A method for controlling temperature of continuous extrusion of high-fluidity PC plastic pellets

Abstract

The present invention relates to a method for continuous extrusion temperature control of high-fluidity PC plastic pellets in the field of PC plastic pellet production temperature control. The method comprises: establishing a nonlinear mathematical model of plastic pellet temperature and multiple process parameters such as screw speed and back pressure through a response surface methodology; solving the optimal parameter combination of the model using a particle swarm optimization algorithm to obtain an optimal temperature control strategy under different working conditions, thereby realizing adaptive parameter optimization control of the extrusion production process; establishing a correlation model between plastic pellet quality, temperature, and rheological properties; classifying and identifying quality defects of extruded products using a support vector machine algorithm; inferring the causes of temperature and rheological property abnormalities based on the defect type, and forming a corresponding knowledge base of quality defects and process parameter deviations for guiding optimization and improvement of the temperature control strategy.

Description

Technical Field

[0001] The present invention relates to the field of information technology, in particular to the field of temperature control in the production of PC plastic pellets, and specifically to a method for controlling the continuous extrusion temperature of high-fluidity PC plastic pellets. Background Art

[0002] The continuous extrusion production of high-flow PC plastic pellets faces the technical challenge of insufficient temperature control precision. Due to the exceptionally excellent flow properties of these pellets, they are highly sensitive to fluctuations in processing temperature. Even slight temperature deviations during the extrusion process can significantly alter the melt flowability, impacting the stability of extrusion molding and the consistency of product quality. Excessively high temperatures can lead to thermal degradation of the pellets, causing molecular chain breakage and decreased fluidity. Excessively low temperatures can lead to incomplete melting and insufficient fluidity, both of which can result in poor extrusion. The existing single constant temperature control model is insufficient to precisely control pellet temperature during production. A more refined zoned temperature control technology is urgently needed. This technology can implement differentiated temperature control strategies based on the melting state of the pellets in different zones of the extruder screw to ensure dynamic equilibrium in pellet temperature. Furthermore, a precise mathematical model linking pellet temperature with process parameters such as screw speed and back pressure must be established. This approach, through real-time closed-loop feedback control, can enhance the intelligence of continuous extrusion production. Summary of the Invention

[0003] The present invention provides a method for controlling the temperature of continuous extrusion of high-fluidity PC plastic pellets, which mainly includes:

[0004] To address the issue of insufficient temperature control accuracy during the continuous extrusion production of high-flow PC plastic pellets, a zoned temperature control technology is used to obtain the melt state parameters of the plastic pellets in different zones of the extruder screw. The target control temperature of each zone is determined through a fuzzy control algorithm, achieving refined temperature control of the plastic pellets.

[0005] An online near-infrared spectrometer is used to monitor the temperature and melt fluidity changes of plastic pellets during the extrusion process in real time. Based on the preset temperature and fluidity thresholds, it is determined whether the melting state of the plastic pellets deviates from the optimal process window. If so, the corresponding temperature adjustment measures are triggered to dynamically compensate for temperature fluctuations, ensure uniform melting of the plastic pellets, and improve the consistency of product quality.

[0006] Through the response surface methodology, a nonlinear mathematical model of multiple process parameters such as plastic pellet temperature, screw speed, and back pressure was established. The particle swarm optimization algorithm was used to solve the optimal parameter combination of the model, obtain the best temperature control strategy under different working conditions, and realize the adaptive optimization control of parameters in the extrusion production process.

[0007] Multiple independent temperature control devices are installed in the feed, compression, and metering areas of the extruder. A modular heating and cooling system is used. Differentiated temperature control modes are implemented according to the melting state of the plastic pellets in each area. Low-temperature preheating is used in the feed area to prevent premature melting of the plastic pellets. Gradual heating is used in the compression area to ensure full melting of the plastic pellets. Precise constant temperature is used in the metering area to prevent temperature fluctuations of the plastic pellets.

[0008] Multiple temperature sensors are installed on the extruder screw. The Kalman filter algorithm is used to denoise and fuse the temperature signals to obtain real-time temperature distribution maps of different areas of the plastic pellets. Based on the color change trend of the image, it is judged whether the melting state of the plastic pellets is uniform. If the temperature distribution is uneven, the zone compensation heating is activated to ensure that the plastic pellets are heated uniformly.

[0009] A correlation model between plastic pellet quality, temperature, and rheological properties was established. A support vector machine algorithm was used to classify and identify quality defects in extruded products. The causes of temperature and rheological property abnormalities were inferred from the defect types, forming a corresponding knowledge base between quality defects and process parameter deviations to guide the optimization and improvement of temperature control strategies.

[0010] An adaptive neural fuzzy inference system is used to input multi-source process parameters such as plastic pellet temperature, melt fluidity, screw speed, back pressure, etc. into the fuzzy controller. Through reasoning and decision-making based on the fuzzy rule base, optimized temperature and speed control instructions are output to achieve multi-parameter collaborative optimization control of the extrusion production process, thereby improving the accuracy and response speed of plastic pellet temperature control.

[0011] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0012] The present invention addresses the problem of insufficient temperature control accuracy during the extrusion process. The present invention adopts zoned temperature control technology, combined with online near-infrared spectroscopy detection and multi-parameter nonlinear modeling, to achieve dynamic and precise control of the temperature of plastic pellets. Specifically, the present invention installs independent temperature control devices in different areas of the extruder, determines the target temperature through a fuzzy control algorithm; uses near-infrared spectroscopy to monitor the melting state of plastic pellets in real time and trigger temperature compensation; uses response surface methodology and particle swarm optimization algorithm to solve the optimal temperature control strategy; and combines support vector machines and neural fuzzy inference systems to achieve multi-parameter collaborative optimization control. The present invention significantly improves the temperature control accuracy and response speed of plastic pellets, ensures melting uniformity, and effectively improves the consistency of product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 The present invention is a flow chart of a method for controlling the temperature of continuous extrusion of high-fluidity PC plastic pellets. DETAILED DESCRIPTION

[0014] To further understand the content of the present invention, the present invention is described in detail with reference to the accompanying drawings and examples. The present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are intended only to illustrate the relevant invention and are not intended to limit the invention. It should also be noted that, for ease of description, only portions relevant to the invention are shown in the accompanying drawings.

[0015] like Figure 1 In this embodiment, a method for controlling the temperature of continuous extrusion of high-fluidity PC plastic pellets may specifically include:

[0016] S101. Based on the problem of insufficient temperature control accuracy of high-flow PC plastic pellets during the continuous extrusion production process, the zone temperature control technology is adopted to obtain the melting state parameters of the plastic pellets in different areas of the extruder screw. The target control temperature of each area is determined through the fuzzy control algorithm to achieve fine regulation of the temperature of the plastic pellets.

[0017] Based on the material properties of high-flow PC plastic pellets, an appropriate extrusion temperature range and temperature gradient distribution are determined as target temperatures for zoned temperature control. Using multiple temperature sensors, temperature collection points are placed in different zones of the extruder screw to acquire real-time temperature parameters of the plastic pellets in each zone. A mapping model between temperature and melt parameters is established based on the melting characteristics of the plastic pellets at different temperatures. A fuzzy control algorithm comprehensively analyzes the temperature and melt parameters of each zone and dynamically adjusts the target control temperature for each zone. If the temperature of a zone deviates from the target temperature range, fuzzy rule reasoning is used to determine the heating or cooling control strategy for that zone based on the degree and duration of the deviation. Based on the determined zoned temperature control strategy, the heating and cooling devices in each zone are controlled to achieve precise temperature adjustment for different zones of the extruder screw. By continuously monitoring the changes in the temperature and melt parameters of the plastic pellets in each zone, closed-loop feedback control is used to continuously optimize and adjust the zoned temperature control process to ensure that the plastic pellet temperature remains within the target range, achieving stable and controllable continuous extrusion production.

[0018] S102. Use an online near-infrared spectrometer to detect the temperature and melt fluidity changes of the plastic pellets during the extrusion process in real time. Based on the preset temperature and fluidity thresholds, determine whether the melting state of the plastic pellets deviates from the optimal process window. If deviated, trigger the corresponding temperature adjustment measures to dynamically compensate for temperature fluctuations, ensure uniform melting of the plastic pellets, and improve the consistency of product quality.

[0019] Near-infrared spectral data of the plastic pellets during the extrusion process is acquired and transmitted to the data processing module. The data processing module preprocesses the near-infrared spectral data, including denoising and normalization, to obtain preprocessed spectral data. Based on the preprocessed spectral data, a partial least squares regression algorithm is used to establish a quantitative relationship model between the temperature and melt fluidity of the plastic pellets and the spectral data. The real-time near-infrared spectral data of the plastic pellets is input into the established quantitative relationship model to predict the real-time temperature and melt fluidity values ​​of the plastic pellets. A determination is made as to whether the predicted temperature and melt fluidity values ​​of the plastic pellets exceed a preset threshold range. If so, a temperature adjustment control signal is triggered. Based on the received control signal, the temperature adjustment module uses a PID control algorithm to dynamically adjust the extruder's heating temperature to compensate for temperature fluctuations of the plastic pellets. These steps are executed continuously in a loop, monitoring changes in the temperature and melt fluidity of the plastic pellets in real time and dynamically adjusting the extrusion temperature to ensure uniform melting of the plastic pellets and improve product quality consistency.

[0020] Specifically, the application of near-infrared spectroscopy during the extrusion process of plastic pellets enables real-time, non-destructive monitoring of temperature and melt flow. To acquire near-infrared spectral data, a fiber optic probe can be directly inserted into the extruder barrel, or an online spectrometer can be installed at the extruder outlet. Spectral data preprocessing is a key step in improving model accuracy. Common preprocessing methods include moving average filtering for denoising and multivariate scattering correction for normalization. Taking PC plastic pellets as an example, their near-infrared spectra exhibit multiple characteristic absorption peaks in the 1100-2500nm range, which are related to their molecular structure and physical state. By analyzing the position and intensity changes of these characteristic peaks, a model can be established that relates the spectrum to temperature and melt flow. The partial least squares regression algorithm can effectively address multicollinearity issues in spectral data and extract the most relevant latent variables. During model development, PC plastic pellet samples can be selected at different temperatures (e.g., 260°C, 280°C, and 300°C) and shear rates, and their near-infrared spectra, actual temperature, and melt flow index can be measured simultaneously.

[0021] 70% of the data was used for modeling and 30% for validation. Cross-validation was used to optimize model parameters, ultimately resulting in a quantitative relationship model with prediction accuracies of ±2°C and ±5%. During real-time prediction, spectral data was collected every 1 second and fed into the model, generating continuous curves of the plastic pellet temperature and melt flowability. Appropriate threshold ranges were set, such as 280±5°C for temperature and 20±2g / 10min for melt flow index. Alarms and control signals were triggered if the predicted values ​​exceeded these limits. This method better reflects the actual state of the plastic pellets than traditional thermocouple temperature measurement, avoiding quality issues caused by localized overheating or underheating. Temperature regulation utilizes a PID control algorithm, which flexibly adjusts heating power based on deviation size and trend. For example, if the temperature is detected to be continuously rising and approaching the upper limit, the controller will proactively reduce heating power to prevent overshoot. Compared to simple on-off control, PID control provides smoother and more precise temperature regulation, effectively suppressing temperature fluctuations. This closed-loop control system based on near-infrared spectroscopy can control temperature fluctuations of PC plastic pellets within ±3°C and melt flowability fluctuations within ±8%. This significantly outperforms traditional control methods and can effectively improve the dimensional stability and mechanical property consistency of finished products. Furthermore, real-time monitoring can promptly detect abnormalities, such as impurities in raw materials or equipment failures, improving production efficiency and product quality.

[0022] S103. Through the response surface methodology, a nonlinear mathematical model of multiple process parameters such as plastic pellet temperature and screw speed and back pressure is established. The particle swarm optimization algorithm is used to solve the optimal parameter combination of the model, obtain the best temperature control strategy under different working conditions, and realize the adaptive optimization control of parameters in the extrusion production process.

[0023] A nonlinear mathematical model was established using the response surface methodology based on multiple process parameters, including pellet temperature, screw speed, and back pressure. This model was then solved using a particle swarm optimization algorithm. Through an iterative optimization search, the optimal parameter combination was obtained, and the process parameter values ​​that achieved optimal temperature control under different operating conditions were determined. Based on the optimal parameter combination obtained by the particle swarm optimization algorithm, the current operating conditions were determined to be within a predefined operating condition category. The optimal values ​​of each process parameter from the corresponding parameter combination were then determined as the control targets for the extrusion process. During the extrusion process, the current values ​​of parameters such as pellet temperature, screw speed, and back pressure were collected in real time and compared with the target values ​​in the optimal parameter combination. The deviations were calculated and corresponding adjustment instructions were generated. Based on the generated adjustment instructions, the extruder's actuators, such as screw speed and back pressure, were controlled to dynamically adjust the process parameters to approximate the target values ​​in the optimal parameter combination, achieving precise control of pellet temperature. During the adjustment process, the temperature of the plastic pellets is continuously monitored to determine whether it has reached the target temperature range of the optimal temperature control strategy. If not, the system returns to the previous step to continue parameter adjustment. When the temperature of the plastic pellets stabilizes within the target range of the optimal temperature control strategy, the current process parameter settings are maintained, completing the adaptive optimization control of the extrusion production process parameters and continuously producing plastic pellet products with qualified temperatures.

[0024] Specifically, the response surface methodology (RSM) is a commonly used experimental design and data analysis method that can be used to model the nonlinear relationship between pellet temperature and process parameters. For example, a central composite design experiment can be conducted using screw speed, back pressure, and barrel temperature as independent variables, with pellet temperature as the dependent variable. This data can be used to obtain a series of experimental data points. A quadratic polynomial regression equation can be fitted to these data points to obtain a nonlinear function describing the relationship between pellet temperature and each process parameter. The particle swarm optimization algorithm is an intelligent optimization method that can be used to solve the aforementioned nonlinear model using the optimal parameter combination. The algorithm simulates the foraging behavior of a flock of birds to search for the optimal solution in parameter space. For example, each particle can be represented as a set of process parameter values, and the optimal solution can be gradually approached by iteratively updating the particle position and velocity. During the search process, the objective function can be set to minimize the pellet temperature to the target value, while also considering other constraints such as production efficiency. Based on the optimization results, several typical operating conditions can be pre-defined, such as high speed and low pressure or low speed and high pressure, and the optimal parameter combination can be determined for each condition. In actual production, by determining which preset type the current operating condition belongs to, the optimal parameter combination can be quickly selected as the control target. This approach can improve system response speed and adapt to different production needs. During the extrusion process, the current values ​​of various process parameters need to be collected in real time. For example, a thermocouple can be used to measure the temperature of the plastic pellets, an encoder can be used to obtain the screw speed, and a pressure sensor can be used to measure the back pressure. These measured values ​​are compared with the target values ​​of the optimal parameter combination, the deviation is calculated, and corresponding adjustment instructions are generated based on the size and direction of the deviation.

[0025] Based on the generated adjustment instructions, the motor speed is adjusted by the frequency converter to control the screw speed, the back pressure is adjusted by adjusting the back pressure valve opening, and the heater power is controlled by the PID controller to control the barrel temperature. These adjustments aim to bring each process parameter closer to the target value in the optimal combination, thereby achieving precise control of the pellet temperature. During the adjustment process, the pellet temperature must be continuously monitored. A temperature tolerance range can be set, such as ±2°C above the target temperature. If the actual temperature exceeds this range, further parameter adjustments are required. This closed-loop control method effectively addresses external interference and system fluctuations, ensuring effective temperature control. When the pellet temperature stabilizes within the target range, optimal temperature control has been achieved. The system then maintains the current process parameter settings and continues stable production. This adaptive optimization control method automatically adjusts process parameters based on varying production conditions and raw material characteristics, improving product quality stability and production efficiency.

[0026] S104. Install multiple sets of independent temperature control devices in the feeding area, compression area and metering area of ​​the extruder, and adopt a modular heating and cooling system. Implement differentiated temperature control modes according to the melting state of the plastic pellets in each area. Use low-temperature preheating for the feeding area to avoid premature melting of the plastic pellets. Use gradual heating for the compression area to ensure sufficient melting of the plastic pellets. Use precise constant temperature for the metering area to prevent temperature fluctuations of the plastic pellets.

[0027] Real-time temperature data from the extruder's feed, compression, and metering zones is acquired. Based on preset temperature thresholds, the system determines whether the temperatures in each zone are within a reasonable range. If the feed zone temperature exceeds the preset low-temperature preheating threshold, the cooling system in the feed zone is activated to lower the temperature until it reaches the preset low-temperature preheating threshold, preventing premature melting of the plastic pellets. If the compression zone temperature falls below the preset gradual temperature increase threshold, the heating system in the compression zone is activated to raise the temperature, using a gradual temperature increase mode to ensure sufficient melting of the plastic pellets within the compression zone. If the metering zone temperature exceeds the preset precise constant temperature threshold, the heating or cooling system in the metering zone is activated, depending on the direction of temperature deviation, to achieve precise constant temperature control and prevent temperature fluctuations in the metering zone. A machine learning algorithm is used to establish a mapping model between temperature and the melting state of the plastic pellets. Based on real-time temperature data collected from each zone, the melting state of the plastic pellets is predicted, and the temperature control strategy for each zone is dynamically adjusted. During the production process, the system continuously records temperature data from each zone and the corresponding melt state of the plastic pellets. Using an incremental learning algorithm, the temperature-melt state mapping model is updated and optimized in real time to improve temperature control accuracy. Based on this optimized temperature-melt state mapping model, the temperature threshold ranges for the feed, compression, and metering zones are adaptively adjusted, enabling intelligent temperature management of the extruder, stabilizing the melt state of the plastic pellets and improving extrusion quality.

[0028] Specifically, extruder temperature control is a critical step in the plastics processing process, directly impacting product quality. Temperature management in the feed, compression, and metering zones is particularly important. For polyethylene (PE) extrusion, for example, the feed zone temperature is typically controlled within a range of 120-140°C. Excessively high temperatures can cause premature melting of the PE, making feeding difficult. Therefore, a low-temperature preheating threshold is set at 130°C. When the temperature exceeds this threshold, the cooling system is activated to reduce the temperature. The compression zone temperature is crucial for PE melting and is generally controlled between 160-180°C. A gradual temperature increase mode ensures uniform PE melting and avoids local overheating. For example, the gradual temperature increase threshold can be set at 165°C. When the temperature falls below this threshold, the heating power is gradually increased to slowly raise the temperature. This approach can reduce thermal degradation of PE and improve product quality. The metering zone temperature directly affects the uniformity and stability of the discharge and is typically controlled between 180-200°C.

[0029] The precise constant temperature threshold range is set to 190±5℃. When the temperature deviates from this range, the heating or cooling system is started according to the direction of deviation. For example, when the temperature reaches 196℃, the cooling system is started; when the temperature drops to 184℃, the heating system is started. This precise control can ensure the fluidity and uniformity of the PE melt. The application of machine learning algorithms in temperature control can greatly improve control accuracy. Taking support vector machines (SVM) as an example, a mapping model between temperature and PE melting state can be established. The input variables include the temperature of each region, the screw speed, etc., and the output variable is the melt index (MI) of PE. By training the model with a large amount of historical data, real-time prediction of the PE melting state can be achieved. Incremental learning algorithms such as online stochastic gradient descent (OnlineSGD) can be used for real-time updates of the model.

[0030] During each production run, the system records temperature data and the corresponding PE melt state, using this new data to fine-tune the model. For example, if the predicted MI value deviates from the actual measured value, the system automatically adjusts the model parameters to improve prediction accuracy. Based on the optimized model, the system can adaptively adjust the temperature threshold range. For example, if the model predicts that the PE MI value is too high at the current temperature setting, the system automatically lowers the temperature threshold for each zone. This intelligent temperature management method dynamically adjusts process parameters based on factors such as raw material batches and ambient temperature to maintain stable product quality. Implementing this temperature control strategy significantly improves the stability of the PE extrusion process and product quality. For example, during a 24-hour production run, using traditional fixed temperature control methods, the product MI value fluctuates within ±10%. With the intelligent temperature management system, the MI value fluctuation is reduced to ±3%, significantly improving product consistency. Furthermore, due to more precise temperature control, thermal degradation of the raw material is reduced, and the mechanical properties of the product are improved, with tensile strength increasing by 5-8%. Furthermore, intelligent temperature management optimizes energy utilization, reducing energy consumption by approximately 10-15% while ensuring product quality.

[0031] S105. Multiple temperature sensors are set on the extruder screw. The Kalman filter algorithm is used to denoise and fuse the temperature signals to obtain real-time temperature distribution maps of different areas of the plastic pellets. Based on the color change trend of the image, it is judged whether the melting state of the plastic pellets is uniform. If uneven temperature distribution occurs, zone compensation heating is started to ensure that the plastic pellets are heated uniformly.

[0032] Multiple high-precision temperature sensors are installed in different areas of the extruder screw to collect real-time temperature data from the plastic pellets at different locations on the screw. The collected temperature data is denoised using a Kalman filter algorithm to eliminate high-frequency noise interference in the temperature signal and improve temperature measurement accuracy. The temperature data collected by multiple temperature sensors is fused to comprehensively analyze the temperature distribution of the plastic pellets in different areas of the screw, generating a real-time temperature distribution map. Image analysis is performed on the generated temperature distribution map to extract color features. Based on color trends, the uniformity of the melt state of the plastic pellets is determined. If the color analysis of the temperature distribution map indicates uneven melt state of the plastic pellets, the zoned compensatory heating control mechanism is triggered to focus heating on the lower-temperature areas. During the zoned compensatory heating process, the heating power and heating time are dynamically adjusted. Through closed-loop control, uniform heating of the plastic pellets is achieved, eliminating temperature inconsistencies. The temperature distribution map is continuously monitored. When the color distribution becomes consistent and the temperature distribution is uniform, the plastic pellets are determined to have reached the optimal melt state, completing the extruder screw temperature optimization control.

[0033] Specifically, installing high-precision temperature sensors in different areas of the extruder screw is essential for achieving precise temperature control. For example, thermocouples or infrared temperature sensors can be installed in the feed, compression, and metering zones. These sensors can quickly respond to temperature changes with an accuracy of ±0.1°C. Real-time temperature data may be affected by factors such as environmental noise and electromagnetic interference, making it essential to employ a Kalman filter algorithm for denoising. Through two steps, prediction and correction, the Kalman filter effectively eliminates high-frequency noise and improves temperature measurement accuracy. Temperature data fusion is key to comprehensively analyzing the melting state of plastic pellets. Weighted averaging or Bayesian fusion algorithms can be used to combine data from multiple sensors to generate a more reliable temperature distribution map. For example, if 10 temperature sensors are installed on the screw, data fusion can produce a continuous temperature curve that intuitively reflects the entire temperature evolution of the plastic pellets from feed to discharge. Image analysis of the temperature distribution map can employ computer vision techniques such as color segmentation and edge detection. For example, temperature values ​​can be mapped to different colors, such as blue for low temperatures and red for high temperatures. By analyzing color gradients, it's possible to determine whether the plastic pellets are melting uniformly. If abrupt color changes or noticeable color blocks are observed in certain areas, this indicates a potential problem with the melting state in those areas. Zoned compensatory heating control is an effective means of addressing uneven melting. For example, if the temperature in the compression zone is low, the heating power can be increased in that area. This compensatory heating approach employs a PID control algorithm to dynamically adjust the heating power based on the deviation between the target and actual temperatures. Fuzzy control theory can also be incorporated to develop a more flexible heating strategy based on the magnitude and trend of temperature deviations. Closed-loop control is key to achieving uniform heating of plastic pellets. By continuously monitoring temperature distribution changes and adjusting heating parameters in a timely manner, temperature inconsistencies can be effectively eliminated. For example, a temperature uniformity metric can be set such that optimal melting is achieved when the temperature difference between zones is less than ±2°C. This closed-loop control approach not only improves product quality but also optimizes energy efficiency. The ultimate goal of temperature optimization control is to achieve optimal melting of the plastic pellets. By continuously monitoring the temperature distribution, optimal melting can be determined when the color distribution becomes consistent and the temperature distribution is uniform. This approach not only ensures product quality but also improves production efficiency. For example, in actual production, the optimal melting temperature range can be pre-set based on the characteristics of different plastic materials. When the temperature distribution is detected to fall within this range and remains stable, the next process is automatically entered, realizing intelligent production.

[0034] S106. Establish a correlation model between the quality of plastic pellets and temperature and rheological properties. Use the support vector machine algorithm to classify and identify the quality defects of extruded products. Infer the causes of abnormal temperature and rheological properties based on the defect type, and form a corresponding knowledge base of quality defects and process parameter deviations to guide the optimization and improvement of temperature control strategies.

[0035] Temperature and rheological property data for plastic pellets, as well as quality defect data for extruded products, are collected to establish a correlation model between pellet quality, temperature, and rheological properties. A support vector machine algorithm is used to classify and identify quality defects in extruded products, generating sample data for different defect types. Based on the quality defect type and the correlation model between pellet quality, temperature, and rheological properties, the cause of the temperature and rheological abnormalities leading to that defect type is determined. The quality defect type is then associated with the corresponding temperature and rheological abnormality causes, forming a knowledge base that correlates quality defects with process parameter deviations. When a new quality defect in an extruded product occurs, the support vector machine algorithm is used to identify the defect type and the corresponding temperature and rheological abnormality causes are searched within the knowledge base. Based on the identified temperature and rheological abnormality causes, the temperature control strategy and process parameters that require optimization are determined. The optimized temperature control strategy is then applied to the extrusion process. By monitoring pellet temperature and rheological property data in real time, the temperature control parameters are dynamically adjusted to ensure the quality stability of the extruded product.

[0036] Specifically, a model correlating plastic pellet quality with temperature and rheological properties is fundamental to quality control of extruded products. By collecting rheological property data of plastic pellets at different temperatures, such as melt flow rate and shear viscosity, a temperature-rheological property curve can be constructed. For example, the melt flow rate of a polyethylene pellet is 2 g / 10 min at 180°C, but increases to 5 g / 10 min at 220°C, indicating that increasing temperature reduces pellet viscosity and improves fluidity. Support vector machine algorithms play a key role in classifying quality defects in extruded products. Support vector machine models are trained by extracting feature vectors of defect images, such as texture, shape, and color. For example, for surface bubble defects, features such as bubble roundness and size distribution can be extracted; for weld line defects, features such as line continuity and width can be extracted. This allows for accurate distinction between different types of defects. A knowledge base that maps quality defects to process parameter deviations is key to problem solving. For example, surface bubble defects may be caused by excessively high pellet temperature or insufficient drying. If a batch of products is found to have a large number of bubbles, the knowledge base can be consulted to quickly identify the possible causes, such as melt temperature exceeding 230°C or moisture content of the plastic pellets exceeding 0.1%. This allows for prompt action such as lowering the temperature or increasing drying.

[0037] Optimizing temperature control strategies is an effective means of improving product quality. For example, for easily degradable plastics, a step-by-step temperature ramp strategy can be employed: maintaining a low temperature, such as 160°C, in the feeding section and gradually increasing it to 200°C in the metering section. This ensures sufficient plastic melting while preventing excessive degradation. Furthermore, screw speed and back pressure can be adjusted according to the characteristics of the plastic to optimize plasticization. Real-time monitoring and dynamic adjustment are key to ensuring consistent extruded product quality. Online viscometers and temperature sensors enable real-time monitoring of the melt state. If a sudden drop in viscosity is detected, possibly due to excessive temperature or shear degradation, the system will automatically reduce the screw speed or lower the temperature setpoint. This closed-loop control approach rapidly responds to process fluctuations and maintains consistent product quality. By establishing a correlation model between pellet quality, processing parameters, and product quality, combined with machine learning algorithms and expert knowledge bases, intelligent control of the extrusion process can be achieved. This not only improves product quality but also reduces energy consumption and raw material waste, providing technical support for the sustainable development of the plastics processing industry.

[0038] S107. Adopting an adaptive neural fuzzy inference system, the multi-source process parameters such as the temperature of the plastic pellets, melt fluidity, screw speed, back pressure, etc. are input into the fuzzy controller. Through the reasoning and decision-making of the fuzzy rule base, the optimized temperature and speed control instructions are output to realize the multi-parameter coordinated optimization control of the extrusion production process, thereby improving the accuracy and response speed of the temperature control of the plastic pellets.

[0039] Data on the temperature, melt fluidity, screw speed, and back pressure of the plastic pellets during extruder operation are collected. A raw database is constructed using this data. Data preprocessing is performed on the raw database to remove outliers and obtain a valid dataset. The adaptive neural-fuzzy inference system is trained using this valid dataset to determine the rules in the fuzzy rule base and obtain a trained inference engine. Real-time data on the temperature, melt fluidity, screw speed, and back pressure of the plastic pellets during extruder operation is obtained and input into the trained inference engine. The trained inference engine outputs preliminary temperature and speed control instructions. Based on the current extruder status, the rationality of these preliminary instructions is determined. If not, the inference engine parameters are adjusted. By adjusting the inference engine parameters, optimized temperature and speed control instructions are obtained and output to the extruder actuator. Based on the operating status of the extruder actuator, the operating status of the extruder is determined to be consistent with the set values. If so, real-time data collection from the extruder continues.

[0040] Specifically, data collection on the extruder's operating status is fundamental to plastics processing control. Taking polyethylene extrusion as an example, key parameters such as melt temperature, melt flow index (MFI), screw speed, and back pressure can be collected. Temperature can be measured using thermocouples, with a typical value of 180-220°C; MFI can be determined using a capillary rheometer, with a common value of 2-10 g / 10 min; screw speed can be measured using a motor tachometer, typically 30-100 rpm; and back pressure can be measured using a pressure sensor, typically 5-15 MPa. Data preprocessing is crucial to ensuring the quality of model training. The 3σ criterion can be used to eliminate outliers, such as data points with temperatures exceeding 250°C or below 150°C. Furthermore, data normalization can be performed to normalize each parameter to the range of 0-1 to eliminate dimensionality effects. Adaptive neuro-fuzzy inference systems (ANFIS) combine the learning capabilities of neural networks with the reasoning capabilities of fuzzy logic.

[0041] During the training process, the backpropagation algorithm can be used to optimize the membership function parameters. For example, a Gaussian membership function can be used to adjust its center value and width. Simultaneously, the least squares method is used to determine the consequent parameters. This results in a fuzzy rule base that accurately describes the input-output relationship. Real-time control is the core of the extrusion process. If a 2°C increase in melt temperature is detected, the ANFIS might output a command to reduce the screw speed by 5 rpm. However, if the back pressure is already low at this point, a direct speed reduction could result in reduced product quality. In this case, expert experience is needed to adjust the ANFIS output weights, perhaps lowering the heating zone temperature setpoint by 3°C. The execution of control commands involves the coordination of multiple actuators. Temperature regulation is achieved through a PID controller, which may take 0.5-2 minutes to reach a new steady-state. Screw speed regulation can be achieved quickly through a frequency converter, with a response time typically in seconds. Therefore, in actual control, the dynamic characteristics of different actuators must be considered and the control cycle appropriately set. Closed-loop control is key to ensuring extrusion process stability. By comparing deviations between actual operating parameters and setpoints, anomalies can be detected and adjusted promptly. For example, if the actual temperature is found to be continuously higher than the set value by more than 2°C, it may be due to excessive shear heat, and it is necessary to re-evaluate whether the screw speed setting is reasonable. Through this continuous monitoring and adjustment, the extrusion process can always be kept within the optimal process window.

[0042] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A method for controlling the temperature of continuous extrusion of high-fluidity PC plastic pellets, characterized in that: The method comprises: Step S101: To address the issue of insufficient temperature control accuracy during the continuous extrusion production of high-flow PC plastic pellets, a zoned temperature control technology is employed to obtain melt parameters of the plastic pellets in different zones of the extruder screw. A fuzzy control algorithm is then used to determine the target control temperature for each zone, thereby achieving refined temperature control of the plastic pellets. Step S102: Using an online near-infrared spectrometer, the temperature and melt fluidity of the plastic pellets during the extrusion process are monitored in real time. Based on preset temperature and fluidity thresholds, the system determines whether the melt state of the plastic pellets deviates from the optimal process window. If so, appropriate temperature adjustment measures are triggered to dynamically compensate for temperature fluctuations, ensuring uniform melting of the plastic pellets and improving product quality consistency. Step S103: Using the response surface methodology, a nonlinear mathematical model of the process parameters of the plastic pellet temperature, screw speed, and back pressure is established. The particle swarm optimization algorithm is used to solve the optimal parameter combination of the model, thereby obtaining the optimal temperature control strategy under different working conditions, thereby achieving adaptive optimization control of the parameters of the extrusion production process. Step S104: Install multiple independent temperature control devices in the feed zone, compression zone, and metering zone of the extruder. A modular heating and cooling system is used. Differentiated temperature control modes are implemented based on the melting state of the plastic pellets in each zone. Low-temperature preheating is used in the feed zone to prevent premature melting of the plastic pellets. A gradual heating is used in the compression zone to ensure sufficient melting of the plastic pellets. A precise constant temperature is used in the metering zone to prevent temperature fluctuations of the plastic pellets. Step S105: Multiple temperature sensors are installed on the extruder screw. The Kalman filter algorithm is used to perform denoising and fusion processing on the temperature signals to obtain real-time temperature distribution maps of different areas of the plastic pellets. Based on the color change trend of the image, it is determined whether the melting state of the plastic pellets is uniform. If the temperature distribution is uneven, zone compensation heating is activated to ensure that the plastic pellets are heated uniformly. Step S106: Establish a correlation model between the quality of plastic pellets and temperature and rheological properties. Use a support vector machine algorithm to classify and identify quality defects of extruded products. Infer temperature and rheological property anomalies based on defect types, and form a corresponding knowledge base of quality defects and process parameter deviations to guide the optimization and improvement of temperature control strategies. In step S107, an adaptive neural fuzzy inference system is used to input the source process parameters of the plastic pellet temperature, melt fluidity, screw speed, and back pressure into the fuzzy controller. Through the reasoning and decision-making of the fuzzy rule base, the optimized temperature and speed control instructions are output to realize the multi-parameter collaborative optimization control of the extrusion production process and improve the accuracy and response speed of the plastic pellet temperature control.

2. The method according to claim 1, characterized in that The step S101 includes: According to the material properties of high-fluidity PC plastic pellets, the extrusion temperature range and temperature gradient distribution are determined as the target temperature reference for zone temperature control; Using multi-point temperature sensors, temperature collection points are arranged in different areas of the extruder screw to obtain the temperature parameters of the plastic pellets in each area in real time; According to the melting state characteristics of plastic particles at different temperatures, a mapping relationship model between temperature and melting state parameters is established; Through fuzzy control algorithm, the temperature parameters and melting state parameters of each area are comprehensively analyzed to dynamically adjust the target control temperature of each area; If the temperature of a certain area deviates from the target temperature range, the heating or cooling control strategy of the area is determined based on the degree and duration of deviation through fuzzy rule reasoning; According to the determined zone temperature control strategy, the heating and cooling devices in each zone are controlled to finely adjust the temperature of different zones of the extruder screw; Continuously monitor the temperature and melting state parameter changes of plastic pellets in each area, and continuously optimize and adjust the zone temperature control process through closed-loop feedback control to ensure that the temperature of plastic pellets is always within the target range, achieving stable and controllable continuous extrusion production.

3. The method according to claim 1, characterized in that The step S102 includes: Acquire near-infrared spectral data of plastic pellets during the extrusion process and transmit the spectral data to the data processing module; The data processing module preprocesses the near-infrared spectral data, including denoising and normalization, to obtain preprocessed spectral data; Based on the pre-processed spectral data, a partial least squares regression algorithm was used to establish a quantitative relationship model between the temperature of plastic pellets, melt fluidity and spectral data. Input the real-time near-infrared spectral data of plastic pellets into the established quantitative relationship model to predict the real-time temperature and melt fluidity values ​​of the plastic pellets; Determine whether the predicted plastic pellet temperature and melt fluidity values ​​exceed the preset threshold range. If so, trigger a temperature adjustment control signal; The temperature regulation module uses the PID control algorithm according to the control signal received to dynamically adjust the heating temperature of the extruder and compensate for the temperature fluctuation of the plastic pellets; Continuous cycle execution monitors the temperature of plastic pellets and changes in melt fluidity in real time, dynamically adjusts the extrusion temperature, ensures uniform melting of plastic pellets, and improves product quality consistency.

4. The method according to claim 1, wherein The step S103 includes: According to the process parameters of plastic pellet temperature, screw speed and back pressure, a nonlinear mathematical model was established using the response surface methodology to obtain a nonlinear function expression describing the relationship between plastic pellet temperature and each process parameter. The particle swarm optimization algorithm is used to solve the established nonlinear mathematical model. The optimal parameter combination of the model is obtained through iterative optimization search, and the values ​​of each process parameter that achieves the best temperature control effect under different working conditions are determined; According to the optimal parameter combination obtained by the particle swarm optimization algorithm, it is determined which preset working condition type the current working condition belongs to, and the optimal value of each process parameter is obtained from the corresponding parameter combination as the control target of the extrusion production process; During the extrusion production process, the current values ​​of the plastic pellet temperature, screw speed and back pressure parameters are collected in real time, compared with the target values ​​in the optimal parameter combination, the deviation value is calculated, and the corresponding adjustment instructions are generated; According to the generated adjustment instructions, the screw speed and back pressure actuator of the extruder are controlled, and each process parameter is dynamically adjusted to approach the target value in the optimal parameter combination, thereby achieving precise control of the temperature of the plastic pellets; During the adjustment process, the temperature changes of the plastic pellets are continuously monitored to determine whether they have reached the target temperature range of the optimal temperature control strategy. If not, the process returns to repeat the execution and continues to adjust the parameters. When the temperature of the plastic pellets is stable within the target range of the optimal temperature control strategy, the current process parameter settings are maintained, the parameter adaptive optimization control of the extrusion production process is completed, and plastic pellet products with qualified temperatures are continuously produced.

5. The method according to claim 1, wherein The step S104 includes: Obtain real-time temperature data from the extruder's feed, compression, and metering zones, and determine whether the temperature in each zone is within a reasonable range based on preset temperature thresholds. If the temperature of the feeding zone is higher than the preset low-temperature preheating threshold, the cooling system of the feeding zone is controlled to start and reduce the temperature of the feeding zone until the temperature reaches the preset low-temperature preheating threshold range to avoid premature melting of the plastic pellets; If the temperature of the compression zone is lower than the preset gradual temperature rise threshold, the heating system of the compression zone is controlled to start, increase the temperature of the compression zone, and adopt a gradual temperature rise mode to ensure that the plastic particles are fully melted in the compression zone; If the temperature in the metering area exceeds the preset precise constant temperature threshold range, the heating system or cooling system in the metering area will be controlled to start according to the temperature deviation direction to achieve precise constant temperature control and prevent the temperature fluctuation of the plastic pellets in the metering area; A mapping model between temperature and the melting state of plastic pellets is established through machine learning algorithms. Based on the real-time temperature data collected in each area, the melting state of plastic pellets is predicted and the temperature control strategy of each area is dynamically adjusted. During the production process, the temperature data of each area and the corresponding melting state of the plastic pellets are continuously recorded, and the temperature-melting state mapping model is updated and optimized in real time using an incremental learning algorithm to improve the accuracy of temperature control; Based on the optimized temperature-melting state mapping model, the temperature threshold ranges of the feeding zone, compression zone, and metering zone are adaptively adjusted to achieve intelligent temperature management of the extruder, stabilize the melting state of the plastic pellets, and improve extrusion quality.

6. The method according to claim 1, characterized in that The step S105 includes: Install multiple high-precision temperature sensors in different areas of the extruder screw to collect real-time temperature data of plastic pellets at different positions of the screw; The Kalman filter algorithm is used to perform denoising on the collected temperature data to eliminate high-frequency noise interference in the temperature signal and improve the accuracy of temperature measurement; The temperature data collected by multiple temperature sensors are integrated and processed to comprehensively analyze the temperature distribution of plastic particles in different areas of the screw and generate a real-time temperature distribution map; Perform image analysis on the generated temperature distribution map, extract the color features of the image, and judge whether the melting state of the plastic pellets is uniform based on the color change trend; If the color analysis of the temperature distribution graph reveals that the melting state of the plastic pellets is uneven, the zone compensation heating control mechanism is triggered to focus on heating the areas where the temperature is lower than the preset value; During the zone compensation heating process, the heating power and heating time are dynamically adjusted. Through closed-loop control, the heating uniformity of the plastic pellets is achieved and the inconsistency of temperature distribution is eliminated. Continuously monitor the changes in the temperature distribution diagram. When the color distribution tends to be consistent and the temperature distribution is uniform, it is determined that the plastic particles have reached the optimal melting state and the temperature optimization control of the extruder screw is completed.

7. The method according to claim 1, characterized in that The step S106 includes: Obtain temperature and rheological property data of plastic pellets, as well as quality defect data of extruded products, and establish a correlation model between plastic pellet quality, temperature, and rheological properties; The support vector machine algorithm is used to classify and identify the quality defects of extruded products, and sample data of different defect types are obtained; Based on the quality defect type, combined with the correlation model between plastic pellet quality, temperature and rheological properties, determine the temperature and rheological property abnormalities that cause the defect type; Correlate quality defect types with corresponding temperature and rheological property anomalies to form a corresponding knowledge base of quality defects and process parameter deviations; When a new quality defect of an extruded product occurs, the support vector machine algorithm is used to identify the defect type, and the temperature and rheological property anomalies corresponding to the defect type are searched in the corresponding knowledge base of quality defects and process parameter deviations; Based on the temperature and rheological property anomalies found, determine the temperature control strategy and process parameters that need to be optimized and adjusted; The optimized temperature control strategy is applied to the plastic pellet extrusion production process. By real-time monitoring of the temperature and rheological properties of the plastic pellets, the temperature control parameters are dynamically adjusted to ensure the quality stability of the extruded products.

8. The method according to claim 1, characterized in that The step S107 includes: Collect the temperature, melt fluidity, screw speed and back pressure data of the plastic pellets when the extruder is running, and build an original database based on these data; Based on the data in the original database, data preprocessing is performed to remove outliers and obtain a valid data set; By training the adaptive neuro-fuzzy inference system with valid data sets, the rules in the fuzzy rule base are determined and a trained inference engine is obtained; Obtain the real-time data of plastic pellet temperature, melt fluidity, screw speed, and back pressure from the extruder, and input the real-time data into the trained inference engine; According to the trained inference engine, the initial temperature and speed control instructions are output. Combined with the current extruder status, the rationality of the initial instructions is judged. If it is unreasonable, the inference engine parameters are adjusted. By adjusting the inference engine parameters, the optimized temperature and speed control instructions are obtained and the optimized instructions are output to the extruder actuator; According to the operating status of the extruder actuator, it is judged whether the operating status of the extruder is consistent with the set value. If it is consistent, the real-time data of the extruder will continue to be collected.

Images