A method for controlling the extrusion of waste resin based on electrical parameters

Abstract

The application provides a kind of waste resin extrusion forming control method based on electric parameter.Start waste resin mixing extrusion system, preheat heating module;Predictive control is carried out to motor control module through database;Motor output shaft under motor control module drives screw torque movement;When base material is plasticized, after being rotated and extruded by screw and being output, the pressure and flow rate of waste resin mixing extrusion system plasticizing extrusion are detected, whether the parameters of pressure and flow rate meet target value is judged, if yes, the plasticized plastic is continuously output, if not, the detected parameters of pressure and flow rate are fed back to motor control module, and motor control module is adjusted to control pressure and flow rate.The application is used to solve the problems of redundant control of waste resin mixing extrusion system in prior art and inability to carry out process quality prediction.

Description

Technical Field

[0001] This invention belongs to the field of waste resin extrusion prediction and molding control technology based on electrical parameters. Background Technology

[0002] Extrusion equipment consists of several production steps, including feeding, transmission, heating and plasticizing, and extrusion output. During the production process, temperature sensors, confluence core pressure sensors, and speed sensors detect and provide feedback, enabling closed-loop feedback automated control. However, because the sensors operate in a high-temperature and high-pressure environment, they are easily damaged, affecting the continuity of production. Higher material pressure during the production process is more beneficial for the molding process of waste resin, but higher pressure will reduce the service life of the screw.

[0003] When the pressure sensor provides feedback on the pressure value, if the screw torque is continuously adjusted to ensure the pressure output, the screw load cycle will fluctuate periodically. This adds a pulsating torque load to the pressure closed-loop control process, which has a significant impact on the screw's lifespan and directly reduces its lifespan. The screw is the core component of the waste resin extruder and undertakes more than 80% of the extruder's tasks.

[0004] Existing technologies utilize temperature and pressure sensors to detect the state of the production process, and then use set thresholds for feedback control to achieve process control. This is a traditional automation control strategy. However, when sensors fail and threshold adjustments are made, they are done through trial and error without using data analysis techniques for model analysis and prediction. Therefore, it is impossible to model and predict the historical state of the equipment, and it can only achieve real-time state detection.

[0005] Existing waste resin extrusion equipment includes a die, a punch located directly above the die and movable up and down, a drive mechanism connected to the punch for driving its up and down movement, a screw extruder located on one side of the die for extruding thermoplastic material into the die, and a drive module for driving the screw extruder to move laterally closer to or away from the die; the outlet end of the screw extruder is close to the die. This invention effectively reduces the manufacturing process and cost of the equipment by changing the operating mode of traditional waste resin extrusion equipment. Meanwhile, existing methods directly mold and composite thermoplastic materials with fabrics, which, compared to traditional composite methods, have the advantages of strong bonding, simple processes, and high bonding efficiency. However, they lack the function of intelligent detection and identification of data status. Summary of the Invention

[0006] This invention provides a waste resin extrusion molding control method based on electrical parameters, which solves the problems of redundant control and inability to predict process quality in existing waste resin mixing and extrusion systems.

[0007] This invention is achieved through the following technical solution: A method for controlling waste resin extrusion molding based on electrical parameters, the method comprising: Start the waste resin mixing and extrusion system and preheat the heating module; Predictive control of the motor control module is achieved through a database. The motor output shaft under the motor control module drives the screw torque to perform motion. When the base material is plasticized, it is extruded and output by the screw through the extrusion port. The pressure and flow rate of the plasticizing extrusion of the waste resin compounding extrusion system are detected to determine whether the pressure and flow rate parameters meet the target values. If so, the plasticized plastic is continuously output. If not, the detected pressure and flow rate parameters are fed back to the motor control module, and the motor control module is adjusted to control the pressure and flow rate.

[0008] A data-predictive method for controlling resin compounding and extrusion, wherein the predictive control specifically involves... The electrical parameters of the motors during the production process are collected, filtered, and then dynamically adjusted to achieve dynamic parameter regulation.

[0009] A data-predictive-based method for controlling resin compounding and extrusion, wherein the database is specifically established as follows: Based on the adjustment of dynamic parameters, after sampling and screening of abnormal parameter values, a time series-based process parameter database is established, and a production process parameter model is built.

[0010] A production process parameter model is established based on a time-series database of process parameters. Based on the established production process parameter model, the machine is trained to autonomously judge the production status and adjust the parameter status during the production process, thereby realizing production process parameter control and fault early warning and diagnosis.

[0011] A data-predictive-based resin compounding and extrusion control method, wherein the step of collecting parameters during the production process, filtering out abnormal parameter values, and dynamically adjusting electrical parameters specifically involves: using sensors to collect the output electrical parameters of the motor; using the KNN clustering algorithm to filter and remove abnormal parameter values ​​from the motor's output electrical parameters; using the remaining motor output electrical parameters as a database of input values; and using a decision tree algorithm to automatically adjust the control coefficients of the electrical parameters based on the changing patterns of the input and output values, thereby achieving dynamic adjustment of the electrical parameters.

[0012] A method for controlling waste resin extrusion molding based on electrical parameters, wherein the dynamic adjustment of electrical parameters specifically involves the following: the electrical parameters are the input values ​​of the resin mixing and extrusion equipment, and the pressure and flow rate of the plasticizing and extruding of the resin mixing and extrusion equipment are the output values. By adjusting the input electrical parameters, the deviation of the output values ​​is reduced, so that the output values ​​meet the target values. The dynamic adjustment of electrical parameters is achieved through dynamic parameter adjustment.

[0013] A waste resin extrusion molding control method based on electrical parameters is proposed. This method involves dynamic parameter training and tuning. During extruder operation, motor control electrical parameters are set according to production process requirements. A frequency converter is configured to control the motor's frequency and speed. Pressure and flow rate sensors detect and control the output of the heated, plasticized, and extruded plastic. When the detected pressure and flow rate do not meet the set target requirements, the motor's input parameters are automatically adjusted to achieve closed-loop feedback control. When the production process changes, and the target value changes, a decision tree algorithm is used. After setting threshold parameters, if the output value does not match the target value, the decision tree is used to adjust the output value within the threshold range of current and voltage changes during actual production. When the distance between the adjusted output value and the actual target value decreases, it indicates that the parameter adjustment direction is correct. Dynamic matching will continue in this direction until the actual output value meets the target value requirements.

[0014] A waste resin extrusion molding control method based on electrical parameters, wherein the decision tree algorithm is specifically as follows: The collected electrical parameters or motor operating electrical parameters are filtered using the KNN clustering algorithm to form a filtered motor parameter dataset, which is then divided into a training set and a validation set. The selected motor parameter dataset is subjected to pattern analysis to obtain a decision tree model; The decision tree model is trained using the training set; The trained decision tree model is validated using a validation set. Training is stopped when the error rate calculated by the decision tree model is within the acceptable range of 3%-5% in actual production; otherwise, training continues until the error rate reaches the standard.

[0015] A method for controlling waste resin extrusion molding based on electrical parameters, wherein the method involves adjusting dynamic parameters, sampling and screening abnormal parameter values, and then establishing a time-series-based database. The time-series-based database stores process parameters for different production targets and processes them using the KNN clustering algorithm to filter out interfering values.

[0016] A waste resin extrusion molding control method based on electrical parameters, wherein the method, based on an established production process parameter model, achieves production process parameter control and fault early warning diagnosis, specifically as follows: Integrate the process parameters and various sensor parameters during the production process; A linear model is fitted to the production data based on the integrated data; Based on the linear model, when outlier changes occur during the reproduction process, the support vector machine calculation method is used to determine the pattern of outlier changes based on the distance difference between the outlier and the linear model. This allows for troubleshooting or adjusting production parameters in response to abnormal alarms.

[0017] A waste resin extrusion molding control method based on electrical parameters, wherein the prediction of outliers in the state parameters specifically involves using a DBSCAN clustering method, where all data points are defined as core points, boundary points, or noise points; core points are data points with at least a minimum number of contained points within a distance; boundary points are neighboring points of core points within a distance, but contain fewer points than the minimum number of contained points; all other data points are noise points and are also identified as outliers; thus, anomaly detection depends on the required minimum number of contained points, the distance, and the selected distance metric.

[0018] The beneficial effects of this invention are: This invention can improve the extrusion process control in the extrusion molding production of waste resin, increase the yield, reduce the cyclic stress of the extrusion screw, and extend the service life of the screw.

[0019] This invention effectively controls the stress cycle of the screw during the extrusion molding process of waste resin, establishes a mapping relationship between the active control parameters of the drive motor and the output pressure and speed, and creates a time-series database of the electrical parameters of the screw drive motor and sensor parameters. A clustering algorithm (KNN nearest neighbor average distance) is used to filter the data, establishing a production process prediction model and a fault prediction model. The control motor and the confluence core pressure sensor work together, achieving redundant control through mutual verification with the pressure sensor without increasing hardware costs. This enables precise control of the screw extrusion pressure and reduces screw failure rate.

[0020] This invention achieves redundant control of sensor detection and observation values ​​(extrusion pressure observation and flow rate). This redundant control utilizes parameters returned by sensors during production to control motor output. Once the parameters are adjusted, the motor control parameters become a mapping relationship with the sensors. When the sensors are off or malfunctioning, the motor can still reflect the parameter values ​​of the pressure and flow rate sensors based on its own electrical parameter status. The sensors do not need to operate continuously; periodic calibration is sufficient. This achieves redundant feedback control based on input electrical parameters and sensor detection.

[0021] This invention uses process parameters during production as target values ​​and adjusts electrical parameters to control torque and speed, ensuring that extrusion pressure and flow rate meet process requirements. However, these electrical parameters are inputs, and the pressure and flow rate at the final plasticizing and extrusion output deviate from the target process values ​​due to various interferences. The confluence core pressure sensor provides feedback on pressure, and the speed sensor provides flow parameters. By adjusting the input electrical parameters, the deviation is reduced, and the actual output value is controlled to meet the target value. Based on the change pattern of the output value after adjusting the input, the decision tree algorithm mentioned later is used to automatically adjust the control coefficient of the electrical parameters, achieving automatic parameter adjustment. When the production process changes, online parameter adjustment is completed autonomously. This not only meets the needs of the production process but also provides redundant control. In this production state, the matching of electrical parameters and output values ​​is a dual feedback control, thus enabling automatic parameter adjustment.

[0022] An outlier detection algorithm is used to observe the status of key parameters in the production process, eliminate interference factors, and establish a production process status information database based on data observation. Supervised learning (KNN, K-nearest neighbor) is used to train the machine to autonomously judge the production status and adjust the parameter status in the production process, thereby realizing production process parameter control and fault early warning diagnosis. Attached Figure Description

[0023] Figure 1 This is a flowchart of the method of the present invention.

[0024] Figure 2 This is a schematic diagram of the structure of the present invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] A method for controlling waste resin extrusion molding based on electrical parameters, the method comprising: Start the waste resin mixing and extrusion system and preheat the heating module; Predictive control of the motor control module is achieved through a database. The motor output shaft under the motor control module drives the screw torque to perform motion. When the base material is plasticized, it is extruded and output by the screw through the extrusion port. The pressure and flow rate of the plasticizing extrusion of the waste resin compounding extrusion system are detected to determine whether the pressure and flow rate parameters meet the target values. If so, the plasticized plastic is continuously output. If not, the detected pressure and flow rate parameters are fed back to the motor control module, and the motor control module is adjusted to control the pressure and flow rate.

[0027] A data-predictive method for controlling resin compounding and extrusion, wherein the predictive control specifically involves... The electrical parameters of the motors during the production process are collected, filtered, and then dynamically adjusted to achieve dynamic parameter regulation.

[0028] Based on dynamically adjusted electrical parameters, dynamic parameter regulation is achieved.

[0029] A data-predictive-based method for controlling resin compounding and extrusion, wherein the database is specifically established as follows: Based on the adjustment of dynamic parameters, after sampling and screening of abnormal parameter values, a time series-based process parameter database is established, and a production process parameter model is built.

[0030] A production process parameter model is established based on a time-series database of process parameters. Based on the established production process parameter model, supervised learning (KNN, K-nearest neighbor) is used to train the machine to autonomously judge the production status and adjust the parameter status in the production process, thereby realizing production process parameter control and fault early warning diagnosis.

[0031] A data-predictive-based resin compounding and extrusion control method, wherein the step of collecting parameters during the production process, filtering out abnormal parameter values, and dynamically adjusting electrical parameters specifically involves: using sensors to collect the output electrical parameters of the motor; using the KNN clustering algorithm to filter and remove abnormal parameter values ​​from the motor's output electrical parameters; using the remaining motor output electrical parameters as a database of input values; and using a decision tree algorithm to automatically adjust the control coefficients of the electrical parameters based on the changing patterns of the input and output values, thereby achieving dynamic adjustment of the electrical parameters.

[0032] A method for controlling waste resin extrusion molding based on electrical parameters, wherein the dynamic adjustment of electrical parameters specifically involves the following: the electrical parameters are the input values ​​of the resin mixing and extrusion equipment, and the pressure and flow rate of the plasticizing extrusion of the resin mixing and extrusion equipment are the output values. Since the output values ​​are affected by various interferences and deviate from the target process values, the pressure sensor of the merging core provides feedback on the pressure and the flow rate parameters of the velocity sensor. By adjusting the input electrical parameters, the deviation of the output values ​​is reduced, so that the output values ​​meet the target values. The dynamic adjustment of electrical parameters achieves the regulation of dynamic parameters.

[0033] Once the electrical parameters are adjusted and the motor control parameters are mapped to the sensors, the motor can still reflect the parameter values ​​of the pressure sensor and flow rate sensor based on its own electrical parameter status when the sensors are turned off or fail. The sensors only need to be calibrated periodically to maintain operation.

[0034] The adjusted parameters are specifically designed to be adjusted according to the production process. Generally, this depends on the motor selection. For example, permanent magnet synchronous motors usually come with a frequency converter. By adjusting the frequency, the current and motor speed can be controlled. If it is a dual feedback system, it can achieve bidirectional adjustment of current and speed. While adjusting, the production status is monitored. For example, if the required extrusion pressure and extrusion speed are set, the motor frequency is adjusted. If the applied pressure and extrusion speed are lower than the target values, the frequency and speed values ​​are adjusted to increase the current and speed.

[0035] A control method for waste resin extrusion molding based on electrical parameters involves dynamic parameter training and tuning. During extruder operation, motor control electrical parameters are set according to production process requirements. A frequency converter is configured to achieve frequency conversion control of the motor, enabling current and speed regulation. Pressure and flow rate sensors detect and control the output of the plastic during heating, plasticizing, and extrusion. When the detected pressure and flow rate do not meet the set target requirements, the motor input parameters are automatically adjusted. Closed-loop feedback control is achieved through PID control. These parameters require adjustment by experienced engineers based on the on-site environment and are fixed after production stabilizes. However, when the production process changes and the target value changes, these parameters need to be readjusted, requiring manual matching. A decision tree algorithm is used. After setting threshold parameters, if the output value does not match the target value, the decision tree is used to adjust the parameters within the threshold range of current and voltage changes during actual production. When the distance between the adjusted output value and the actual target value decreases, it indicates that the parameter adjustment direction is correct, and dynamic matching will continue in this direction until the actual output value meets the target value requirements.

[0036] A waste resin extrusion molding control method based on electrical parameters, wherein the decision tree algorithm is specifically as follows: The collected electrical parameters or motor operating electrical parameters are filtered using the KNN clustering algorithm to form a filtered motor parameter dataset, which is then divided into a training set and a validation set. The selected motor parameter dataset is subjected to pattern analysis. For example, the motor current required when it is in the start-up state is large and the voltage speed control is small. When the current decreases and the voltage increases to speed up during the production process, such a cyclical process is observed. The pattern analysis is carried out by extracting patterns based on clustering algorithms to obtain a decision tree model. The decision tree model is trained using the training set; The trained decision tree model is validated using a validation set. Training is stopped when the error rate calculated by the decision tree model is within the acceptable range of 3%-5% in actual production. That is, the error rate is acceptable when the impact on the process production is negligible. Otherwise, training continues until the error rate reaches the standard.

[0037] A method for controlling waste resin extrusion molding based on electrical parameters, wherein the method involves adjusting dynamic parameters, sampling and screening abnormal parameter values, and then establishing a time-series-based database. The time-series-based database stores process parameters for different production targets and processes them using the KNN clustering algorithm to filter out interfering values.

[0038] A waste resin extrusion molding control method based on electrical parameters, wherein the method, based on an established production process parameter model, achieves production process parameter control and fault early warning diagnosis, specifically as follows: Integrate the process parameters and various sensor parameters during the production process; A linear model is fitted to the production data based on the integrated data; Based on the linear model, when outlier changes occur during the reproduction process, the support vector machine calculation method is used to determine the pattern of outlier changes based on the distance difference between the outlier and the linear model. This allows for troubleshooting or adjusting production parameters in response to abnormal alarms.

[0039] When abnormal changes occur in the equipment, abnormal values ​​of status parameters can be predicted, and the parameter database of the trained model can guide the production process. Data traceability can quickly and accurately locate the problem points in the production process.

[0040] A method for controlling waste resin extrusion molding based on electrical parameters, wherein the prediction of outliers in the state parameters specifically involves using a DBSCAN clustering method. DBSCAN is a non-parametric, density-based outlier detection method in a one-dimensional or multi-dimensional feature space. In DBSCAN clustering, all data points are defined as core points, border points, or noise points. Core points are data points with at least a minimum number of contained points (minPTs) within a distance; border points are the nearest neighbors of core points, but with a number of contained points less than the minimum number of contained points (minPTs); all other data points are noise points and are also identified as outliers. Thus, outlier detection depends on the required minimum number of contained points, the distance, and the chosen distance metric, such as Euclidean or Manhattan distance.

[0041] The working principle and process of this invention are described in detail.

[0042] The basic principle of an extruder is that a rotating screw drives resin plastic forward. During this forward movement, the plastic is heated and plasticized. After plasticization, the plastic is extruded through the extrusion port by the rotating screw and output into the mold cavity. Then, it is molded and die-cast, ultimately completing the production of the plastic product. The screw is powered by the torque output of a motor. When the extrusion pressure increases, the motor load increases, and the corresponding current also rises. When the flow rate increases, the motor speed needs to increase, and the current will also increase. Therefore, the extrusion pressure and flow rate can be controlled through dual feedback control of the motor's speed loop and current loop. However, the output torque and speed, and the corresponding pressure output value, represent an ideal state. This working process requires feedback control, and once this parameter is adjusted, it is fixed. This state value can also achieve redundant control of pressure and flow sensors. An anomaly detection and prediction model is used to analyze the sensor parameters of the production equipment. The anomaly detection model estimates the working state of each sensor through data distribution clustering or probability density, establishes a process parameter database, constructs a supervised training production state prediction model, and predicts data entering the abnormal region, thereby achieving process quality prediction.

[0043] like Figure 2 As shown, the block diagram of the extruder equipment and control system includes an extruder power system module, a motor control module, a process processing module (extrusion pressure and temperature heating module, vacuuming, feeding device, temperature detection), a pressure / flow rate detection module, and an extrusion output module.

[0044] like Figure 1 The system shown starts up and enters the normal production work cycle. Set target values ​​for process parameters, adjust motor drive electrical parameters, and control screw rotation. Based on the real-time detection values ​​of the pressure sensor and flow rate sensor, a closed-loop feedback control is formed with the motor control parameters to adjust the electrical parameters and control the output to meet the target requirements. The parameter adjustment process introduces a decision tree algorithm. By adjusting the PID control parameters of the motor, the decision tree adjusts the three parameters P, I, and D to control the output torque and speed of the motor. The status feedback from the sensor is used to determine whether the parameters controlled by the decision tree meet the requirements. The feedback status verifies the actual effect of the parameters adjusted by the decision tree. Through iterative decision adjustment, the control target requirements are finally met. Motor frequency converter parameter control controls the output torque and speed of the motor. When the frequency converter duty cycle increases and the frequency increases, the output torque of the motor increases accordingly and the speed increases, and vice versa.

[0045] When the output extrusion pressure is detected to be lower than the rated value, the motor drive torque is increased; when the extrusion pressure is greater than the target value, the motor torque is reduced. Similarly, the flow rate control is also a closed loop. The motor's torque and speed are controlled by feedback from the current and encoder, and the motor's speed and torque operate within the rated threshold range. When the system enters a normal cycle, a database is established for the time-series parameters, and tables are created according to time, with fields for temperature sensor parameters, pressure sensor parameters, flow rate sensor parameters, and real-time electrical parameters of the motor. During the database establishment process, the DBSCAN outlier detection algorithm is used to filter out outliers. The filtered outliers are entered into another database table. When the outlier is verified to be caused by accidental interference without equipment downtime for maintenance, it can be entered into the database and archived for future time series tracing. If an anomaly is detected within the sensor's threshold range, and the equipment has been shut down for maintenance, it is determined that the anomaly is not accidental. A fault data tag is automatically created for future fault prediction. (The anomalies given here are definitely within the rated threshold range. In normal production, parameters such as temperature, pressure, and flow rate are set with boundaries. Based on production needs, these parameters should be within a reasonable range of minimum and maximum values. Exceeding a certain value is considered a fault. This is the simplest threshold judgment. However, anomaly detection can be achieved through cluster analysis to identify anomalies within the threshold range and their cluster density.) A process status database is established between the motor control electrical parameters and the sensor detection values. Outlier detection algorithms are used to filter out outliers, and the filtered data is then used to build a parameter database. After time series storage, a mapping relationship between the data change pattern of the pressure sensor and the electrical parameters can be established. Based on the mapping relationship, the motor electrical parameters can be directly controlled. The response pattern after adjusting torque / speed can be predicted through the time series database, thereby reducing or decreasing the number and magnitude of motor electrical parameter adjustments, thus reducing the magnitude and number of cyclic stress changes and extending the screw's service life. By establishing a database of motor electrical parameters, a set of reliable operating control parameters can be obtained based on the electrical parameters. For sensors, a time period can be set for power-on verification, thereby reducing the power-on time of the sensors and increasing their lifespan. The two work together to achieve redundant control. The database model established based on the various state parameters of the extruder can verify the operating status of the equipment and establish a database by comparing the real-time detected state values ​​with historical production states. The principle for establishing the process database is as follows: During normal operation, the data model of conventional production process parameters is established by storing various electrical parameters and sensor status parameters. Then, linear logistic regression is used to perform regression prediction processing on the conventional parameters to generate a process parameter prediction model. Based on this model, the motor electrical parameters are controlled, replacing the original PID feedback control strategy and reducing the number of cycles. When outliers are detected, they are filtered and categorized into several types, such as occasional outliers, periodically occurring outliers, and progressively changing outliers. A data model library is established for these categories. The status of the equipment is judged by outlier detection. For outliers exceeding the threshold range, a warning is issued to indicate potential faults. Occasional outliers are accumulated over time, and after long-term tracking, it can be verified whether they prove to be recurring patterns or occasional occurrences that will never reappear. For outliers with progressively changing relationships, timely responses are taken to detect fault points and issue alarms.

Claims

1. A method for controlling waste resin extrusion molding based on electrical parameters, characterized in that, The control method includes: Start the waste resin mixing and extrusion system and preheat the heating module; Predictive control of the motor control module is achieved through a database. The motor output shaft under the motor control module drives the screw torque to perform motion. When the base material is plasticized, it is squeezed out by the screw through the extrusion port. The pressure and flow rate of the plasticizing extrusion of the waste resin compounding extrusion system are detected to determine whether the pressure and flow rate parameters meet the target values. If so, the plasticized plastic is continuously output. If not, the detected pressure and flow rate parameters are fed back to the motor control module, and the motor control module is adjusted to control the pressure and flow rate. The predictive control specifically refers to... The electrical parameters of the motors during the production process are collected, filtered, and then dynamically adjusted to achieve dynamic parameter regulation. The database is established specifically as follows: Based on the adjustment of dynamic parameters, after sampling and screening of abnormal parameter values, a time series-based process parameter database is established. A production process parameter model is established based on a time-series database of process parameters. Based on the established production process parameter model, the machine is trained to autonomously judge the production status and adjust the parameter status during the production process, thereby realizing production process parameter control and fault early warning and diagnosis. The process of collecting parameters during the production process, filtering out abnormal parameter values, and then dynamically adjusting electrical parameters specifically involves: using sensors to collect electrical parameters output by the motor; using the KNN clustering algorithm to filter and remove abnormal parameter values ​​from the output electrical parameters; using the remaining electrical parameters output by the motor as a database of input values; and using a decision tree algorithm to automatically adjust the control coefficients of the electrical parameters based on the changing patterns of the input and output values, thereby achieving dynamic adjustment of the electrical parameters. Dynamic adjustment of electrical parameters is achieved by adjusting the input values ​​of the resin mixing and extrusion equipment and the output values ​​of the plasticizing and extruding pressure and flow rate of the resin mixing and extrusion equipment. By adjusting the input electrical parameters, the deviation of the output values ​​is reduced so that the output values ​​meet the target values. Dynamic parameter adjustment is achieved by dynamically adjusting the electrical parameters. During extruder operation, the electrical parameters for motor control are set according to production process requirements. The motor is equipped with a frequency converter for frequency conversion control, enabling current and speed regulation. Pressure and flow rate sensors detect and control the output of plastic during heating, plasticizing, and extrusion. When the detected pressure and flow rate do not meet the set target requirements, the input electrical parameters are automatically adjusted to achieve closed-loop feedback control. When the production process changes, as the target value changes, a decision tree algorithm is used to set threshold parameters. If the output value does not match the target value, the decision tree is used to adjust the output value within the range of threshold points for current and voltage changes during actual production. When the distance between the adjusted output value and the actual target value decreases, it indicates that the direction of parameter adjustment is correct. Dynamic matching will continue in this direction until the actual output value meets the target value requirements.

2. The waste resin extrusion molding control method according to claim 1, characterized in that, The decision tree algorithm is specifically as follows: The collected electrical parameters are filtered using the KNN clustering algorithm to form a filtered electrical parameter dataset, which is then divided into a training set and a validation set. The selected electrical parameter dataset was subjected to pattern analysis to obtain a decision tree model; The decision tree model is trained using the training set; The trained decision tree model is validated using a validation set. Training is stopped when the error rate calculated by the decision tree model is within the acceptable range of 3%-5% in actual production; otherwise, training continues until the error rate reaches the standard.

3. The waste resin extrusion molding control method according to claim 1, characterized in that, The database is established based on time series data after sampling and screening abnormal parameter values ​​by adjusting dynamic parameters. The time-series-based database stores process parameters for different production targets and processes them using the KNN clustering algorithm to filter out interfering values.

4. The waste resin extrusion molding control method according to claim 3, characterized in that, The aforementioned production process parameter control and fault early warning diagnosis based on the established production process parameter model specifically involves: Integrate the process parameters and various sensor parameters during the production process; A linear model is fitted to the production data based on the integrated data; Based on the linear model, when outlier changes occur during the production process, the support vector machine calculation method is used to determine the pattern of outlier changes based on the distance difference between the outlier and the linear model. This allows for troubleshooting or adjusting production parameters in response to abnormal alarms.

5. The waste resin extrusion molding control method according to claim 4, characterized in that, To predict outliers in parameter states, the DBSCAN clustering method is used, where all data points are defined as core points, boundary points, or noise points. Core points are data points that have at least a minimum number of contained points within a distance. Boundary points are the nearest points within a distance of the core point, but contain fewer points than the minimum number of points; all other data points are noise points and are also identified as outliers; thus, outlier detection depends on the required minimum number of points, the distance, and the chosen distance metric.

Images