Method for realizing real-time optimization and reverse control of cable wrapping
By using multi-source data acquisition, structured storage, and large model-driven process analysis, combined with reverse control components, the problem of inaccurate parameter adjustment during cable wrapping was solved. This enabled adaptive optimization and closed-loop control of the wrapping process, reducing scrap rate and improving product quality consistency.
Patent Information
- Application Number
- CN202511846066.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-09
- Publication Date
- 2026-03-17
AI Technical Summary
Existing technologies lack the ability to fuse and analyze multi-source heterogeneous data during cable manufacturing, making it impossible to achieve real-time quality feedback and dynamic adjustment of process parameters. This results in the wrapping pitch not meeting requirements and a high scrap rate.
By acquiring multi-source process data in real time, storing structured data, performing process analysis driven by a large model, and establishing a human-machine collaborative confirmation mechanism, the wrapping process is adaptively optimized and controlled in a closed loop by combining reverse control components. Parameter adjustment is performed using the Modbus TCP protocol and the PLC.
It achieves dynamic optimization of the copper strip wrapping process, reduces scrap rate, improves product quality consistency and production line intelligence, and supports rapid adaptation to new cable specifications.
Smart Images
Figure CN121680185A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of industrial artificial intelligence, and in particular relates to a method for realizing real-time optimization and reverse control of cable wrapping. BACKGROUND
[0002] In the cable manufacturing process, the wrapping process is one of the core links affecting product quality. Copper strip wrapping needs to strictly control parameters such as wrapping pitch, copper strip tension, cable traction speed, motor speed, etc., to ensure that the shielding layer is uniform, wrinkle-free, and has no overlap or gap. The traditional process relies on the experience of operators to set fixed parameters. Due to the numerous and complex factors in the wrapping process, it is easy to cause the pitch data to not meet the process requirements, and manual detection of wrapping pitch changes in real time is difficult, resulting in an increase in scrap rate.
[0003] Although existing automatic systems can monitor some parameters, they generally lack the ability to fuse and analyze multi-source heterogeneous data, and cannot dynamically adjust process parameters based on real-time quality feedback.
[0004] In recent years, artificial intelligence models have been tried for process prediction, but there are three major bottlenecks:
[0005] (1) The data acquisition protocol is not unified, making it difficult to access PLCs, encoders, visual quality inspection instruments, and other types of equipment;
[0006] (2) The model has weak generalization ability and cannot quickly adapt to new specifications of cables or new materials with a small number of samples;
[0007] (3) There is a lack of closed-loop linkage mechanism with production equipment, and the analysis results are only for reference, and cannot automatically execute process adjustment.
[0008] Therefore, there is an urgent need for an intelligent system that integrates "perception-analysis-decision-execution" to realize the adaptive optimization and closed-loop control of the wrapping process.
[0009] Based on the above, the present application provides a method for realizing real-time optimization and reverse control of cable wrapping. SUMMARY
[0010] The present application provides a simple and efficient method for realizing real-time optimization and reverse control of cable wrapping to overcome the shortcomings of the prior art.
[0011] The present application is achieved by the following technical solutions:
[0012] A method for realizing real-time optimization and reverse control of cable wrapping, comprising the following steps:
[0013] Step S1, real-time acquisition of multi-source process data
[0014] In the cable production process, the data acquisition component is used to obtain real-time data of the wrapping process, including wrapping pitch, main motor speed, cable traction speed, copper strip width, tension sensor readings, and surface defect information output by the line vision inspection system.
[0015] In step S1, the data acquisition component connects the servo motor through the Modbus TCP industrial protocol to obtain the main motor speed, calculates the wrapping pitch through the encoder, obtains the cable traction speed through the laser speedometer, and outputs the wrinkle, no wrinkle, or overlap label through the AI inspection module by shooting the cable surface image every second through the industrial camera.
[0016] Step S2, structured data storage
[0017] The collected time series data, image metadata, and device status information are written into the time series database and the relational database through the data storage component to establish a process quality association data set.
[0018] In step S2, the real-time data of the wrapping process during the cable production process is written into the time series database in real time.
[0019] The image data collected by the line vision inspection system in real time is stored in the MinIO object storage.
[0020] In the relational database, the meta information of each cable is recorded, including process batch ID, material batch, and operator, and these meta information is associated with the corresponding time series data and image data.
[0021] Step S3, process analysis driven by large model
[0022] The process analysis component loads a pre-trained large model, fine-tunes the large model based on historical high-quality process data and abnormal data, and fine-tunes the large model.
[0023] The fine-tuned large model set is used to analyze whether there is deviation in the current wrapping state in real time, and generate process optimization suggestions.
[0024] In step S3, the process analysis component sets a timing task, calls the large model periodically, inputs multi-dimensional time series data, receives feedback from the large model, and feeds back to the terminal for confirmation by the on-site engineer according to the actual production situation.
[0025] In step S3, the large model uses a time series Transformer model or a multi-modal large model to support joint inference of multi-source heterogeneous time series data.
[0026] Step S4, human-machine collaborative confirmation mechanism
[0027] Considering on-site production, to reduce the problem of inaccurate model adjustment recommendations, process optimization recommendations are pushed to the operation terminal, and whether to adopt them is confirmed by the on-site engineer;
[0028] If it is confirmed to be adopted, a reverse control process is triggered;
[0029] If not, return to step S1;
[0030] Step S5, production equipment reverse control
[0031] The reverse control component sends control signals to the programmable logic controller PLC or servo driver through the industrial communication protocol according to the confirmed instructions, controls the dynamic adjustment of the wrapping machine, and realizes the closed-loop adjustment of the process parameters;
[0032] In step S5, the reverse control component has a safety control strategy, including parameter adjustment range limitation, permission verification, operation trace, and emergency stop linkage mechanism, and writes the adjustment suggestions output by the process analysis component to the programmable logic controller PLC through the Modbus TCP communication protocol;
[0033] At the same time, the wrapping pitch acquisition device measures the wrapping pitch in real time and feeds back to the operation terminal.
[0034] Step S6, effect feedback and model iteration
[0035] Continuously collect new process data and quality inspection results after adjustment to evaluate the optimization effect, and use them as incremental data for online fine-tuning and continuous optimization of large models.
[0036] A system for realizing real-time optimization and reverse control of cable wrapping, for realizing the above method, comprising:
[0037] Data acquisition component: integrated protocol compatible engine, supporting Modbus TCP, Modbus RTU and other industrial protocols, supporting access to encoders (measuring pitch), servo motors (reading speed), laser speed meters (measuring line speed) and industrial cameras, realizing millisecond-level synchronous sampling;
[0038] Data storage component: build a unified data lake architecture, store time series parameters in a time series database, store images and quality inspection results in an object storage (such as MinIO object storage), and store process events and label information in a relational database, support efficient query and backtracking;
[0039] Process analysis component: deploy an intelligent agent based on a large model, the large model is pre-trained using historical high-quality wrapping data in the offline stage, and in the online stage, it combines real-time input for inference, outputs process health score and specific adjustment suggestions;
[0040] Reverse control component: configure safety control strategy (such as pitch adjustment, emergency stop linkage), send target parameter value to the winding machine programmable logic controller (PLC) through communication with the winding machine, and ensure the safety and reliability of the control process.
[0041] An apparatus for realizing real-time optimization and reverse control of cable winding, comprising a memory and a processor; the memory is used to store a computer program, and the processor is used to execute the computer program to realize the method steps described above.
[0042] A readable storage medium, the readable storage medium has a computer program stored thereon, the computer program is executed by a processor to realize the method steps described above.
[0043] The beneficial effects of the present application are: the method for realizing real-time optimization and reverse control of cable winding is suitable for dynamic optimization and automatic adjustment of key parameters such as pitch, tension and speed in the copper strip winding process, supports quick adaptation to new specification cables, can significantly reduce the winding defect rate, and improves product quality consistency and production line intelligent level. BRIEF DESCRIPTION OF DRAWINGS
[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.
[0045] Appendix Figure 1 The method for realizing real-time optimization and reverse control of cable winding of the present application is shown in the figure. DETAILED DESCRIPTION
[0046] In order to make those skilled in the art better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should be within the scope of protection of the present application.
[0047] The method for realizing real-time optimization and reverse control of cable winding comprises the following steps:
[0048] Step S1, real-time acquisition of multi-source process data
[0049] In the cable production process, the data acquisition component is used to obtain real-time data of the wrapping process, including wrapping pitch, main motor speed, cable traction speed, copper strip width, tension sensor readings, and surface defect information output by the line vision inspection system.
[0050] In step S1, the data acquisition component connects the servo motor through the Modbus TCP industrial protocol to obtain the main motor speed, calculates the wrapping pitch through the encoder, obtains the cable traction speed through the laser speedometer, and outputs the wrinkle, no wrinkle, or overlap label by the AI inspection module by shooting the cable surface image every second through the industrial camera.
[0051] Step S2, structured data storage
[0052] The collected time series data, image metadata, and equipment status information are written into the time series database and the relational database through the data storage component to establish a process quality association data set.
[0053] In step S2, the real-time data of the wrapping process during cable production is written into the time series database in real time.
[0054] The image data collected by the line vision inspection system in real time is stored in the MinIO object storage.
[0055] In the relational database PostgreSQL, the meta information of each cable is recorded, including process batch ID, material batch, and operator, and these meta information is associated with the corresponding time series data and image data.
[0056] Step S3, process analysis driven by large models
[0057] The process analysis component loads a pre-trained large model, fine-tunes the large model based on historical high-quality process data and abnormal data, and generates process optimization suggestions (such as "adjust the pitch to a certain value").
[0058] The fine-tuned large model set is used to analyze whether the current wrapping state has deviations in real time, and generate process optimization suggestions (such as "adjust the pitch to a certain value").
[0059] In step S3, the process analysis component customizes a timing task, periodically (e.g., every 15 seconds) calls the large model, inputs the multi-dimensional time series data of the last 30 seconds, receives the feedback of the large model, and feeds back to the terminal, which is confirmed by the on-site engineer according to the actual production situation.
[0060] In step S3, the large model uses a time series Transformer model or a multi-modal large model to support joint inference of multi-source heterogeneous time series data.
[0061] Step S4, human-machine collaborative confirmation mechanism
[0062] To mitigate issues arising from inaccurate model adjustment suggestions during on-site production, process optimization suggestions are pushed to the operation terminal for confirmation by on-site engineers regarding their adoption.
[0063] If the adoption is confirmed, the reverse control flow is triggered;
[0064] If not adopted, return to step S1;
[0065] Step S5: Reverse control of production equipment
[0066] The reverse control component sends control signals to the programmable logic controller (PLC) or servo driver via industrial communication protocol after receiving the confirmed instructions. This controls the wrapping machine to dynamically adjust the wrapping parameters and achieve closed-loop adjustment of the process parameters.
[0067] In step S5, the reverse control component is equipped with a safety control strategy, including parameter adjustment range limitation, permission verification, operation logging and emergency stop linkage mechanism, and writes the adjustment suggestions output by the process analysis component to the programmable logic controller PLC through the Modbus TCP communication protocol.
[0068] Meanwhile, the wrapping pitch acquisition device measures the wrapping pitch in real time and feeds it back to the operation terminal.
[0069] Step S6: Feedback and Model Iteration
[0070] We continuously collect and adjust the new process data and quality inspection results to evaluate the optimization effect, and use them as incremental data to achieve online fine-tuning and continuous optimization of the large model.
[0071] This system, which enables real-time optimization and reverse control of cable wrapping, is used to implement the above method, including:
[0072] Data acquisition component: Integrated protocol compatible engine, supports industrial protocols such as Modbus TCP and Modbus RTU, supports access to encoders (pitch measurement), servo motors (speed reading), laser velocimeters (linear speed measurement) and industrial cameras, and achieves millisecond-level synchronous sampling;
[0073] Data storage components: Construct a unified data lake architecture, store time series parameters in a time series database, store images and quality inspection results in object storage (such as MinIO object storage), and store process events and label information in a relational database PostgreSQL, supporting efficient querying and backtracking;
[0074] Process analysis component: Deploy an intelligent agent based on a large model. The large model is pre-trained using historical high-quality wrapping data in the offline stage, and inference is performed in the online stage by combining real-time input to output process health score and specific adjustment suggestions;
[0075] Reverse control component: Configures safety control strategies (such as pitch adjustment and emergency stop linkage), and sends target parameter values to the programmable logic controller (PLC) of the wrapping machine through communication to ensure the safety and reliability of the control process.
[0076] The device for real-time optimization and reverse control of cable wrapping includes a memory and a processor; the memory is used to store a computer program, and the processor is used to execute the computer program to implement the above-described method steps.
[0077] The readable storage medium stores a computer program that, when executed by a processor, implements the above-described method steps.
[0078] Compared with existing technologies, this method for real-time optimization and reverse control of cable wrapping has the following characteristics:
[0079] (1) Full-element perception: covering key parameters and quality results of the wrapping process, realizing the full-link data connection of "parameter-quality";
[0080] (2) Strong intelligent decision-making ability: It uses large models to capture complex nonlinear process laws, surpassing traditional rule engines or shallow models.
[0081] (3) Closed-loop control: For the first time, a semi-automatic closed loop of "AI analysis → manual confirmation → equipment execution" is realized, which takes into account both intelligence and safety;
[0082] (4) Quickly adapt to new specifications: By fine-tuning the large model with a small number of samples, it can quickly support the wrapping process of new cables or copper strip materials;
[0083] (5) Reduce scrap rate: It can significantly reduce the wrapping defect rate and reduce the frequency of manual intervention;
[0084] (6) Complies with industrial safety standards: The reverse control is equipped with permission verification, operation logging and physical limit, which can further ensure production safety.
[0085] The embodiments described above are merely one specific implementation of the present invention. Ordinary changes and substitutions made by those skilled in the art within the scope of the technical solution of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for implementing real-time optimization and inverse control of cable wrapping, characterized by: Comprising the following steps: Step S1, real-time acquisition of multi-source process data In the cable production process, real-time acquisition of wrapping process data is realized through a data acquisition component, including wrapping pitch, main motor speed, cable traction speed, copper strip width, tension sensor readings, and surface defect information output by the line visual inspection system; Step S2, structured data storage The collected time series data, image metadata and equipment status information are written into the time series database and the relational database through the data storage component, and the process quality correlation data set is established; Step S3, process analysis driven by large model The process analysis component loads a pre-trained large model, fine-tunes the large model based on historical high-quality process data and abnormal data, and generates process optimization suggestions; Step S4, man-machine collaborative confirmation mechanism The process optimization suggestions are pushed to the operation terminal, and the on-site engineer confirms whether to adopt them; If it is confirmed, the reverse control process is triggered; If not, return to step S1; Step S5, reverse regulation of production equipment According to the confirmed instructions, the reverse control component sends control signals to the programmable logic controller PLC or servo driver through the industrial communication protocol to control the wrapping machine to dynamically adjust the wrapping parameters, realizing closed-loop adjustment of process parameters; Step S6, effect feedback and model iteration Continuously collect new process data and inspection results after adjustment to evaluate the optimization effect, and use the incremental data to realize online fine-tuning and continuous optimization of the large model.
2. The method for realizing real-time optimization and reverse control of cable wrapping according to claim 1, wherein: In step S1, the data acquisition component connects the servo motor through the Modbus TCP industrial protocol to obtain the main motor speed, calculates the wrapping pitch through the encoder, obtains the cable traction speed through the laser speedometer, and outputs the wrinkle, no wrinkle or overlap label through the AI inspection module by shooting the cable surface image every second through the industrial camera.
3. The method for realizing real-time optimization and reverse control of cable wrapping according to claim 1, wherein: In step S2, the real-time acquisition of wrapping process data in the cable production process is written into the time series database in real time; The image data collected by the line visual inspection system is stored in the object storage; In the relational database, the meta information of each cable is recorded, including process batch ID, material batch and operator, and these meta information is associated with corresponding time series data and image data.
4. The method for realizing real-time optimization and reverse control of cable wrapping according to claim 1, wherein: In step S3, the process analysis component customizes a timing task, periodically calls the large model once, inputs multi-dimensional time series data, receives the feedback of the large model, and feeds back to the terminal for confirmation by the on-site engineer according to the actual production situation.
5. The method for realizing real-time optimization and reverse control of cable wrapping according to claim 4, wherein: In the step S3, the large model adopts a time series Transformer model or a multi-modal large model to support joint inference on multi-source heterogeneous time series data.
6. The method for realizing real-time optimization and reverse control of cable wrapping according to claim 1, wherein: In the step S5, the reverse control component is provided with a safety control strategy, including parameter adjustment range limitation, permission verification, operation trace leaving and emergency stop linkage mechanism, and writes the adjustment suggestion output by the process analysis component to the programmable logic controller (PLC) through the Modbus TCP communication protocol; Meanwhile, the wrapping pitch acquisition device measures the wrapping pitch in real time and feeds back to the operation terminal.
7. A system for implementing real-time optimization and inverse control of cable wrapping, characterized by: A device for implementing the method of any one of claims 1 to 6, comprising a data acquisition component: an integrated protocol compatible engine supporting Modbus TCP and Modbus RTU industrial protocols, supporting access to encoders, servo motors, laser speed meters and industrial cameras, and realizing millisecond-level synchronous sampling; a data storage component: a unified data lake architecture is constructed, time series parameters are stored in a time series database, images and quality inspection results are stored in an object storage, and process events and label information are stored in a relational database, supporting efficient query and backtracking; a process analysis component: an intelligent agent based on a large model is deployed, the large model is pre-trained using historical high-quality wrapping data in an offline stage, and in an online stage, real-time input is combined for inference, outputting process health score and specific adjustment suggestions; a reverse control component: a safety control strategy is configured, target parameter values are sent to the programmable logic controller (PLC) of the wrapping machine through communication with the wrapping machine, and the control process is ensured to be safe and reliable.
8. An apparatus for implementing real-time optimization and inverse control of cable wrapping, the apparatus comprising: a cable wrapping system; a cable wrapping controller; a cable wrapping optimizer; and a cable wrapping optimizer controller. comprising a memory and a processor; the memory is used to store a computer program, and the processor is used to execute the computer program to realize the method of any one of claims 1 to 6.
9. A readable storage medium characterized by: The readable storage medium has a computer program stored thereon, and the computer program is executed by a processor to realize the method of any one of claims 1 to 6.