Plastic wire drawing parameter self-adaptive regulation and control method and system fusing multi-sensor data

Abstract

The invention belongs to the technical field of plastic wire drawing, and particularly relates to a plastic wire drawing parameter self-adaptive regulation and control method and system fusing multi-sensor data, and the method comprises the steps: judging unqualified products, recording fluctuation nodes, merging continuous fluctuation nodes to generate a quality monitoring section, and extracting process parameters in the section as profile reference parameters; the to-be-operated parameters are compared with the reference parameters, temporary characteristic parameters are determined, and then the checking model is called to generate final characteristic parameters so as to regulate and control the wire drawing system; according to the method, the technological parameters associated with the continuous quality fluctuation section are traced, future parameters are corrected based on historical problem data, and cross-cycle systematic problems are processed through upgrading, so that the regulation and control accuracy and the long-term operation stability are improved.

Description

Technical Field

[0001] This invention belongs to the field of plastic filament technology, specifically relating to an adaptive control method and system for plastic filament parameters that integrates data from multiple sensors. Background Technology

[0002] Plastic filament drawing is a fundamental process in modern industrial production, and its products are widely used in key fields such as packaging, textiles, construction, and medicine. With the in-depth advancement of intelligent manufacturing technology, automated and intelligent production methods have become the core means for the plastic filament drawing industry to improve production efficiency and ensure product consistency.

[0003] Existing plastic filament drawing parameter control technologies still have many shortcomings in practice. Traditional control systems mainly rely on the equipment's own operating parameters (such as heating temperature and traction speed) for open-loop or semi-open-loop control, neglecting the online monitoring and closed-loop feedback of real-time quality indicators of the drawn samples (such as filament diameter uniformity, surface smoothness, and tensile strength). This results in a serious disconnect between control behavior and quality objectives, making it difficult to guarantee the stability of product quality. Summary of the Invention

[0004] This invention provides an adaptive control method for plastic filament drawing parameters that integrates data from multiple sensors, in order to solve the problems existing in the prior art.

[0005] To achieve the above-mentioned objectives, the technical solution adopted by the present invention is as follows:

[0006] When generating a quality monitoring section based on the quality information of the drawing cycle, the final characteristic parameters of the drawing parameters to be run are determined, and the profile reference parameters are extracted based on the quality monitoring section.

[0007] The drawing parameters to be run are compared with the profile reference parameters to output provisional feature parameters, and the verification model is called to process the provisional feature parameters to generate the final feature parameters.

[0008] The operation of the wire drawing system is controlled based on the final characteristic parameters.

[0009] Optionally, the steps for generating quality monitoring sections include:

[0010] Output time points where the sample deviation is greater than a preset deviation threshold are recorded as fluctuation nodes, and output time points where the sample deviation is not greater than a preset deviation threshold are recorded as stable nodes. The time range between adjacent stable nodes is defined as the fluctuation interval.

[0011] Based on the distribution of fluctuation nodes within the fluctuation range, at least one continuous fluctuation segment is identified, and at least one continuous fluctuation segment is merged to generate a quality monitoring segment.

[0012] Optionally, the step of identifying at least one continuous fluctuation segment based on the distribution of fluctuation nodes within the fluctuation range includes:

[0013] Calculate the time difference between adjacent fluctuation nodes ordered by time within the fluctuation range. If the time difference is greater than a preset time difference threshold, the time period corresponding to the time difference is identified as a separation gap, and the fluctuation range is divided into at least one continuous fluctuation segment based on the separation gap.

[0014] Optionally, the steps for extracting profile reference parameters based on the quality monitoring section include:

[0015] The time center point of the quality monitoring section is determined, and a time profile is formed by extending a preset amount of time along the time axis before and after the time center point. Then, the drawing speed, drawing temperature and drawing material recorded in the time profile are extracted as profile reference parameters.

[0016] Optionally, the step of comparing the wire drawing parameters to be run with the profile reference parameters to output provisional characteristic parameters includes:

[0017] The difference between the drawing parameters to be run and the corresponding parameters in the profile reference parameters is calculated or the material is compared and analyzed. Based on the preset judgment rules, the parameters that need to be corrected are determined as provisional characteristic parameters.

[0018] Optionally, the verification model call includes: selectively calling at least one of the following, depending on the type of the provisional characteristic parameter: a verification model based on the time difference between the start and end points of wire drawing, a verification model based on the time consumed per unit mass, and a verification model based on the material replacement cycle.

[0019] Optionally, the method further includes:

[0020] After adjusting the operation of the wire drawing system based on the final characteristic parameters, the adjustment results are evaluated; and,

[0021] If the assessment determines that the deviation of the sample produced after adjustment is greater than the preset deviation threshold, then compensation control will be implemented.

[0022] Optionally, the method further includes:

[0023] After generating the quality monitoring section, if the quality monitoring section appears continuously in multiple consecutive drawing cycles and exceeds the preset continuous cycle threshold, the duration of the quality monitoring section is extended to form a calibration section, and the calibration section is used to replace the quality monitoring section for subsequent processing.

[0024] This application also provides an adaptive control system for plastic drawing parameters that integrates multi-sensor data, applied to the adaptive control method for plastic drawing parameters that integrates multi-sensor data as described above. The system includes:

[0025] The state sensing unit is used to monitor quality information during the wire drawing process and generate quality monitoring sections based on the quality information;

[0026] The decision-making strategy unit extracts the profile reference parameters associated with the quality monitoring section, compares the wire drawing parameters to be run with the profile reference parameters to output provisional feature parameters, and calls the verification model to process the provisional feature parameters to generate the final feature parameters.

[0027] The execution control unit receives the final characteristic parameters from the decision-making strategy unit and adjusts the operation of the wire drawing equipment accordingly.

[0028] The control and evaluation unit is used to evaluate the production results after the control unit performs the control, and to trigger compensation control when it determines that the deviation of the sample produced after control is greater than the preset deviation threshold.

[0029] Optionally, the state sensing unit is used to record the output time point where the sample deviation is greater than a preset deviation threshold as a fluctuation node, and the output time point where the sample deviation is not greater than the preset deviation threshold as a stable node; it is used to define the time range between adjacent stable nodes as a fluctuation interval; and it is used to identify and merge at least one continuous fluctuation segment based on the distribution of fluctuation nodes within the fluctuation interval to generate a quality monitoring segment.

[0030] Beneficial effects

[0031] This invention distinguishes between qualified and unqualified products within the drawing cycle, records the output time of unqualified products as fluctuation nodes, defines the fluctuation range between stable nodes, further calculates the time difference between fluctuation nodes, and filters out isolated fluctuation events based on the time difference threshold. It merges temporally continuous fluctuation nodes into quality monitoring segments and extracts the drawing speed, temperature, and raw materials within the corresponding time profile of the segment as profile reference parameters. Thus, it can accurately locate the core time period that causes continuous fluctuations in product quality and effectively eliminate the interference of occasional quality fluctuation nodes on the cause analysis.

[0032] After obtaining the drawing parameters to be run, this invention compares them with the profile reference parameters extracted in the previous steps that are directly related to historical quality problems. Based on preset rules, it determines the provisional characteristic parameters that need to be corrected, and calls the verification model for temperature, drawing speed or raw material type to verify the provisional characteristic parameters, generating the final characteristic parameters for control. Thus, a predictive control mechanism based on historical data is established, realizing the adaptive adjustment of parameters.

[0033] In this invention, when a quality monitoring segment continuously appears and exceeds a preset continuous cycle threshold in multiple consecutive drawing cycles, the segment is automatically extended to form a calibration segment with a wider coverage. Based on this, subsequent parameter extraction and analysis are performed, thereby effectively identifying and handling recurring systemic or trend-based quality problems, rather than treating them as independent single events. Attached Figure Description

[0034] Figure 1 This is a flowchart of the method provided by the present invention. Detailed Implementation

[0035] Example 1

[0036] This embodiment provides an adaptive control method for plastic filament drawing parameters that integrates multi-sensor data. By dynamically monitoring and real-time correcting parameters during the production process, stable control of the quality of plastic filament drawing products can be achieved. The method includes the following steps:

[0037] When generating a quality monitoring section based on the quality information of the drawing cycle, the final characteristic parameters of the drawing parameters to be run are determined, and the profile reference parameters are extracted based on the quality monitoring section.

[0038] The drawing parameters to be run are compared with the profile reference parameters to output provisional feature parameters, and the verification model is called to process the provisional feature parameters to generate the final feature parameters.

[0039] The operation of the wire drawing system is controlled based on the final characteristic parameters.

[0040] Baseline calibration is performed while the wire drawing equipment operates in an idling state. The multi-sensor module collects energy consumption information, such as standby power and rotational inertia consumption. This data is integrated into baseline parameters, providing a zero-point reference for subsequent energy consumption anomaly analysis. At the start of a production task, system operating status information is collected and recorded. This information includes not only initial wire drawing temperature, initial wire drawing speed, and initial raw material as initialization parameters, but also real-time data continuously collected by the multi-sensor module during system operation. The real-time data constitutes a dynamic profile of the production process, specifically including wire drawing temperature, wire drawing speed, speed variation amplitude, instantaneous tension, and rotational power. The real-time data, together with the initialization parameters, constitutes the complete initial conditions for the wire drawing system's operation, used to initialize the system.

[0041] During the drawing cycle, the output time point and corresponding deviation degree of each sample are acquired through data acquisition equipment, such as the fluctuation value of the wire diameter. A deviation threshold is preset, which is the maximum acceptable deviation range determined according to the product process requirements. The deviation degree of each sample is compared with the deviation threshold in real time. If the deviation degree of a sample is greater than the deviation threshold, the sample is judged as a defective product, and its corresponding output time point is recorded by the system as a fluctuation node. If the deviation degree of a sample is not greater than the deviation threshold, it is judged as a qualified product, and its output time point is recorded as a stable node. The node information constitutes the basic data for subsequent quality problem diagnosis.

[0042] The time range between two temporally adjacent stable nodes is defined as the fluctuation interval. Specifically, the output time points of all fluctuation nodes within this fluctuation interval are extracted and sorted chronologically. To distinguish between isolated quality defects and continuous production anomalies, the time difference between adjacent sorted output time points is calculated, and a time difference threshold is set. If the calculated time difference is greater than the threshold, it means that there is a stable production time window between the two non-conforming products. Therefore, the time period between the two adjacent output time points corresponding to this time difference is identified as a separation gap. Based on all identified separation gaps, the original fluctuation interval is divided into at least one continuous fluctuation segment, each representing a period of continuous substandard production quality. All these continuous fluctuation segments are merged to generate a quality monitoring segment.

[0043] To address systemic issues across cycles, this method also includes a segment extension mechanism. After each quality monitoring segment is generated, it is determined whether the segment extends to the end of the current drawing cycle. If so, it is marked as a segment to be continued, and the number of consecutive cycles in which it occurs is counted. A preset continuous cycle threshold is set. If the count of the number of cycles exceeds the threshold, it is determined to be a long-term problem, and the duration of the quality monitoring segment is extended to form an extension segment. This extension segment is merged with the quality monitoring segment to generate a calibration segment, and this calibration segment is used to replace the quality monitoring segment for subsequent processing to provide a more macroscopic data foundation.

[0044] This application uses the generated quality monitoring or calibration sections as the core to plan time profiles. Specifically, it determines the time center point of each quality monitoring section, and uses this time center point as a reference to extend a preset amount of time along the time axis before and after this time center point to form a time profile. This profile is used to capture the complete parameter change process that causes quality fluctuations. From the time period corresponding to this time profile, the system-recorded drawing speed, drawing temperature, and drawing raw materials are extracted, and the extracted parameters are marked as profile reference parameters.

[0045] When parameters need to be determined for a new production task, the drawing parameters to be run are obtained, and these parameters are compared with the profile reference parameters. The comparison process includes: calculating the difference between the drawing temperature to be run and the drawing temperature in the profile reference parameters to obtain the temperature difference; calculating the difference between the drawing speed to be run and the drawing speed in the profile reference parameters to obtain the speed difference; and simultaneously, performing a material comparison analysis between the drawing raw material to be run and the drawing raw material in the profile reference parameters, for example, by comparing the batch number of the raw material or calling the chemical composition analysis data of the material system to obtain the raw material comparison result. This application internally presets parameter thresholds and compares the temperature difference and speed difference with their respective corresponding parameter thresholds. Based on the preset judgment rules, combined with the comparison results of the temperature difference, the speed difference, and the raw material comparison results, the parameters that need to be corrected are determined, and the parameters that need to be corrected are output as provisional characteristic parameters.

[0046] A verification model is invoked to process provisional characteristic parameters to generate corrected values. The model selection dynamically adapts based on the type of the provisional characteristic parameter. Specifically, if the provisional characteristic parameter is drawing temperature, a verification model based on the time difference between the start and end points of drawing is invoked. This model compensates for thermal inertia by analyzing the time difference between temperature commands and corresponding fluctuation nodes in historical data. If the provisional characteristic parameter is drawing speed, a verification model based on time consumed per unit mass is invoked. This model calculates the optimal speed correction value by establishing a correlation between speed, efficiency, and mass. If the provisional characteristic parameter is drawing raw material, a verification model based on the material replacement cycle of the extension section is invoked. This model is used to optimize the timing of raw material switching and transition parameters. This application verifies the provisional characteristic parameters using the corrected values ​​generated by the model to obtain the final characteristic parameters.

[0047] The verification model based on the time difference between the start and end points of wire drawing refers to the calculation model used to compensate for the temperature control delay caused by thermal inertia, and its specific definition is as follows:

[0048]

[0049] In the formula, Temperature correction value means the suggested adjustment amount to the current temperature setting; The thermal inertia compensation coefficient is a proportional constant that characterizes the degree of hysteresis in the system's thermal response, and is obtained through calibration using historical data. The time difference function represents the time delay between the temperature command observed at time point t and the mass fluctuation. This represents the target time difference, which is the system response time under ideal conditions with no delay. It is usually set to zero or a very small value. This represents the integration time interval, which is the continuous fluctuation range for analysis.

[0050] The verification model based on unit mass consumption time refers to a computational model used to calculate the optimal speed correction value by correlating production efficiency with product quality. Its specific definition is as follows:

[0051]

[0052] In the formula, This indicates the speed correction value, which represents the suggested adjustment amount for the current wire drawing speed. The speed-efficiency gain coefficient is a proportional constant that characterizes the sensitivity of speed adjustment to production efficiency; it is calibrated experimentally. This represents the time taken to produce a unit mass of product, specifically the time actually spent producing that unit mass. The calculation method is as follows: ; This indicates the time taken to achieve the target unit mass, which represents the ideal time taken under the standard production efficiency required by the process.

[0053] The verification model based on the material replacement cycle of the extended segment refers to a computational model used to optimize the timing of raw material switching and reduce quality problems caused by unstable raw material transitions. Its specific definition is as follows:

[0054]

[0055] In the formula, This indicates the lead time for switching, which means the amount of time that is recommended to advance the next switching operation for the same type of raw materials. The material transition sensitivity coefficient is a dimensionless proportionality coefficient that characterizes the relationship between the duration of quality problems caused by improper switching and the required adjustment time. It is obtained through regression analysis of historical data. This indicates the duration of the extended segment, which represents the total duration of continuous quality issues across cycles. This indicates the theoretical residence time, which is the theoretical time required for new raw materials to completely replace the old raw materials after entering the equipment.

[0056] This application can also send the final characteristic parameters to the drawing equipment to regulate the operation of the drawing system. After regulation, the deviation of newly produced samples is continuously collected and compared with the deviation threshold to evaluate the regulation result. If the deviation is not greater than the deviation threshold, it is recorded as effective regulation. If, in three or more consecutive regulations, the deviation of the produced samples after each regulation is not greater than the deviation threshold, and the root cause of the deviation points to the same drawing parameter, the last regulation in this series is also recorded as effective regulation. If the deviation of the sample after regulation is greater than the deviation threshold, compensation control is executed. Compensation control is a continuous closed-loop fine-tuning process, including continuously adjusting the drawing temperature and drawing speed until the deviation of the produced samples after regulation is not greater than the deviation threshold, ensuring that production recovers to stability as soon as possible.

[0057] Example 2

[0058] This embodiment provides an adaptive control system for plastic filament drawing parameters that integrates data from multiple sensors. The system includes a filament drawing device and a control platform. The control platform is divided into the following collaboratively working units:

[0059] The state sensing unit continuously monitors the quality information during the wire drawing process and generates quality monitoring sections based on the quality information. The quality information may include, but is not limited to, real-time data collected by laser diameter gauges, tension sensors, surface defect detectors, etc.

[0060] In a specific execution process, the state perception unit compares the collected real-time quality data with the preset standard product specifications, calculates the sample deviation at each time point, records the output time point where the sample deviation is greater than a preset deviation threshold as a fluctuation node, and records the output time point where the sample deviation is not greater than the preset deviation threshold as a stable node. Thus, by identifying two adjacent stable nodes on the time axis, the time range between them is defined as a fluctuation interval, which represents a period of potential quality instability.

[0061] Furthermore, the state perception unit analyzes the distribution of fluctuation nodes ordered chronologically within the fluctuation range and calculates the time difference between adjacent fluctuation nodes. If the time difference is greater than a preset time difference threshold, this longer time interval is considered a separation gap, indicating that the quality fluctuation may be interrupted or belong to fluctuation groups with different causes. Based on one or more identified separation gaps, the unit divides the entire fluctuation range into at least one continuous fluctuation segment and merges all identified continuous fluctuation segments to generate the final quality monitoring segment. This segment reflects the critical time range for parameter adjustment.

[0062] The state-aware unit can also record the occurrence of quality monitoring segments in multiple consecutive drawing cycles. If a certain type of quality monitoring segment is found to appear continuously, and its duration exceeds a preset continuous cycle threshold, it indicates that the problem is systematic or periodic. In this case, the unit will actively extend the duration of the quality monitoring segment to form a wider calibration segment, and use this calibration segment to replace the original quality monitoring segment, so that subsequent units can perform more in-depth analysis and adjustment.

[0063] Once the decision strategy unit identifies the quality monitoring section, it is activated and generates a set of optimized and verified final characteristic parameters for the drawing parameters to be run (e.g., parameters planned for the next production batch or the next stage).

[0064] The decision-making strategy unit receives the quality monitoring segment (or calibration segment) determined by the state perception unit. Based on the time range of this segment, it determines its time center point. Using this time center point as a reference, it extends forward and backward along the time axis by a preset time amount, thus forming a time profile. The unit extracts key process parameters such as the drawing speed, temperature, and raw materials used at that time from the historical database covered by this time profile. These parameters together constitute the profile reference parameters, which represent the actual operating conditions before and after the period that caused the current quality fluctuation problem.

[0065] The decision-making strategy unit compares the drawing parameters to be executed with the extracted profile reference parameters. Specifically, for numerical parameters (such as drawing speed and temperature), difference calculation is performed; for categorical parameters (such as raw materials), material comparison analysis is performed. Thus, based on preset judgment rules (e.g., a rule base or a machine learning-based decision model), by determining which parameters to be executed have significant differences from the reference parameters and may lead to quality problems, the parameters that need to be corrected are identified as provisional feature parameters.

[0066] The decision-making strategy unit invokes one or more verification models to process provisional characteristic parameters. Specifically, the invocation is selective based on the type of provisional characteristic parameter: for example, if the provisional characteristic parameter involves speed adjustment, a verification model based on the time difference between the start and end points of wire drawing may be invoked to account for the system's inertial delay; if it involves raw material ratio or extrusion rate, a verification model based on the time consumed per unit mass may be invoked; if it involves changing different batches or types of raw materials, a verification model based on the material replacement cycle of the extension section is invoked to account for the time required for material replacement within the pipeline.

[0067] The execution control unit receives the final characteristic parameters from the decision-making strategy unit. After receiving the parameters, it parses them into specific control commands (such as voltage, frequency, valve opening, etc.) and sends them to the corresponding actuators in the wire drawing system (such as variable frequency motors, heating controllers, raw material supply pumps, etc.) through the control bus or network interface. Based on the final characteristic parameters, it precisely controls the operating status of the wire drawing equipment, thereby applying the optimized process parameters to actual production.

[0068] After the control unit completes the control and production begins, the control and evaluation unit, like the state sensing unit, continuously monitors the quality information of the newly produced samples and evaluates the control results. Specifically, it calculates the degree of deviation of the sample after control and compares it with a preset deviation threshold. If the evaluation determines that the degree of deviation of the produced sample after control is still greater than the preset deviation threshold, it indicates that the single control has not completely solved the problem or that there is unforeseen interference. At this time, compensation control can be triggered. Compensation control can be used to start a new round of "sensing-decision-execution" cycle, or to issue an alarm to the system operator and provide diagnostic suggestions for further manual intervention to adjust the strategy.

[0069] Through the coordinated operation of the aforementioned state perception unit, decision-making strategy unit, execution control unit, and regulation and evaluation unit, this application can form a complete adaptive closed-loop control loop. This not only enables rapid response to quality fluctuations during the production process but also allows for more accurate and reliable decision-making through historical data profile analysis and multi-model verification, thereby improving the automation level, product quality, and production efficiency of plastic filament production.

[0070] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for adaptive control of plastic filament drawing parameters by fusing data from multiple sensors, characterized in that, include: When generating a quality monitoring section based on the quality information of the drawing cycle, the final characteristic parameters of the drawing parameters to be run are determined, and the profile reference parameters are extracted based on the quality monitoring section. The drawing parameters to be run are compared with the profile reference parameters to output provisional feature parameters, and the verification model is called to process the provisional feature parameters to generate the final feature parameters. The operation of the wire drawing system is controlled based on the final characteristic parameters.

2. The method for adaptive control of plastic filament drawing parameters by fusing multi-sensor data according to claim 1, characterized in that, The steps for generating quality monitoring sections include: Output time points where the sample deviation is greater than a preset deviation threshold are recorded as fluctuation nodes, and output time points where the sample deviation is not greater than a preset deviation threshold are recorded as stable nodes. The time range between adjacent stable nodes is defined as the fluctuation interval. Based on the distribution of fluctuation nodes within the fluctuation range, at least one continuous fluctuation segment is identified, and at least one continuous fluctuation segment is merged to generate a quality monitoring segment.

3. The method for adaptive control of plastic filament drawing parameters by fusing multi-sensor data according to claim 2, characterized in that, The steps for identifying at least one continuous fluctuation segment based on the distribution of fluctuation nodes within the fluctuation range include: Calculate the time difference between adjacent fluctuation nodes ordered by time within the fluctuation range. If the time difference is greater than a preset time difference threshold, the time period corresponding to the time difference is identified as a separation gap, and the fluctuation range is divided into at least one continuous fluctuation segment based on the separation gap.

4. The method for adaptive control of plastic filament drawing parameters by fusing multi-sensor data according to claim 1, characterized in that, The steps for extracting profile reference parameters based on quality monitoring sections include: The time center point of the quality monitoring section is determined, and a time profile is formed by extending a preset amount of time along the time axis before and after the time center point. Then, the drawing speed, drawing temperature and drawing material recorded in the time profile are extracted as profile reference parameters.

5. The method for adaptive control of plastic filament drawing parameters by fusing multi-sensor data according to claim 1, characterized in that, The steps for comparing the wire drawing parameters to be executed with the profile reference parameters to output provisional characteristic parameters include: The difference between the drawing parameters to be run and the corresponding parameters in the profile reference parameters is calculated or the material is compared and analyzed. Based on the preset judgment rules, the parameters that need to be corrected are determined as provisional characteristic parameters.

6. The method for adaptive control of plastic filament drawing parameters by fusing multi-sensor data according to claim 1, characterized in that, The verification model call includes: selectively calling at least one of the following, depending on the type of provisional characteristic parameter: a verification model based on the time difference between the start and end points of wire drawing, a verification model based on the time consumed per unit mass, and a verification model based on the material replacement cycle.

7. The method for adaptive control of plastic filament drawing parameters by fusing multi-sensor data according to claim 1, characterized in that, The method also includes: After adjusting the operation of the wire drawing system based on the final characteristic parameters, the adjustment results are evaluated; and, If the assessment determines that the deviation of the sample produced after adjustment is greater than the preset deviation threshold, then compensation control will be implemented.

8. The method for adaptive control of plastic filament drawing parameters by fusing multi-sensor data according to claim 3, characterized in that, The method also includes: After generating the quality monitoring section, if the quality monitoring section appears continuously in multiple consecutive drawing cycles and exceeds the preset continuous cycle threshold, the duration of the quality monitoring section is extended to form a calibration section, and the calibration section is used to replace the quality monitoring section for subsequent processing.

9. A plastic filament drawing parameter adaptive control system integrating multi-sensor data, characterized in that, The system, applied to the adaptive control method for plastic filament drawing parameters fused from multi-sensor data as described in any one of claims 1-8, comprises: The state sensing unit is used to monitor quality information during the wire drawing process and generate quality monitoring sections based on the quality information; The decision-making strategy unit extracts the profile reference parameters associated with the quality monitoring section, compares the wire drawing parameters to be run with the profile reference parameters to output provisional feature parameters, and calls the verification model to process the provisional feature parameters to generate the final feature parameters. The execution control unit receives the final characteristic parameters from the decision-making strategy unit and adjusts the operation of the wire drawing equipment accordingly. The control and evaluation unit is used to evaluate the production results after the control unit performs the control, and to trigger compensation control when it determines that the deviation of the sample produced after control is greater than the preset deviation threshold.

10. The adaptive control system for plastic filament drawing parameters fused with multi-sensor data according to claim 9, characterized in that, The state sensing unit is used to record the output time point where the sample deviation is greater than the preset deviation threshold as a fluctuation node, and the output time point where the sample deviation is not greater than the preset deviation threshold as a stable node; it is used to define the time range between adjacent stable nodes as a fluctuation interval; and it is used to identify and merge at least one continuous fluctuation segment based on the distribution of fluctuation nodes within the fluctuation interval to generate a quality monitoring segment.

Images