then, a temperature sensor array is configured, and 3 to 5 temperature sensors are installed in each of the four temperature control areas to form a temperature sensor array. The sensor spacing is 50 to 80 mm. This configuration is based on the principle of spatial distribution sampling to ensure the representativeness and accuracy of temperature measurement. Then, a pressure sensor is installed. A strain gauge pressure transmitter is installed in each of the plasticizing and homogenizing sections of the barrel. The measurement range is 0 to 50 MPa and the accuracy reaches ±0.25%. The sensor is installed on the side wall of the barrel and connected to the barrel cavity through a joint. Then, a torque sensor is configured. A torque sensor is installed between the main drive motor and the screw. The measurement range is 0 to 5000 N·m and the accuracy reaches ±0.5%. This sensor is based on the strain measurement principle and can monitor the torque load borne by the screw in real time. Next, a photoelectric encoder was installed. An incremental photoelectric encoder with a resolution of 1024 pulses per revolution was installed on the main motor shaft. Based on the principle of photoelectric conversion, this encoder measures motor speed and calculates screw speed and extrusion speed. Current sensors were then configured. Hall effect current sensors were installed on each of the three-phase power lines of the main motor. These sensors, based on the Hall effect principle, provide a non-contact measurement of motor current intensity with a measurement range of 0 to 100A and an accuracy of ±1%. Finally, a data acquisition system was established, using a high-speed data acquisition card with a sampling frequency set to every three seconds. The data acquisition system utilizes a time-triggered sampling principle to ensure synchronization and timing consistency between sensor data. The collected data is then transmitted to a host computer via a fieldbus for processing and storage.

[0054] The specific implementation of step S05 involves first establishing a historical database. This database collects process parameter and product quality data from over 500 consecutive batches of wood-plastic composite extrusion experiments. For each batch, key information such as temperature setpoints, actual temperature curves, product mechanical properties test results, and energy consumption data is recorded. This database contains multidimensional data on temperature-performance relationships, temperature-energy consumption relationships, and temperature-stability relationships, providing training and validation samples for the genetic algorithm. Next, data preprocessing is performed. The collected historical data is divided into a training set and a validation set at a ratio of 70% and 30%, respectively. The data is standardized to eliminate dimensionality effects, and outlier data points are removed using an outlier detection algorithm. This step, based on data mining preprocessing theory, ensures the quality of the algorithm input data. A genetic algorithm optimization model is then established, with a population size of 50 to 100 individuals. Each individual represents a set of temperature parameter combinations, including target temperature values ​​for the four temperature control zones. Based on biological evolution theory, the genetic algorithm searches for the optimal solution by simulating the processes of natural selection, crossover, and mutation. A fitness function was then established, using the weighted sum of product quality, energy consumption, and temperature stability as the fitness evaluation criteria. The weight coefficients were adjusted according to process requirements: the quality index had a weight of 0.5, the energy consumption index had a weight of 0.3, and the stability index had a weight of 0.2. The fitness function was based on multi-objective optimization theory, balancing various performance indicators. A genetic algorithm iteration process was then performed, with the number of iterations set between 200 and 500 generations, a crossover probability of 0.6 to 0.8, and a mutation probability of 0.05 to 0.15. Roulette selection was used as the selection operator, single-point crossover as the crossover operator, and uniform mutation as the mutation operator. The optimal temperature parameter combination was found through multi-generational evolution. The optimal target temperature for each zone was then determined. Genetic algorithm optimization revealed that the optimal target temperatures for the feeding section, the plasticizing section, the homogenizing section, and the extrusion section were 150±5°C, 170±5°C, 190±5°C, and 180±5°C, respectively. These temperature settings were based on a balance between the material's thermodynamic properties and the processing requirements. Finally, a temperature classification control strategy was established. According to the thermal degradation characteristics of wood fiber, temperature control was divided into three categories: a safe temperature range of 145-155°C, a suitable temperature range of 155-185°C, and a dangerous temperature range of 185-195°C. The proportion of the suitable temperature range was controlled to be maintained at 60-80%, the safe temperature range was 15-30%, and the dangerous temperature range was controlled at 5-15%. This strategy is based on the theory of material degradation kinetics and balances processing efficiency and material properties.

[0055] The specific implementation method of step S06 is to first establish an upper-level game model with temperature stability as the objective function. This model is based on the master-slave game theory in game theory and converts the temperature control problem into a multi-objective optimization problem. The input parameters of the objective function of the upper-level model include the actual temperature of the four temperature control areas, barrel pressure, screw torque, extrusion speed, and motor current intensity. The output is the temperature control strategy optimization plan. The objective function adopts the form of adding the square of the temperature deviation and multiplying it by the weight coefficient, plus the coupling exponential function of the barrel pressure fluctuation term and the temperature stability. The weight coefficient is determined based on the statistical analysis of historical temperature control data. Then, a lower-level game model is established with energy consumption minimization as the objective function. This model is also based on game theory and forms a master-slave relationship with the upper-level model. The input parameters of the objective function of the lower-level model are the same as those of the upper-level model, and the output is an energy-optimization control strategy. The objective function takes the form of the 1.5th power term of the temperature in the temperature control area multiplied by the corresponding energy consumption coefficient, plus the square of the motor current intensity divided by the efficiency coefficient, and finally subtracting the coupled logarithmic function of the barrel pressure and energy consumption optimization. This function form is based on energy consumption modeling theory and historical data statistical analysis results. A two-level game solution algorithm is then established, using an iterative solution method. The upper-level model first provides a temperature control strategy, and the lower-level model performs energy consumption optimization calculations based on this strategy. The optimization results are then fed back to the upper-level model to adjust the temperature control parameters. This solution process is based on two-level programming theory and converges to the optimal solution through iteration. The game model startup condition is then set. When the actual temperature of any temperature control area deviates from the target temperature by more than ±2°C, the game optimization adjustment is initiated. The threshold is determined based on the temperature control accuracy requirements and system response characteristics. Next, a model parameter update mechanism was established. Based on real-time process parameters and product quality feedback, the game model's weight coefficients and coupling function parameters were regularly updated every 100 batches of product. This mechanism, based on adaptive control theory, ensured the model's adaptability and accuracy. Finally, a control strategy execution system was established to convert the optimized control strategy output by the game model into specific control instructions, including adjustments to heating power in each zone, cooling water flow, and screw speed fine-tuning. These control instructions were executed via a PLC control system, with a response time of less than 5 seconds.

[0056] The specific implementation method of step S07 is to first establish an intelligent monitoring hardware platform, integrating multiple monitoring equipment such as thermal imagers, contact temperature sensors, non-contact infrared thermometers, pressure sensors, and online rheometers. The platform is based on multimodal information fusion technology to achieve all-round monitoring of the process. Then configure the thermal imager monitoring system, using a long-wave infrared detector with a wavelength of 8 to 14 microns, with a temperature resolution of 0.1°C and a spatial resolution of 0.5 milliradians. The device is based on the principle of thermal radiation detection and is used to detect the uniformity of the temperature distribution on the barrel surface. The thermal imager is installed 1 to 2 meters above the barrel, and the scanning angle covers the entire length of the barrel. Then configure a contact temperature sensor array, embedding multiple temperature sensors on the inner wall of the barrel to monitor the actual temperature of the melt. The sensor is encapsulated with high-temperature resistant materials and can withstand temperatures above 200°C. This configuration is based on the principle of direct contact temperature measurement to ensure the accuracy of temperature measurement. Subsequently, an online rheometer is installed, and a capillary rheometer is installed at the extruder head to measure the change in melt viscosity in real time. The measurement range is 10 to 10 5 The device measures viscosity in Pa·s, achieving a measurement accuracy of ±2%. Based on the capillary flow principle, it calculates viscosity by measuring pressure drop and flow rate. A data feature analysis system was then established, employing a similarity calculation algorithm to analyze the characteristics of continuously collected process parameters. When the standard deviation of continuously collected temperature data is less than 1°C, indicating temperature stability, the sampling interval is increased to every 10 seconds. This determination is based on the statistical principle of variability analysis. An adaptive sampling mechanism was then established. When temperature data show a sudden change or the similarity coefficient falls below 0.9, the sampling frequency is increased to every 1 second, and an early warning mechanism is activated. The similarity coefficient is calculated using Pearson correlation analysis. Based on adaptive sampling theory, this mechanism dynamically adjusts the monitoring frequency based on process changes. The temperature uniformity index (TUI) was then calculated, defined as the ratio of the temperature standard deviation to the mean temperature minus 1. This index is based on the coefficient of variation. A TUI closer to 1 indicates a more uniform temperature distribution. An index below 0.85 indicates significant localized overheating or undercooling. Finally, a process parameter adjustment system was established to iteratively adjust the three control parameters: screw speed, heating power distribution, and cooling water flow rate according to the numerical changes in the temperature uniformity index. When the temperature uniformity index was in the range of 0.8 to 0.85, the screw speed was adjusted to the current speed multiplied by 0.95, and the heating power was adjusted to the current power minus 5%. When the index was lower than 0.8, the screw speed was adjusted to the current speed divided by 1.1, the heating power was adjusted to the current power multiplied by 0.9, and the cooling water flow rate was adjusted to the current flow rate plus 20%. The adjustment process adopted an iterative control method. After each adjustment, the temperature uniformity index was recalculated after waiting 30 seconds until the index reached above 0.85.

[0057] The specific implementation of step S08 is to first establish an online quality inspection system, setting up three inspection stations for surface finish inspection, dimensional accuracy inspection, and mechanical property inspection after the cooling section of the extrusion production line. The system is based on online non-destructive testing technology to achieve real-time monitoring of product quality. Then, a surface finish inspection device is configured, using a laser surface roughness measuring instrument with a laser wavelength of 650nm and a measurement range of Ra0.1~50μm. The device calculates the surface roughness by analyzing the intensity distribution of surface scattered light based on the principle of laser scattering. The detection head is 10~20mm away from the product surface, the scanning speed is 50~100mm / s, and the qualified threshold range is set to Ra≤3.2μm. Then, a dimensional accuracy inspection device is installed, using a laser diameter gauge to continuously measure the outer diameter of the extruded product with a measurement accuracy of ±0.01mm. The device is based on the principle of laser triangulation and calculates the size by detecting the position change after the laser beam is blocked by the product. The measurement frequency is 10 times per second, and the qualified threshold range is set to the design size ±0.1mm. Subsequently, a mechanical properties testing system was configured, using an online tensile testing machine to test the mechanical properties of the products. The testing machine has a maximum tensile force of 10kN and a displacement accuracy of 0.01mm. This equipment is based on the principle of material mechanics testing and measures the stress-strain relationship of the material by applying a tensile load. The qualified threshold for tensile strength is set at ≥25MPa, and the qualified threshold for flexural strength is set at ≥40MPa. A quality data processing system was then established to perform real-time analysis and processing of the quality data collected at each testing station. A control chart was established using statistical process control methods to monitor the changing trends of quality parameters. An alarm signal was issued when the quality parameters exceeded the control limit. Based on the theory of quality statistics, the system uses control chart analysis to determine whether the process is in a controlled state. A quality judgment and sorting system was then established to make a comprehensive judgment based on the results of various quality tests. Products that meet all the quality indicators are judged to be qualified and enter the subsequent cooling and shaping process. Products that do not meet any of the requirements are judged to be unqualified and separated from the production line by an automatic sorting device. This judgment is based on the principle of logical AND operation to ensure strict control of product quality. Finally, a quality feedback control system is established to feed back the quality inspection results to the process control system, and the extrusion process parameters are automatically adjusted according to the quality deviation. For example, if the surface finish does not meet the standard, the extrusion temperature is appropriately lowered or the pulling speed is adjusted. If the dimensional accuracy is found to be out of tolerance, the extruder pulling speed and mold temperature are adjusted. If the mechanical properties do not meet the standard, the mixing ratio or the amount of modifier added is adjusted. This feedback control is based on closed-loop control theory to achieve continuous improvement of product quality and process optimization.

[0058] It should be noted that the segmented temperature control system divides the extruder barrel into four independent control areas: feeding section, plasticizing section, homogenizing section, and extrusion section. Each area is equipped with independent temperature sensors and heating elements, which has significant technical advantages over the traditional overall temperature control method. The traditional method uses a unified temperature set value for full barrel control, which cannot adapt to the differentiated heat treatment requirements of wood fiber and plastic matrix at different processing stages, resulting in uneven temperature distribution and local overheating. The segmented control system achieves refined management of the material heat treatment process by setting targeted optimal target temperatures for each area. The lower temperature of the feeding section avoids premature thermal degradation of wood fiber, the moderate temperatures of the plasticizing and homogenizing sections ensure sufficient melting and uniform mixing of the plastic matrix, and the optimized temperature of the extrusion section ensures molding quality, effectively solving the problem of material property fluctuations caused by the extensive temperature control of traditional methods.

[0059] The temperature classification control strategy divides temperature control into three categories: safe, suitable, and dangerous, based on the thermal degradation characteristics of wood fiber. By controlling the proportion of each temperature category, it balances processing efficiency and material performance, which is an essential technical improvement compared to the traditional single temperature control mode. Traditional control methods only focus on maintaining temperature values, ignoring the stage characteristics of wood fiber thermal degradation and the gradual process of plastic matrix melting. This easily leads to a dilemma where either the temperature is too low, resulting in insufficient plasticization, or the temperature is too high, resulting in a large amount of fiber degradation. The temperature classification control strategy scientifically divides the temperature intervals and controls the proportion of each interval, so that most of the processing process is carried out within the appropriate temperature range. This not only ensures the sufficient melting of the plastic matrix, but also controls the degradation of wood fibers within an acceptable range, achieving an organic unity of the processing needs of the two materials.

[0060] The two-layer game model constructs an upper-layer game with the goal of temperature stability and a lower-layer game with the goal of minimizing energy consumption. It achieves multi-objective collaborative optimization through the Nash equilibrium principle of game theory, and has significant technical advantages over traditional single-objective control methods. Traditional control systems usually only consider one of the objectives of temperature maintenance or energy consumption control, and cannot handle the contradictory relationship between temperature stability and energy consumption optimization, resulting in limited control effects. By establishing the upper and lower-layer game relationship, the two-layer game model enables the temperature control strategy to consider energy minimization while pursuing stability, and the energy consumption optimization strategy to ensure temperature control effects while reducing consumption. The two objectives reach a dynamic balance through the game process, achieving global optimization of the control system.

[0061] The synergistic effect of three key technical approaches—segmented temperature control, temperature classification control strategy, and a two-tiered game model—has created a multi-dimensional, multi-level intelligent temperature control system, achieving a qualitative leap compared to traditional single-control methods. Segmented control provides a spatial basis for temperature classification control, enabling different regions to adopt different temperature classification strategies based on material properties; temperature classification control provides a decision-making basis for the two-tiered game model, making the game optimization process more accurate and effective; and the two-tiered game model provides an intelligent decision-making mechanism for the entire control system, enabling dynamic optimization of temperature control. The three elements support and reinforce each other, forming an organic technical whole that enables precise control within the narrow processing temperature window of wood fiber and plastic matrix. This thoroughly resolves the technical difficulty of traditional methods in balancing the different heat treatment requirements of the two materials, ensuring the stability and consistency of the finished wood-plastic composite material.

[0062] The present invention pre-treats the wood fiber raw material with a silane coupling agent for surface modification and controls the moisture content and particle size, establishes a segmented temperature control system to divide the extruder barrel into four independent temperature control zones and sets targeted optimal target temperatures, adopts a multi-sensor data acquisition platform to monitor five process parameters in real time, uses genetic algorithm optimization and temperature classification control strategies to balance the thermal stability of the wood fiber and the melting effect of the plastic matrix, constructs a double-layer game model to achieve coordinated optimization of temperature stability and energy consumption, and uses an intelligent multi-parameter monitoring platform to iteratively adjust parameters according to the temperature uniformity index. Finally, online quality inspection of the extruded product is performed, thereby solving the technical problem that wood fiber is prone to thermal degradation at high temperature, while the plastic matrix requires a sufficiently high temperature to fully melt, and the processing temperature window of the two is very narrow, which easily causes local overheating and leads to insufficient performance of the finished wood-plastic composite material.

[0063] Specifically, the principle of the present invention is: the present invention can solve the technical problem that wood fiber is prone to thermal degradation at high temperature, while the plastic matrix requires a sufficiently high temperature to fully melt. The processing temperature window of the two is very narrow, which easily causes local overheating and leads to insufficient performance of the finished wood-plastic composite material. Its technical principle is to achieve precise matching of the different heat treatment requirements of wood fiber and plastic matrix through multi-dimensional temperature control and intelligent optimization algorithm. First, the segmented temperature control system divides the extruder barrel into four independent control areas: feeding section, plasticizing section, homogenizing section, and extrusion section. Each area sets a different optimal target temperature according to the heat treatment characteristics of the material. The feeding section is 150±5℃ mainly used for preheating and moisture removal of wood fiber, the plasticizing section is 170±5℃ to achieve preliminary melting of the plastic matrix, the homogenizing section is 190±5℃ to ensure that the plastic matrix is ​​fully melted and evenly mixed with the wood fiber, and the extrusion section is 180±5℃ to maintain an appropriate extrusion temperature to avoid excessive thermal degradation.

[0064] The temperature classification control strategy establishes a scientific temperature distribution management system based on the thermal degradation characteristics of wood fiber and the melting characteristics of the plastic matrix. The safe temperature range of 145-155°C corresponds to a fiber degradation rate of 0-2%, the suitable temperature range of 155-185°C corresponds to a fiber degradation rate of 2-8%, and the dangerous temperature range of 185-195°C corresponds to a fiber degradation rate of 8-15%. By controlling the suitable temperature range to account for 60-80%, sufficient melting of the plastic matrix is ​​ensured while keeping wood fiber degradation within an acceptable range. A two-layer game model further optimizes the temperature control strategy. The upper-layer game model minimizes temperature deviations in each region with the goal of temperature stability, while the lower-layer game model optimizes heating power distribution with the goal of minimizing energy consumption. The two achieve collaborative optimization through the Nash equilibrium principle of game theory.

[0065] The intelligent multi-parameter monitoring platform uses a thermal imager to detect the uniformity of barrel surface temperature distribution, a contact temperature sensor to monitor the actual melt temperature, and an online rheometer to measure changes in melt viscosity. This multi-dimensional monitoring ensures precise control of temperature distribution. When the temperature uniformity index falls below 0.85, the system automatically adjusts the screw speed, heating power distribution, and cooling water flow rate to eliminate localized overheating. This multi-level intelligent control system precisely balances the different heat treatment requirements of the wood fiber and plastic matrix, achieving optimal processing results within a narrow temperature window.

[0066] A specific embodiment 1 of the present invention is provided below. The specific implementation of each step in this embodiment 1 is described in detail as follows.

[0067] The specific implementation of step S01 in this embodiment is the same as that described above and will not be described in detail here.

[0068] The specific implementation of step S02 is to perform accurate material calculation, and the mass ratio of wood fiber to plastic matrix is ​​determined by the following formula:

[0069]

[0070] Where R is the mass ratio of wood fiber to plastic matrix; m f is the mass of wood fiber, in kg; m p R is the mass of the plastic matrix, in kg. The ratio range is controlled within 0.3≤R≤1.0, i.e. a mass ratio of 30:70 to 50:50. The amount of compatibilizer added is calculated based on the total mass fraction:

[0071]

[0072] Where, C is the mass fraction of the compatibilizer, ranging from 2% to 4%; m c is the mass of the compatibilizer, in kg. fObtained by electronic weighing, weighing accuracy ±0.1g; m p Obtained by electronic weighing, weighing accuracy ±0.1g; m c Calculated based on the set mass fraction C and total mass.

[0073] The specific implementation of steps S03-S04 is the same as above and will not be repeated here.

[0074] The specific implementation of step S05 is to establish a multi-objective optimization model based on a genetic algorithm, and the fitness function is expressed as:

[0075] F = w1·Q+w2·E+w3·S;

[0076] Where F is the comprehensive fitness value; Q is the product quality index, dimensionless; E is the energy consumption index, unit is kJ / kg; S is the temperature stability index, dimensionless, calculated by the temperature fluctuation coefficient where σ temp is the process temperature standard deviation, in °C, obtained by statistical calculation of the temperature data of each temperature control area in 500 batches of experiments. is the average temperature of the process, in °C, obtained by the arithmetic mean of the corresponding experimental data; w1, w2, and w3 are weight coefficients, which are 0.5, 0.3, and 0.2 respectively. The product quality index Q is calculated by standardizing the mechanical properties test results:

[0077]

[0078] Where σ t is the measured value of tensile strength, in MPa; σ b is the measured value of bending strength, in MPa; α1 and α2 are weight factors, both of which are 0.5. The energy consumption index E is calculated by the heating power and motor power consumption of each area:

[0079]

[0080] Where, P h，i is the heating power of the ith temperature control area, in kW; t i is the heating time of the corresponding area, in h; P m is the motor power in kW; t m is the motor running time, in h; m total is the total output, in kg. t and σ b Obtained by measuring with an online tensile testing machine; P h，i Obtained by measuring with a power meter; P m Obtained through the motor power monitoring device.

[0081] The specific implementation of step S06 is to establish a two-layer game optimization model, and the objective function of the upper-layer game model is:

[0082]

[0083] Where, J1 is the upper layer objective function value; T i，actual is the actual temperature of the ith temperature control area, in °C; T i，target is the target temperature of the ith temperature control area, in °C; β i is the temperature deviation weight coefficient; P barrel is the barrel pressure, in MPa; ΔT std is the temperature standard deviation, in °C; is the average temperature of the temperature control area, in °C; γ is the coupling coefficient, ranging from 0.1 to 0.3, determined by analyzing the correlation between temperature and pressure in historical data. The objective function of the lower-level game model is:

[0084]

[0085] Where J2 is the lower layer objective function value; λ i is the energy consumption coefficient of the ith region, in kJ / (℃ 1.5 kg), which is obtained by fitting historical energy consumption data. The calculation method is: Among them E i，avg is the average energy consumption of the i-th region, obtained by statistically analyzing the power meter measurement data of the corresponding region in 500 batches of experiments, T i，avg is the average temperature of the i-th region, obtained by statistically analyzing the temperature sensor measurement data in the corresponding experiment; T i Equivalent to T i，actual , is the actual temperature of the ith temperature control area, in °C; I is the motor current intensity, in A; μ is the current power consumption coefficient, in kJ / (A 2 kg), obtained through the motor power consumption calibration experiment, the calibration steps include: Step 1: measuring the relationship between motor current and power consumption under different loads; Step 2: fitting the μ value, the typical value is 0.02-0.05; η is the motor efficiency coefficient, the value range is 0.85-0.95; δ is the pressure coupling coefficient, the value range is 0.05-0.15, determined by fitting the experimental data of the correlation between energy consumption and pressure; P ref is the reference pressure, which is 10 MPa. i，actual Obtained through real-time measurement of temperature sensor array; P barrel Obtained by measuring with a pressure sensor; I obtained by measuring with a current sensor; β iThe calculation method is determined by statistical analysis of historical data and is the inverse of the difficulty of temperature control in each area. The specific calculation steps include: Step 1: Calculate the temperature fluctuation variance of each temperature control area in 500 batches of experiments Step 2: Calculate the temperature control difficulty coefficient in is the reference variance, which is 4℃ 2 ; Step 3: Calculate the weight coefficient The value range is 0.2 to 2.0.

[0086] The specific implementation of step S07 is to establish an intelligent monitoring algorithm. The temperature uniformity index calculation formula is:

[0087]

[0088] Where U is the temperature uniformity index, dimensionless; σ T is the temperature standard deviation, in °C; is the average temperature in °C. The temperature standard deviation is calculated as:

[0089]

[0090] Where n is the number of temperature measurement points; T j is the temperature value of the jth measurement point, in °C. The average temperature calculation formula is:

[0091]

[0092] The similarity of data features is calculated using the Pearson correlation coefficient:

[0093]

[0094] Where ρ is the similarity coefficient; x k 、y k is the temperature data of the continuous time series; m is the number of data points; are the means of the corresponding sequences respectively. The process parameter adaptive adjustment algorithm is:

[0095] When 0.8≤U<0.85, screw speed adjustment: N new =N current ×0.95; Heating power adjustment: P new =P current ×0.95;

[0096] When U<0.8, screw speed adjustment: Heating power adjustment: P new =P current ×0.9; Cooling water flow adjustment: Q new =Qcurrent ×1.2;

[0097] Where N current is the current screw speed, in r / min; N new is the screw speed after adjustment, in r / min; P current is the current heating power, in kW; P new is the adjusted heating power in kW; Q current is the current cooling water flow rate, in L / min; Q new is the cooling water flow rate after adjustment, in L / min. j Acquired by measuring with thermal imager and temperature sensor array; N current Obtained by measuring with a photoelectric encoder; P current Obtained by measuring with a power meter; Q current Obtained through flow meter measurement.

[0098] The specific implementation of step S08 is to establish an online quality detection algorithm, and the surface finish evaluation is calculated by roughness parameters:

[0099]

[0100] Where Ra is the arithmetic mean roughness, in μm; L is the measured length, in mm; z(x) is the surface profile height function, in μm. The dimensional accuracy deviation calculation formula is:

[0101] ΔD=D measured -D design ;

[0102] Where ΔD is the dimensional deviation, in mm; D measured D is the measured outer diameter in mm; design is the design outer diameter, in mm. Calculation of comprehensive evaluation index of mechanical properties:

[0103]

[0104] Where, M is the comprehensive index of mechanical properties, dimensionless; σ t，measured is the measured tensile strength, in MPa; σ b，measured is the measured bending strength, in MPa. The comprehensive quality judgment function is:

[0105] Q total =ω1f1(Ra)+ω2f2(ΔD)+ω3f3(M);

[0106] Where Q totalis the comprehensive quality evaluation value; f1(Ra) is the surface finish evaluation function; f2(ΔD) is the dimensional accuracy evaluation function; f3(M) is the mechanical performance evaluation function; ω1, ω2, and ω3 are weight coefficients, which are 0.2, 0.3, and 0.5 respectively. The evaluation function is defined as:

[0107]

[0108] When Q total =1 is judged as qualified product, when Q total When the value is less than 1, it is judged as unqualified. Among them, Ra is measured by a laser surface roughness measuring instrument. The measurement steps include: Step 1: Adjust the distance between the laser probe and the product surface to 15mm; Step 2: Set the scanning length to 25mm and the scanning speed to 75mm / s; Step 3: Collect surface profile data and calculate the Ra value. measured It is obtained by measuring with a laser diameter gauge. The measurement steps include: Step 1: Calibrate the zero point of the laser diameter gauge; Step 2: Continuously scan the outer diameter of the product with a sampling frequency of 10Hz; Step 3: Calculate the average outer diameter value. t，measured and σ b，measured The strength was obtained by measuring with an online tensile testing machine. The measurement steps included: Step 1: cutting a standard specimen with a length of 50 mm; Step 2: setting a tensile speed of 5 mm / min for a tensile test; Step 3: recording the maximum tensile force and bending force, and calculating the strength value.

[0109] It should be noted that, in this embodiment, the product quality index and energy consumption indicators Based on standardized evaluation theory and energy balance principles, mechanical properties and energy consumption data are converted into dimensionless indicators for unified comparison. Compared to existing absolute numerical evaluation methods, this standardized process eliminates the impact of scale differences between product batches, improves the comparability and stability of process parameter optimization, and enables the genetic algorithm to more accurately identify the optimal combination of process conditions.

[0110] Temperature stability index Based on the coefficient of variation principle, temperature control accuracy is evaluated through relative change. Compared to existing absolute temperature deviation evaluation methods, this metric better reflects control stability at different temperature levels, avoids the situation where high-temperature areas mask control accuracy issues due to large base numbers, and improves the temperature control system's sensitivity and accuracy to process stability.

[0111] Two-level game model objective function and Based on game theory and nonlinear optimization theory, a master-slave game relationship between temperature stability and energy consumption minimization was established. Compared with existing single-layer optimization control methods, this model can better address the conflicting relationship between temperature control accuracy and energy consumption. By iteratively searching for the Nash equilibrium point between these two objectives, it achieves the goal of minimizing system energy consumption while maintaining temperature control accuracy, significantly improving production efficiency and economic benefits.

[0112] Temperature Uniformity Index Based on the principles of statistical variation analysis, this index quantitatively evaluates the uniformity of temperature distribution. Compared to existing qualitative observation and local measurement methods, this index provides a global quantitative assessment of temperature distribution uniformity. It can promptly detect local overheating or undercooling, guide the precise adjustment of process parameters, and effectively avoid product quality defects and material degradation caused by uneven temperature distribution.

[0113] Data feature similarity formula Based on Pearson correlation analysis theory, this algorithm achieves intelligent identification of process parameter trends. Compared to existing fixed-frequency sampling methods, this algorithm can dynamically adjust the monitoring frequency based on process changes. When the process is stable, the sampling frequency is reduced to save computing resources, while when the process is abnormal, the sampling frequency is increased to ensure monitoring accuracy. This achieves adaptive optimization of the monitoring system and rational allocation of resources.

[0114] The adaptive process parameter adjustment algorithm uses a piecewise function to implement a nonlinear mapping between the parameter adjustment amplitude and the temperature uniformity index. Compared to existing linear adjustment methods, this algorithm can adopt different adjustment strategies based on the degree of deviation. It can perform fine-tuning to avoid system oscillation when the deviation is small, and perform rapid adjustment to ensure system stability when the deviation is large. This significantly improves the control system's response speed and adjustment accuracy.

[0115] In online quality inspection, the surface finish formula Based on surface morphology analysis theory, comprehensive evaluation index of mechanical properties Based on the weighted evaluation method, the quality comprehensive judgment function Q total =ω1f1(Ra)+ω2f2(ΔD)+ω3f3(M) is based on multi-criteria decision-making theory. Compared to existing offline sampling methods, this quality evaluation system achieves real-time, comprehensive evaluation and automatic determination of product quality. Through continuous monitoring and timely feedback control, it effectively improves product qualification rates and production efficiency, while reducing defective product rates and raw material waste.

[0116] To better understand and implement the present invention, Example 2 of a specific application scenario is provided below: A materials engineering research team was tasked with producing high-performance wood-plastic composite extruded profiles for the outdoor construction industry. These profiles required excellent mechanical properties, good surface quality, and low production energy consumption. Conventional extrusion processes presented technical challenges such as low temperature control precision, large fluctuations in product quality, and high energy consumption. The research team decided to address these challenges using the present invention's method for optimizing continuous extrusion molding of wood-plastic composites.

[0117] The research team first performed raw material pretreatment and ingredient design. They selected poplar wood fiber as the raw material, sieved it to a 60-mesh particle size using a vibrating screen, and dried it in a vacuum oven at 85°C for 5 hours to reduce the moisture content to 1.8%. They then used γ-aminopropyltriethoxysilane as a coupling agent, at a ratio of 2.5% by weight to the fiber surface, and heat-treated it at 125°C for 2.5 hours. Polypropylene was used as the plastic matrix, and maleic anhydride-grafted polypropylene was used as the compatibilizer. The specific ingredient ratios are shown in Table 1.

[0118] Table 1 Raw material ratio parameters

[0119] Components Mass ratio Quality score Remark Poplar fiber 40 38.5％ Silane modified Polypropylene 60 57.7％ Base material Compatibilizer - 3.8％ Maleic anhydride grafted polypropylene

[0120] The research team established a segmented temperature control system, dividing the twin-screw extruder barrel into four independent temperature control zones. Each zone is 180 mm long and equipped with a PT100 temperature sensor and a 3kW resistance heater. A multi-sensor data acquisition platform was built, integrating a temperature sensor array, a strain gauge pressure sensor, a torque sensor, a photoelectric encoder, and a Hall effect current sensor. The data acquisition frequency was set to every three seconds. The technical parameters of each sensor are shown in Table 2.

[0121] Table 2 Sensor configuration parameters

[0122] Sensor Type Model specifications Measuring range Accuracy Installation location Temperature sensor PT100 -50～300℃ ±0.1℃ Each temperature control area pressure sensor Strain gauge 0~50MPa ±0.25％ Plasticizing section, homogenizing section Torque sensor Strain gauge 0~5000N·m ±0.5％ Main motor shaft Photoelectric encoder Incremental 1024 pulses / revolution - Motor shaft end Current sensor Hall type 0～100A ±1％ Three-phase power cord

[0123] The research team used a genetic algorithm to optimize temperature control parameters. They established a multidimensional database containing 500 batches of historical data, using 70% of the data as a training set and 30% as a validation set. They set the population size to 80 individuals, the number of iterations to 300 generations, the crossover probability to 0.7, and the mutation probability to 0.1. The fitness function weight coefficients w1 = 0.5, w2 = 0.3, and w3 = 0.2. After genetic algorithm optimization, the optimal target temperatures for each temperature control zone were obtained, as shown in Table 3.

[0124] Table 3. Temperature parameters optimized by genetic algorithm

[0125] Temperature control area Optimal target temperature Control accuracy <![CDATA[Weight coefficient β i > Feeding section 152℃ ±3℃ 0.8 Plasticizing section 173℃ ±3℃ 1.2 Homogenization section 187℃ ±3℃ 1.5 Extrusion section 178℃ ±3℃ 1.0

[0126] Figure 2 The figure shows the convergence of the optimized temperature parameters in the four temperature control zones during 300 iterations of the genetic algorithm. The solid black line represents the temperature change in the feeding zone, which gradually converged from an initial 165°C to the optimal target temperature of 152°C; the dashed black line represents the temperature change in the plasticizing zone, which converged from 185°C to 173°C; the dashed black line represents the temperature change in the homogenization zone, which converged from 200°C to 187°C; and the dotted black line represents the temperature change in the extrusion zone, which converged from 190°C to 178°C. The gray horizontal lines indicate the optimal target temperature for each zone.

[0127] The research team established a two-layer game model to fine-tune and optimize temperature control. The upper-layer game model targets temperature stability, with a coupling coefficient of γ = 0.2. The lower-layer game model aims to minimize energy consumption, with energy consumption coefficients λ1 = 0.15, λ2 = 0.18, λ3 = 0.22, and λ4 = 0.16, current power consumption coefficient μ = 0.03, efficiency coefficient η = 0.9, and pressure coupling coefficient δ = 0.1. Game optimization is initiated when the temperature deviation exceeds ±2°C. Figure 3 The figure depicts the convergence process of the two-layer game model over 50 iterations. The solid black line in the figure represents the change in the upper-layer game objective function J1, which gradually converges from an initial value of 15.2 to 2.1, reflecting the improvement in temperature stability. The dashed black line represents the change in the lower-layer game objective function J2, which converges from 28.5 to 8.3, demonstrating the energy consumption optimization effect. The black dotted line corresponding to the right y-axis shows the temperature deviation decreasing from 4.2°C to 0.5°C. This figure demonstrates how the two-layer game model achieves coordinated optimization of temperature control accuracy and energy consumption minimization through a master-slave game relationship. Table 4 shows a comparison of the game model before and after optimization.

[0128] Table 4 Comparison of game model optimization effects

[0129] Control parameters Before optimization After optimization Improvement Temperature stability index 0.78 0.91 16.7％ System energy consumption 285kJ / kg 251kJ / kg 11.9％ Temperature deviation ±4.2℃ ±1.8℃ 57.1％ Control response time 12 seconds 6 seconds 50.0％

[0130] The research team configured an intelligent multi-parameter monitoring platform and used a long-wave infrared thermal imager to monitor the temperature distribution on the barrel surface with a temperature resolution of 0.1°C. The online rheometer measures the melt viscosity in real time with a measurement range of 10 2 ~10 5 Pa·s. A data feature similarity analysis algorithm was established. When the temperature data standard deviation was less than 1°C, the collection interval was extended to 10 seconds. When the similarity coefficient fell below 0.9, the collection frequency was increased to 1 second and an early warning was initiated. Temperature uniformity index calculations showed that when the system was operating stably, the U value remained above 0.87. When the U value dropped to 0.82, the system automatically adjusted the screw speed and heating power distribution.

[0131] Figure 4This chart records the real-time changes in the temperature uniformity index (TUI) over 72 hours of continuous production. The black curve shows that the TUI fluctuated around 0.87, maintaining a good overall level. The gray dashed line indicates the control threshold of 0.85; when the TUI falls below this threshold, the system automatically initiates parameter adjustments. The three gray vertical dotted lines in the figure mark the automatic adjustment events that occurred at 10.5 hours, 38.25 hours, and 56.8 hours. Arrows indicate the adjustment times, demonstrating the intelligent monitoring system's real-time responsiveness and adaptive adjustment effectiveness.

[0132] The research team implemented an online quality inspection system, installing a laser surface roughness meter, a laser diameter gauge, and an online tensile testing machine after the cooling section. Surface finish testing was performed with a scan length of 25 mm and a scan speed of 75 mm / s. Dimensional accuracy testing was performed with a sampling frequency of 10 Hz, continuously monitoring changes in outer diameter. Mechanical properties were tested using 50 mm standard specimens at a tensile speed of 5 mm / min. The quality inspection results are shown in Table 5.

[0133] Table 5 Product quality test results

[0134]

[0135]

[0136] Figure 5 This chart comprehensively displays the test results for four key quality indicators. The black bar chart on the left y-axis shows the measured values, while the gray bar chart shows the pass thresholds. The numerical labels clearly indicate the specific data. The black line chart on the right y-axis shows the pass rate for each indicator, ranging from 94.5% to 97.2%. This chart comprehensively demonstrates the monitoring effectiveness of the online quality inspection system and verifies the impact of process optimization methods on improving product quality.

[0137] During the actual production process, at 10:30 am on the first day, the monitoring system detected that the temperature uniformity index of the homogenization section dropped to 0.81. The system automatically adjusted the screw speed from 95r / min to 86r / min, the heating power from 2.8kW to 2.5kW, and the cooling water flow from 5.2L / min to 6.2L / min. After 30 seconds of adjustment, the temperature uniformity index returned to 0.86. At 14:15 pm the next day, the pressure sensor detected that the barrel pressure suddenly rose to 28MPa and the temperature deviation reached +3.1℃. The two-layer game model immediately started the optimization adjustment and obtained the optimal control strategy through iterative calculation. The heating power of each area was adjusted to 2.6kW, 2.9kW, 3.1kW, and 2.7kW respectively. At the same time, the screw speed was fine-tuned to 92r / min. After 5 seconds, the system returned to a stable state.

[0138] After 72 hours of continuous production, the research team conducted a comprehensive evaluation of the entire process. Temperature control accuracy remained stable within ±1.8°C, a 10% improvement over the target of ±2°C. The system's average energy consumption was 251 kJ / kg, an 11.9% reduction from the initial design target. The overall product qualification rate reached 94.1%, including a 96.8% surface quality qualification rate, a 94.5% dimensional accuracy qualification rate, and a 96.6% mechanical performance qualification rate. Throughout the production process, the system automatically adjusted 47 times, with only three manual interventions, achieving highly intelligent control.

[0139] Traditional wood-plastic composite extrusion processes primarily utilize a segmented, fixed-value temperature control method, using PID controllers to maintain constant temperatures in each zone. Process parameter adjustments rely on operator experience, and quality testing is performed using offline spot checks. Traditional methods suffer from technical drawbacks such as low temperature control accuracy, difficulty optimizing parameters, and delayed quality monitoring, leading to significant product quality fluctuations, low production efficiency, and high energy consumption. This present invention utilizes advanced technologies such as genetic algorithm multi-objective optimization, real-time fine-tuning using a two-layer game model, intelligent multi-parameter monitoring, and online quality testing, achieving significant technological advancements over traditional methods. Temperature control accuracy is improved from ±4.2°C to ±1.8°C, a 57.1% improvement. System energy consumption is reduced from 285kJ / kg to 251kJ / kg, achieving an 11.9% energy saving. The overall product qualification rate is increased from 82.3% to 94.1%, a 14.3% improvement in quality. The temperature uniformity index is increased from 0.78 to 0.91, a 16.7% improvement in uniformity. The control response time is shortened from 12 seconds in traditional methods to 6 seconds, a 50% increase in response speed. These technological advances demonstrate that the present invention effectively solves key technical difficulties in traditional wood-plastic composite extrusion processes and provides an advanced technical solution for the industrial production of high-quality wood-plastic composites.

[0140] It should be noted that the detailed explanation of the variables involved in the present invention is shown in Table 6.

[0141] Table 6 Variable explanation table

[0142]

[0143]

[0144] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed by the present invention, which should be covered by the scope of protection of the present invention.

Claims

1. A process optimization method for continuous extrusion molding of wood-plastic composite materials, characterized in that: include: The wood fiber raw material was surface-modified and pretreated with a silane coupling agent to control the moisture content of the wood fiber to below 2% and the particle size to be within the range of 40-80 mesh. The wood fiber and plastic matrix were mixed in a mass ratio of 30:70-50:50, and a compatibilizer with a mass fraction of 2%-4% was added, and melt blending was carried out through a twin-screw extruder. A segmented temperature control system was established, and the extruder barrel was divided into four temperature control zones according to the feeding zone, plasticizing zone, homogenizing zone, and extrusion zone. Each zone was equipped with an independent temperature sensor and heating element. Five process parameters, including temperature in the temperature control zone, barrel pressure, screw torque, extrusion speed, and motor current intensity, were collected through a multi-sensor data acquisition platform. The optimal target temperature of each temperature control zone was determined based on genetic algorithm optimization, and a temperature classification control strategy was established. An upper-level game model with temperature stability as the goal and a lower-level game model with energy consumption minimization as the goal were established to achieve real-time fine-tuning of temperature control. The barrel temperature distribution and process parameter changes were monitored in real time through an intelligent multi-parameter monitoring platform, and the acquisition parameters were iteratively optimized based on the similarity of data features. Online quality inspection of extruded wood-plastic composites, including surface finish, dimensional accuracy and mechanical properties.

2. The process optimization method for continuous extrusion molding of wood-plastic composite materials according to claim 1, characterized in that: The surface modification treatment with silane coupling agent refers to chemical modification of the wood fiber surface using silane coupling agent, through which the alkoxy groups in the silane molecules react with the hydroxyl groups on the wood fiber surface to form stable chemical bonds, thereby reducing the hydrophilicity of the wood fiber and improving the interfacial bonding strength with the plastic matrix.

3. The process optimization method for continuous extrusion molding of wood-plastic composite materials according to claim 2, characterized in that: The segmented temperature control system refers to dividing the extruder barrel into multiple independent temperature control zones along the axial direction. Each zone is equipped with an independent temperature sensor and heating element to achieve precise control of the temperature of different zones and avoid local overheating or overcooling caused by overall temperature control.

4. The process optimization method for continuous extrusion molding of wood-plastic composite materials according to claim 3, characterized in that: The multi-sensor data acquisition platform refers to a comprehensive monitoring device that integrates a temperature sensor array, a pressure sensor, a torque sensor, a photoelectric encoder, and a current sensor. The temperature sensor uses a platinum resistance PT100 type with a measurement accuracy of ±0.1°C. The pressure sensor uses a strain gauge pressure transmitter with a measurement range of 0 to 50 MPa and an accuracy of ±0.25%. The acquisition frequency is once every 3 seconds.

5. The process optimization method for continuous extrusion molding of wood-plastic composite materials according to claim 4, characterized in that: The genetic algorithm optimizes and determines the steps of the optimal target temperature of each temperature control zone, specifically, the optimal target temperature of the feeding section is 150±5°C, the optimal target temperature of the plasticizing section is 170±5°C, the optimal target temperature of the homogenizing section is 190±5°C, and the optimal target temperature of the extrusion section is 180±5°C.

6. The process optimization method for continuous extrusion molding of wood-plastic composite materials according to claim 5, characterized in that: The temperature classification control strategy refers to dividing the temperature control into three categories according to the thermal degradation characteristics of wood fiber: a safe temperature range of 145-155°C, a suitable temperature range of 155-185°C, and a dangerous temperature range of 185-195°C. The processing efficiency and material properties are balanced by controlling the proportion of temperature classification in each area.

7. The process optimization method for continuous extrusion molding of wood-plastic composite materials according to claim 6, characterized in that: The fiber degradation rate corresponding to the safe temperature range is 0-2%, the fiber degradation rate corresponding to the suitable temperature range is 2-8%, and the fiber degradation rate corresponding to the dangerous temperature range is 8-15%. The proportion of the suitable temperature range is maintained at 60-80%, the proportion of the safe temperature range is 15-30%, and the proportion of the dangerous temperature range is controlled at 5-15%.

8. The process optimization method for continuous extrusion molding of wood-plastic composite materials according to claim 7, characterized in that: The genetic algorithm historical data acquisition method refers to collecting process parameters and product quality data through more than 500 consecutive batches of wood-plastic composite material extrusion experiments, establishing a multidimensional database containing temperature-performance relationships, temperature-energy consumption relationships, and temperature-stability relationships, using 70% of the data as a training set for genetic algorithm parameter optimization, and 30% of the data as a validation set for algorithm performance evaluation.

9. The process optimization method for continuous extrusion molding of wood-plastic composite materials according to claim 8, characterized in that: When the temperature deviation exceeds ±2°C, game optimization adjustment is initiated. The intelligent multi-parameter monitoring platform includes a combination of a thermal imager, a contact temperature sensor, a non-contact infrared thermometer, a pressure sensor, and an online rheometer. The thermal imager uses a long-wave infrared detector with a wavelength of 8 to 14 microns and a temperature resolution of 0.1°C.

10. The process optimization method for continuous extrusion molding of wood-plastic composite materials according to claim 9, characterized in that: The method of iteratively optimizing acquisition parameters based on data feature similarity refers to extending the acquisition interval to once every 10 seconds when the standard deviation of continuously acquired temperature data is less than 1°C, indicating that the temperature is stable. When the temperature data suddenly changes or the similarity coefficient is lower than 0.9, the acquisition frequency is increased to once every 1 second and an early warning mechanism is activated.

Images