A winding device and a winding method for tire steel wires

Abstract

The application discloses a winding device and method for tire steel wire, relates to the technical field of tire steel wire winding, and comprises a multi-station rotary steel wire winding machine, an adjustable magneto-rheological damper tension control system, an intelligent guide wheel set, a main winding wheel, an integrated vibration isolation winding wheel seat, a 360-degree omnibearing monitoring frame, an intelligent winding cavity and a data processing and control center; the main winding wheel is arranged in the core working area of the multi-station rotary steel wire winding machine; the integrated vibration isolation winding wheel seat; the intelligent guide wheel set and the adjustable magneto-rheological damper tension control system are arranged upstream of the main winding wheel; the 360-degree omnibearing monitoring frame and the intelligent winding cavity are arranged outside the main winding wheel in a surrounding manner; and the data processing and control center communicates data. Through real-time multi-dimensional sensing and predictive analysis, the application establishes active closed-loop control, realizes the transformation from post-failure detection to pre-failure prevention, generates holographic quality data, and comprehensively improves the quality of steel wire winding.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tire steel wire winding, in particular to a winding device and method for tire steel wire. BACKGROUND

[0002] The strength, durability and driving stability of a tire are largely dependent on the quality of its internal framework material, i.e. steel cord. Steel wire winding is one of the core basic processes in tire manufacturing, which aims to tightly wind multiple independent steel wires into a steel wire disc with precise and uniform tension and geometric arrangement for subsequent use in the cord calendering process. The quality of this process directly determines the overall performance of the final tire product. In an ideal winding state, each steel wire should be arranged flatly and orderly under constant tension. Any uneven tension, cross-overlapping or excessive slack will form potential quality problems, which may develop into structural damage under the harsh conditions of high load and high speed, thereby seriously affecting driving safety and shortening the service life of the tire.

[0003] However, the existing steel wire winding monitoring technology generally has significant limitations in ensuring high-quality output. First, its monitoring means has serious hysteresis. The industry generally relies on offline sampling inspection after production or monitors some macro process parameters (such as average tension, motor speed) during production. This method cannot track the quality of the full-length and full-volume steel wire in real time online, and defects are often discovered after the entire steel wire winding is completed, which not only causes a lot of material and time waste, but also allows a large number of potentially problematic products to flow into the next process. Second, the existing technology has obvious detection blind spots. For early and root cause defect signals such as microscopic geometric unevenness between layers and layers, subtle tension differences between multiple parallel steel wires at the sub-Newton level, and microscopic cracks in the steel wire material caused by stress concentration, traditional sensors lack effective detection means and sufficient sensitivity. Finally, the existing technology is essentially a passive defect detection logic rather than an active defect prevention. The goal of the entire monitoring system is to discover physical defects that have already formed, rather than to actively intervene and dynamically correct process parameters at the early stage of defect solidification. These problems collectively result in the inability of the existing production mode to fundamentally eliminate winding quality problems, which restricts the further improvement of the performance consistency of high-end tire products. SUMMARY

[0004] The present application aims to provide a winding device and method for tire steel wire, which solves the problems in the background art.

[0005] To solve the above technical problems, the application provides a winding device for tire steel wire, comprising: a multi-station rotary steel wire winding machine, an adjustable magneto-rheological damper tension control system, an intelligent guide wheel group, a main winding wheel, an integrated vibration isolation winding wheel seat, a 360-degree omnidirectional monitoring frame, an intelligent winding cavity, and a data processing and control center.

[0006] The main winding wheel is arranged in the core working area of the multi-station rotary steel wire winding machine and is used for winding the steel wire.

[0007] The integrated vibration isolation winding wheel seat is used for supporting the main winding wheel, and a high-frequency vibration sensor array is arranged inside the integrated vibration isolation winding wheel seat.

[0008] The intelligent guide wheel group and the adjustable magneto-rheological damper tension control system are arranged upstream of the main winding wheel, wherein the steel wire is sequentially threaded through the adjustable magneto-rheological damper tension control system and the intelligent guide wheel group after being drawn from the source, and is finally wound on the main winding wheel.

[0009] The 360-degree omnidirectional monitoring frame and the intelligent winding cavity are arranged outside the main winding wheel, and together form a monitoring space, wherein the 360-degree omnidirectional monitoring frame and the cavity are provided with a thermal field imaging system, a wideband acoustic emission monitoring system, and a laser displacement sensor group.

[0010] The data processing and control center is in data communication with the high-frequency vibration sensor array, the intelligent guide wheel group, the thermal field imaging system, the wideband acoustic emission monitoring system, and the laser displacement sensor group, and is electrically connected with the adjustable magneto-rheological damper tension control system, so as to form a monitoring-analysis-prediction-control closed loop.

[0011] Preferably, the integrated vibration isolation winding wheel seat is isolated from the interference vibration of the external environment through multi-layer damping materials and a suspension design, wherein the high-frequency vibration sensor array is designed by bionics and is coated with a graphene / piezoelectric ceramic composite nano coating.

[0012] Preferably, the intelligent guide wheel group comprises a plurality of independent guide wheels, wherein each guide wheel is independently provided with a composite MEMS sensor chip, and the chip is used for synchronously measuring the local vibration, temperature, and strain of the steel wire at the contact point of the guide wheel.

[0013] Preferably, the data processing and control center is connected with a distributed edge computing node, wherein the distributed edge computing node is arranged near each sensor group and is used for denoising and preliminary feature extraction of the original data, and high-value feature information is uploaded to the data processing and control center.

[0014] Preferably, a high-speed camera system is also arranged on the 360-degree omnidirectional monitoring framework, which cooperates with the laser displacement sensor group to scan the surface profile of the wound steel wire layer and the spatial position of the incoming steel wire.

[0015] Also provided is a winding method for tire steel wire, comprising:

[0016] System initialization and reference library establishment, using standard process parameters and qualified steel wire for winding, synchronously collecting full set of sensor data, establishing standard winding mode library containing ideal voiceprint fingerprint and thermal field distribution, and initially training winding process digital twin neural network;

[0017] Real-time acquisition and edge preprocessing, starting winding task, all sensors synchronously collecting data; through distributed edge computing nodes deployed on the equipment end, the original signal is denoised, filtered and preliminarily feature extracted;

[0018] Multi-scale fusion analysis and pattern recognition: the preprocessed feature data is sent into the digital twin neural network model in the data center; the digital twin neural network model identifies the spatial pattern of vibration voiceprint spectrum and thermal field distribution, and compares it with the standard winding mode library in real time, while correlating the readings of different sensors in space and time;

[0019] Predictive analysis and defect early warning: based on the fused features, the long short-term memory network layer in the model is used to deduce the future time series data, predict the winding state trend and evaluate the defect probability; when the predicted defect probability exceeds the preset safety threshold, the system starts the early warning;

[0020] Adaptive closed-loop control and optimization: at the same time of the early warning, the optimal tension adjustment strategy is calculated through the reinforcement learning algorithm, and the control instruction is generated and sent to the adjustable magneto-rheological damper tension control system to complete the fine tuning of the tension of the corresponding steel wire.

[0021] Preferably, in the multi-scale fusion analysis and pattern recognition step, the digital twin neural network model is based on the steel wire winding voiceprint resonance theory, and quantifies the inconsistency of tension by identifying the beat frequency phenomenon formed by the vibration interference caused by the inconsistent tension between multiple steel wires.

[0022] Preferably, before the adaptive closed-loop control and optimization step is started, a cross-validation step is further included; wherein the cross-validation step identifies the hot spots generated by friction through the thermal field imaging system, confirms the early warning issued by the vibration voiceprint spectrum, to exclude false positives.

[0023] Preferably, it also includes monitoring of internal damage to the material; wherein high-frequency stress wave signals released by the steel wire due to micro-crack propagation inside the material are captured by a broadband acoustic emission monitoring system, and an alarm is issued at the early stage of irreversible damage to the material.

[0024] Preferably, it also includes a holographic visualization and quality traceability step; wherein the entire winding process is reconstructed in real time as a three-dimensional digital twin model, and a holographic data report containing the quality state of the steel wire is generated after winding is completed.

[0025] Compared with the prior art, the present application has the following beneficial effects:

[0026] 1. By capturing and deeply analyzing the micro-physical signals generated during the winding process, the root cause of winding defects can be identified in advance, and a warning can be issued before physical defects are formed, thereby improving quality control from traditional post-detection to pre-prevention, and fundamentally avoiding the occurrence of quality problems.

[0027] 2. A comprehensive monitoring system with multiple physical dimensions and cross-verification is constructed, not only monitoring traditional parameters, but also fusing various information sources such as vibration, heat distribution, and acoustic emission inside the material to form a comprehensive three-dimensional perception of the winding state. The information sources confirm each other, effectively eliminating the misjudgment that may be caused by single monitoring method, greatly improving the identification accuracy and reliability of the system under complex working conditions, and ensuring comprehensive coverage of various potential problems.

[0028] 3. An adaptive closed-loop control system with fast and precise adjustment is established. Once potential quality deviations are predicted, the optimal process adjustment strategy can be calculated immediately, and the control unit can be driven for fast and accurate fine-tuning intervention. The response speed far exceeds the solidification speed of physical defects, ensuring that the correction action is always completed before the defect is formed, achieving active and real-time optimization of the winding process, and ensuring the high stability of the production process.

[0029] 4. By constructing a digital twin model of the physical process, transparent management of the entire production process and holographic traceability of quality are achieved, which can generate a digital archive containing the quality state of every detail, providing unprecedented detailed data support for quality management throughout the product life cycle. Not only does it provide a reliable basis for downstream processes, but it also accumulates valuable insights for subsequent process continuous improvement and optimization. BRIEF DESCRIPTION OF DRAWINGS

[0030] In order to make the technical solutions in the embodiments of the present application or the prior art clearer, the accompanying drawings needed in the embodiments or prior art description will be briefly introduced. Obviously, the accompanying drawings in the following description only need to be some embodiments of the present application, and all other embodiments obtained by a person of ordinary skill in the art without any creative work on the basis of the embodiments in the present application shall fall within the protection scope of the present application.

[0031] Figure 1 is a schematic diagram of the overall structure of the device outside the device;

[0032] Figure 2 is a schematic diagram of the structure of the intelligent guide wheel group and the main winding wheel;

[0033] Figure 3 is a schematic diagram of the structure of the adjustable magneto-rheological damper tension control system;

[0034] Figure 4 is a schematic diagram of the structure of the 360-degree omnidirectional monitoring frame;

[0035] Figure 5 is a flowchart of the method in the present application.

[0036] 100, multi-station rotary wire winding machine; 101, main winding wheel; 102, integrated vibration isolation winding wheel seat; 200, adjustable magneto-rheological damper tension control system; 300, intelligent guide wheel group; 301, guide wheel; 302, composite MEMS sensing chip; 400, 360-degree omnidirectional monitoring frame; 401, intelligent winding cavity; 402, thermal field imaging system; 403, wideband acoustic emission monitoring system; 404, laser displacement sensor group; 405, high-speed camera system; 5, data processing and control center. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without any creative work shall fall within the protection scope of the present application.

[0038] Please refer to Figures 1-5The application provides a winding device for tire steel wire, which comprises a multi-station rotary steel wire winding machine 100, an adjustable magneto-rheological damper tension control system 200, an intelligent guide wheel set 300, a main winding wheel 101, an integrated vibration isolation winding wheel seat 102, a 360-degree omnidirectional monitoring frame 400, an intelligent winding cavity 401 and a data processing and control center 5; wherein the main winding wheel 101 is arranged in the core working area of the multi-station rotary steel wire winding machine 100 and is used for winding the steel wire; the integrated vibration isolation winding wheel seat 102 is used for supporting the main winding wheel 101, and a high-frequency vibration sensor array is arranged in the integrated vibration isolation winding wheel seat 102; the intelligent guide wheel set 300 and the adjustable magneto-rheological damper tension control system 200 are arranged upstream of the main winding wheel 101; wherein the steel wire is sequentially arranged through the adjustable magneto-rheological damper tension control system 200 and the intelligent guide wheel set 300 after being drawn from the source and is finally wound on the main winding wheel 101; the 360-degree omnidirectional monitoring frame 400 and the intelligent winding cavity 401 are arranged outside the main winding wheel 101 and jointly form a monitoring space; wherein the 360-degree omnidirectional monitoring frame 400 and the cavity are provided with a thermal field imaging system 402, a wideband acoustic emission monitoring system 403 and a laser displacement sensor group 404; the data processing and control center 5 is in data communication with the high-frequency vibration sensor array, the intelligent guide wheel set 300, the thermal field imaging system 402, the wideband acoustic emission monitoring system 403 and the laser displacement sensor group 404 respectively and is electrically connected with the adjustable magneto-rheological damper tension control system 200 to form a closed loop of monitoring-analysis-prediction-control.

[0039] The embodiment provides a winding device for tire steel wire; the winding monitoring method in the prior art has the technical problems of monitoring lag, blind area for detecting micro defects and passive intervention; the integrated multi-physical quantity sensing and control system solves the above problems;

[0040] The device takes a multi-station rotary steel wire winding machine 100 as the main body; on the running path of the steel wire, an adjustable magneto-rheological damper tension control system 200 and an intelligent guide wheel set 300 are sequentially arranged from the source; the steel wire is finally wound on the main winding wheel 101 after passing through the above components; the main winding wheel 101 is arranged in the core working area of the winding machine and is supported by an integrated vibration isolation winding wheel seat 102, and a high-frequency vibration sensor array is arranged in the wheel seat and is used for collecting vibration signals in the winding process;

[0041] Outside the main winding wheel 101, a 360-degree omnidirectional monitoring frame 400 is deployed around it, which together with the intelligent winding cavity 401 forms a three-dimensional monitoring space, and inside it is provided with a thermal field imaging system 402, a wideband acoustic emission monitoring system 403 and a laser displacement sensor group 404 for monitoring the winding state from multiple physical dimensions;

[0042] All sensors, including the high-frequency vibration sensor array, the sensors in the intelligent guide wheel group 300, the thermal field imaging system 402, the wideband acoustic emission monitoring system 403 and the laser displacement sensor group 404, are in data communication with a data processing and control center 5; the data processing and control center 5 is an industrial computer responsible for analyzing all incoming data; the center is also electrically connected with the adjustable magneto-rheological damper tension control system 200; in this way, the system builds a complete monitoring-analysis-prediction-control operation logic, which can predict defect trends based on sensor data and actively adjust tension to eliminate defects before they form.

[0043] The integrated vibration isolation winding wheel seat 102 isolates external environmental interference vibrations through multiple layers of damping materials and a suspension design; the high-frequency vibration sensor array is designed by bionics and coated with a graphene / piezoelectric ceramic composite nano coating;

[0044] This embodiment further limits the integrated vibration isolation winding wheel seat 102 and the high-frequency vibration sensor array inside it; to solve the technical problem of signal interference caused by environmental vibrations (such as motor and rack vibrations) on high-sensitivity measurements, the integrated vibration isolation winding wheel seat 102 adopts a special structural design; specifically, its internal structure contains multiple layers of damping materials with different mechanical properties, combined with a suspension design; this design can effectively isolate more than 99% of external environmental interference vibrations, ensuring that the signals collected by the sensor array can accurately reflect the microscopic vibrations generated by the steel wire winding process itself;

[0045] To solve the technical problem of insufficient sensitivity in detecting weak vibration signals, the high-frequency vibration sensor array itself has also been optimized; the array adopts a bionic design, with a layout that mimics the structure of biological vibration perception; in addition, a layer of graphene / piezoelectric ceramic composite nano coating is applied to the sensing surface of the sensor; this coating material has excellent piezoelectric effect and mechanical properties, and its application has improved the detection sensitivity of the sensor by three orders of magnitude, enabling the system to capture early vibration precursors caused by small imbalances in tension.

[0046] The intelligent guide wheel group 300 comprises a plurality of independent guide wheels 301; wherein each guide wheel 301 is independently equipped with a composite MEMS sensing chip 302, and the composite MEMS sensing chip 302 is used for synchronously measuring the local vibration, temperature and strain of the steel wire at the contact point of the guide wheel 301;

[0047] The embodiment further limits the composition of the intelligent guide wheel group 300; in order to solve the technical problem that fault tracing is difficult and the abnormal source cannot be accurately positioned, the intelligent guide wheel group 300 is composed of a plurality of guide wheels 301 which are independent in structure and function; these guide wheels 301 are arranged along the path of the steel wire before entering the main winding wheel 101;

[0048] Each independent guide wheel 301 is equipped with a composite micro-electro-mechanical system (MEMS) sensing chip; here, the composite means that the single chip integrates multiple sensing functions and can synchronously measure three key physical quantities of the steel wire at the contact point of the guide wheel 301: local vibration, local temperature and local strain; when the data processing and control center 5 monitors the systematic tension abnormality, the data of the sensing chip of each independent guide wheel 301 along the way can be compared and analyzed to determine whether the abnormal state is caused by a specific guide wheel 301 (such as bearing failure) or has existed before entering the intelligent guide wheel group 300; this distributed sensing arrangement significantly enhances the spatial resolution capability and maintainability of the system.

[0049] The data processing and control center 5 is connected with a distributed edge computing node; wherein the distributed edge computing node is deployed near each sensor group, used for noise reduction and preliminary feature extraction of the original data, and uploading high-value feature information to the data processing and control center 5;

[0050] The embodiment further limits the data processing architecture; in order to solve the technical problem that the direct transmission of massive original sensor data to the central processing unit may cause data congestion, network delay, and further affect the real-time performance of the control system, the device adds a distributed edge computing node between the data processing and control center 5 and each sensor group;

[0051] The distributed edge computing nodes are some small computing units, which are physically deployed near each sensor group (such as a high-frequency vibration sensor array, a smart guide wheel group 300, etc.); the functions of these nodes are to locally pre-process the massive raw data generated by the sensors connected thereto; the pre-processing process mainly includes two tasks: one is to filter out irrelevant components in the signal through a digital filtering algorithm; the other is to perform preliminary feature extraction on the denoised signal, such as calculating the frequency spectrum and time-domain statistical features of the vibration signal; after completing the pre-processing, only the extracted feature data containing high-value state information is uploaded to the data processing and control center 5; this architecture significantly reduces the requirements for the central processor and data transmission bandwidth, ensuring the millisecond-level response capability of the entire closed-loop control system.

[0052] The 360-degree omnidirectional monitoring framework 400 is also provided with a high-speed camera system 405, which cooperates with the laser displacement sensor group 404 to scan the surface profile of the wound steel wire layer and the spatial position of the incoming steel wire;

[0053] The embodiment enhances the function of the 360-degree omnidirectional monitoring framework 400; in order to solve the errors that may exist in a single measurement method and provide high-precision geometric basis for the training of an artificial intelligence model, a high-speed camera system 405 is additionally provided on the 360-degree omnidirectional monitoring framework 400;

[0054] The high-speed camera system 405 cooperates with the original laser displacement sensor group 404; the laser displacement sensor group 404 continuously scans the steel wire layer that has been wound on the main winding wheel 101 point by point with its 0.01 mm measurement accuracy to obtain its accurate surface profile three-dimensional data; at the same time, the high-speed camera system 405 is responsible for capturing the dynamic spatial position and posture of the steel wire entering the main winding area; the two cooperate to not only provide final geometric evidence about the winding quality for confirming whether there are physical defects such as crossing and indentation, but more importantly, the high-precision ground true data generated thereby is used to continuously train and calibrate the digital twin model in the data processing and control center 5, ensuring the accuracy and reliability of the model prediction and avoiding the risk of incorrect control instructions due to the model deviating from the physical reality.

[0055] Embodiment 2:

[0056] Please refer to Figure 5The application also provides a winding method for tire steel wire, comprising: system initialization and reference library establishment, winding with standard process parameters and qualified steel wire, synchronous acquisition of full set of sensor data, establishment of standard winding mode library containing ideal voiceprint fingerprint and thermal field distribution, and initial training of winding process digital twin neural network; real-time acquisition and edge preprocessing, starting of winding task, synchronous data acquisition of all sensors; noise reduction, filtering and preliminary feature extraction of original signals by distributed edge computing nodes deployed on equipment end; multi-scale fusion analysis and pattern recognition: feeding the preprocessed feature data into the digital twin neural network model of the data center; the digital twin neural network model identifies the spatial pattern of vibration voiceprint spectrum and thermal field distribution, and compares it with the standard winding mode library in real time, while correlating the readings of different sensors in space and time; predictive analysis and defect early warning, based on the fused features, the long short-term memory network layer in the model is used to deduce the future time series data, predict the winding state trend and evaluate the defect probability; when the predicted defect probability exceeds the preset safety threshold, the system starts the early warning; adaptive closed-loop control and optimization, at the same time of the early warning, the optimal tension adjustment strategy is calculated through the reinforcement learning algorithm, and the control instruction is generated and sent to the adjustable magneto-rheological damper tension control system 200 to complete the fine tuning of the tension of the corresponding steel wire.

[0057] The embodiment provides a tire steel wire winding method applied to any of the foregoing devices, which converts the hardware capability of the device into actual effect of preventing defects.

[0058] The system initialization and reference library establishment step is performed; before formal production, the verified qualified steel wire is wound under the standard process parameters; in this process, the full set of sensors (vibration, thermal field, acoustic emission, position, etc.) on the device synchronously acquire data; these data are used to establish a standard winding mode library, which stores data patterns such as vibration voiceprint fingerprint and thermal field spatial distribution under ideal winding state; at the same time, these reference data are also used to initially train the winding process digital twin neural network in the data processing and control center 5.

[0059] The real-time acquisition and edge preprocessing step is entered; the winding task is formally started, and all sensors synchronously acquire real-time data at a preset frequency (for example, 1 kHz); the distributed edge computing nodes deployed at various places of the equipment perform real-time noise reduction, filtering and preliminary feature extraction on the acquired original signals;

[0060] performing multi-scale fusion analysis and pattern recognition steps; the feature data extracted by the distributed edge computing nodes is transmitted to the digital twin neural network model in the data processing and control center 5; the convolutional neural network layer in the model performs spatial pattern recognition on the input vibration sonogram and thermal field distribution map, and compares it with the reference patterns in the standard winding pattern library in real time to identify deviations; at the same time, the spatio-temporal attention mechanism of the model is responsible for correlating readings from different sensors at different times and spatial positions to construct a complete state causal chain;

[0061] performing predictive analysis and defect warning steps; based on the multi-dimensional features after the previous fusion analysis, the long short-term memory network layer in the model deduces the time series of the data, so as to predict the future trend of the winding state up to 60 seconds in advance and assess the probability of defects such as crossing and relaxation in real time; when the predicted probability of any defect exceeds the preset safety threshold, the system immediately starts the warning mechanism; the preset safety threshold is a critical probability value determined by statistical analysis based on historical production data and experimental results; specifically, a large number of sensor data corresponding to normal winding samples and known defect samples are collected, and a machine learning model (such as logistic regression or support vector machine) is trained to find a probability point that can maximize the distinction between normal and abnormal states, while balancing the false negative rate and false positive rate; for example, the threshold can be set to when the model predicts that the probability of wire crossing in the next 60 seconds exceeds 5%, the system starts the warning; the threshold can also be adjusted online by technical personnel according to product quality level requirements;

[0062] performing adaptive closed-loop control and optimization steps; at the same time when the warning is triggered, a reinforcement learning algorithm in the control system calculates the optimal tension adjustment strategy for correcting the potential defect in real time according to the current system state and the predicted defect type; the strategy is converted into specific control instructions and sent to the adjustable magneto-rheological damper tension control system 200; after receiving the instructions, the target steel wire tension is precisely adjusted within 10 milliseconds; this series of steps constitutes a complete closed loop from prediction to execution, which can eliminate the root cause of physical defects (such as tension imbalance) before they actually form;

[0063] Specifically, the reinforcement learning algorithm model is constructed as follows:

[0064] State space: defined as a vector composed of pre-processed feature data of all sensors at time t, including but not limited to the vibration beat amplitude, local temperature, strain, thermal field distribution characteristics of each steel wire, and the geometric profile deviation value obtained by laser scanning;

[0065] Action space: defined as the amount of voltage or current adjustment applied to each adjustable magneto-rheological damper tension control system 200, which has a corresponding relationship with the amount of tension change, the action can be discrete (such as increase / decrease / keep) or continuous (accurate adjustment value within a certain range);

[0066] Reward function: designed to guide model learning; when the system successfully avoids a predicted defect once (i.e. after the alarm is issued, the subsequent sensor data shows that the state returns to normal), a positive reward is given; when the defect still occurs after the alarm, or the system makes unnecessary adjustments (cross-validation fails), a negative reward is given; the size of the reward is related to the severity of the deviation from the standard state and the accuracy of the adjustment; through continuous iterative training, the algorithm model can learn to take what kind of tension adjustment action under the given system state can obtain the maximum long-term cumulative reward, i.e. the optimal tension adjustment strategy.

[0067] In the multi-scale fusion analysis and pattern recognition step, the digital twin neural network model quantifies the inconsistency of tension by identifying the beat frequency phenomenon formed by the vibration interference between multiple steel wires with inconsistent tension based on the steel wire winding voiceprint resonance theory;

[0068] This embodiment limits a specific technical means in the multi-scale fusion analysis and pattern recognition step; in order to solve the technical problem that the small tension difference between multiple steel wires cannot be quantitatively detected, in this step, the digital twin neural network model applies the steel wire winding voiceprint resonance theory for analysis;

[0069] This theory points out that, in an ideal state, a steel wire moving with a certain tension will produce a stable, inherent vibration frequency spectrum as a reference, i.e. a voiceprint fingerprint; when there is a tension inconsistency between multiple parallel wound steel wires, their respective vibrations will interfere, thereby producing a specific beat frequency phenomenon in the superimposed vibration signal; the model identifies this beat frequency phenomenon through frequency spectrum analysis of the signals collected by the high-frequency vibration sensor array; the theoretical basis is that if the main vibration frequencies of two adjacent steel wires are and , then a beat frequency signal with a frequency of will be produced in the combined vibration; by calculating the frequency and amplitude of the beat frequency, the model can accurately quantify the inconsistency of tension between the steel wires; this quantification result is a key input parameter for subsequent prediction of defect probability and accurate tension compensation.

[0070] Before the start of the adaptive closed-loop control and optimization step, a cross-validation step is further included; wherein, the cross-validation step identifies hot spots generated by friction through the thermal field imaging system 402, and confirms the pre-warning issued by the vibration voiceprint spectrum to exclude false positives;

[0071] The embodiment adds a cross-validation step to address the technical problem that a single sensing system may generate false positives, leading to unnecessary downtime or incorrect adjustments; the cross-validation step is set after the predictive analysis and defect warning step and before the adaptive closed-loop control and optimization step is started;

[0072] When the predictive system based on vibration signal analysis issues a warning (for example, predicting that the steel wire may cross due to uneven tension), the control system does not immediately perform tension adjustment; instead, it first starts the cross-validation step; in this step, the system calls data from the thermal field imaging system 402 to analyze the thermal field distribution in the area near the warning location; abnormal friction of the steel wire (such as the location where the cross is about to occur) will cause local temperature rise and form a hot spot; if the thermal field imaging system 402 also identifies a hot spot at the corresponding location of the warning that conforms to the physical law, the warning is confirmed as valid; only after this multi-physical quantity cross-validation is confirmed, the adaptive closed-loop control and optimization step will be started; this can effectively eliminate false positives caused by accidental vibration interference and other factors, improving the reliability of the entire control system.

[0073] It also includes monitoring of internal damage to the material; the high-frequency stress wave signals released by the steel wire material due to micro-crack propagation are captured by the wideband acoustic emission monitoring system 403, and an alarm is issued at the early stage of irreversible damage to the material;

[0074] The embodiment adds a parallel monitoring dimension to address the technical problem that conventional methods cannot detect damage at the internal germination stage of the material; this step is implemented through the wideband acoustic emission monitoring system 403 set in the intelligent winding cavity 401;

[0075] Acoustic emission monitoring is a passive monitoring technology whose principle is to monitor the stress waves released when micro-damage occurs in the material; during the steel wire winding process, if the steel wire material itself has metallurgical defects or micro-cracks are generated under tension, the germination and expansion of these micro-cracks will release energy in the form of high-frequency stress waves; the wideband acoustic emission monitoring system 403 is specifically designed to capture these high-frequency, low-energy stress wave signals; when the system captures acoustic emission signals that meet the characteristics of micro-crack propagation, it will immediately issue a material damage alarm independent of the defect warning; this monitoring allows the system to detect hidden dangers at the early stage before the material breaks or other macro-irreversible damage occurs, which plays a decisive role in reducing the final steel wire breakage rate, and statistics show that it can reduce the breakage rate by 85%.

[0076] It also includes a holographic visualization and quality traceability step; the entire winding process is reconstructed in real time as a three-dimensional digital twin model, and after winding is completed, a holographic data report containing the quality state of the steel wire is generated;

[0077] The embodiment adds a step throughout the whole process and extends to the step after the process ends; in order to solve the technical problems of opaque winding process and inability to trace the quality data throughout the whole process, the method includes holographic visualization and quality traceability;

[0078] During the whole winding process, the data processing and control center 5 uses the data input by all sensors (including geometric data of laser scanning, vibration data, thermal field data, etc.) to reconstruct a three-dimensional digital twin model corresponding to the physical device in real time; the model is visualized on the operation interface, and the operator can observe the real-time winding state of the steel wire from any angle; any warning or alarm issued by the system will be marked on the accurate position (with an accuracy of 0.1 mm) of the three-dimensional model in a highlighted form;

[0079] After the whole disc of steel wire is wound, the system automatically performs data summarization and generates a holographic data report; the report not only contains the conventional production parameters of the disc of steel wire, but more importantly, it records the key quality state data (such as tension fluctuation, vibration spectrum, temperature change, etc.) of each section of steel wire in the winding process and all potential defect events that have been marked in meters; the report provides detailed quality input basis for subsequent vulcanization and other processes, and realizes quality traceability throughout the product life cycle.

[0080] Compared with the prior art, the following beneficial effects are presented;

[0081] The fundamental change from defect detection to defect prediction is realized, and the forward-looking of quality control is significantly improved;

[0082] The prior art generally relies on offline sampling inspection or macro parameter monitoring, which is essentially passive identification after defect formation; the integrated vibration isolation winding wheel seat 102 in the present scheme, with its multi-layer damping material and suspension design, provides a super-low noise measurement environment for the internal high-frequency vibration sensor array; this environment ensures that the sensor array can capture pure and microscopic vibration signals generated by the steel wire winding process itself; the key is that the data processing and control center 5 can identify the vibration interference caused by the inconsistency of sub-Newton level tension between multiple steel wires, i.e. the beat frequency phenomenon, based on these signals; by quantifying this phenomenon, the system can accurately predict the trend of physical defects such as crossing or indentation before they are formed for tens of seconds; this fundamentally solves the detection blind area problem of the prior art for the core defect precursor of tension imbalance;

[0083] A monitoring system for cross verification of multiple physical quantities is constructed, which greatly improves the accuracy of early warning and the dimension of defect identification;

[0084] The prior art often relies on a single information source, which is prone to misjudgment. The present scheme cooperates with multiple sets of sensing systems in the 360-degree omnidirectional monitoring framework 400 and the intelligent winding cavity 401. When the high-frequency vibration sensor array issues a warning, the system will immediately call the data of the thermal field imaging system 402 to cross-verify whether there is a hotspot caused by abnormal friction at the warning position, effectively eliminating false positives caused by accidental interference. At the same time, the broadband acoustic emission monitoring system 403 captures the high-frequency stress waves released by the micro-crack expansion inside the steel wire material, achieving early warning of irreversible damage inside the material, which is a dimension that traditional monitoring methods cannot reach. The laser displacement sensor group 404 and the high-speed camera system 405 together provide high-precision geometric profile and spatial position information, providing a data foundation for the final confirmation of the defect and the accuracy calibration of the digital twin model.

[0085] An active and adaptive closed-loop control circuit is established to realize real-time intervention and optimization of the winding process.

[0086] The intervention measures of the prior art are passive, usually involving manual intervention after the problem is found, which has serious lag. The operation logic of the present scheme is completely different. After the data processing and control center 5 confirms the problem through predictive analysis and defect warning, it will immediately calculate the optimal tension adjustment strategy through reinforcement learning algorithm. The strategy is in the form of control instructions and is sent to the adjustable magneto-rheological damper tension control system 200. This system uses the physical properties of magneto-rheological fluid to accurately adjust the tension of the target steel wire within 10 milliseconds. This complete closed loop from monitoring, analysis, prediction to control has a response speed much faster than the formation speed of physical defects, ensuring that the correction action is always completed before the defect is solidified, and changing the quality control from passive response to active avoidance.

[0087] Holographic tracing of the production process is realized, providing data support for quality management and process optimization.

[0088] The quality records of the prior art are usually limited to batches and cannot be traced back to the quality status of specific sections in a single steel wire. The present scheme reconstructs the entire winding process in real time into a three-dimensional digital twin model through holographic visualization and quality tracing steps. After winding is completed, the system generates a holographic data report containing the quality status of each meter of steel wire, detailing key indicators such as tension fluctuations and vibration spectrum. This not only provides detailed incoming quality basis for downstream processes, but also accumulates valuable historical data for long-term process improvement and equipment maintenance, fundamentally improving the stability and consistency of product quality.

[0089] The above merely describes the preferred embodiments of the present application, and is not intended to limit the present application in other forms. Any skilled person in the art can modify or change the disclosed technical content into equivalent embodiments with equivalent changes, and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application, without departing from the technical solution content of the present application, still falls within the protection scope of the present application.

Claims

1. A winding device for steel cords of tyres, characterized in that, The application relates to a multi-station rotary steel wire winding machine (100), an adjustable magnetorheological damper tension control system (200), an intelligent guide wheel group (300), a main winding wheel (101), an integrated vibration isolation winding wheel seat (102), a 360-degree omnibearing monitoring frame (400), an intelligent winding cavity (401) and a data processing and control center (5). The main winding wheel (101) is arranged in a core working area of the multi-station rotary steel wire winding machine (100) and is used for winding a steel wire. The integrated vibration isolation winding wheel seat (102) is used for supporting the main winding wheel (101), and a high-frequency vibration sensor array is arranged in the integrated vibration isolation winding wheel seat (102). The intelligent guide wheel group (300) and the adjustable magnetorheological damper tension control system (200) are arranged upstream of the main winding wheel (101); the steel wire passes through the adjustable magnetorheological damper tension control system (200) and the intelligent guide wheel group (300) in sequence after being drawn from a source and is finally wound on the main winding wheel (101). The 360-degree omnibearing monitoring frame (400) and the intelligent winding cavity (401) are arranged outside the main winding wheel (101) and jointly form a monitoring space; the 360-degree omnibearing monitoring frame (400) and the cavity are provided with a thermal field imaging system (402), a wideband acoustic emission monitoring system (403) and a laser displacement sensor group (404). The data processing and control center (5) is in data communication with the high-frequency vibration sensor array, the intelligent guide wheel group (300), the thermal field imaging system (402), the wideband acoustic emission monitoring system (403) and the laser displacement sensor group (404) and is electrically connected with the adjustable magnetorheological damper tension control system (200) to form a monitoring-analysis-prediction-control closed loop. The integrated vibration isolation winding wheel seat (102) is isolated from external interference vibration through multi-layer damping materials and a suspension design; the high-frequency vibration sensor array is designed by bionics and is coated with a graphene / piezoelectric ceramic composite nano coating.

2. A device for winding a steel wire for a tire according to claim 1, characterized in that, The intelligent guide wheel group (300) comprises a plurality of independent guide wheels (301); each guide wheel (301) is independently provided with a composite MEMS sensing chip (302), and the composite MEMS sensing chip (302) is used for synchronously measuring local vibration, temperature and strain of the steel wire at a contact point of the guide wheel (301).

3. A device for winding a steel wire for a tire according to claim 1, characterized in that, The data processing and control center (5) is connected with a distributed edge computing node; the distributed edge computing node is arranged near each sensor group and is used for denoising and preliminarily extracting features of original data and uploading high-value feature information to the data processing and control center (5).

4. A device for winding a steel wire for a tire according to claim 1, wherein ​ 5. A device for winding a steel wire for a tire according to claim 1, wherein The 360-degree omnidirectional monitoring framework (400) is also provided with a high-speed camera system (405) cooperating with the laser displacement sensor group (404) for scanning the surface profile of the wound steel wire layer and the spatial position of the incoming steel wire.

6. A method for winding a steel cord for a tire, applied to a device for winding a steel cord for a tire according to any one of claims 1 to 5, characterized in that, Comprise: System initialization and reference library establishment, using standard process parameters and qualified steel wire for winding, synchronously collecting full set of sensor data, establishing standard winding mode library containing ideal voiceprint fingerprint and thermal field distribution, and initially training winding process digital twin neural network; Real-time acquisition and edge preprocessing, starting winding task, all sensors synchronously collecting data; through the distributed edge computing node deployed on the equipment end, the original signal is denoised, filtered and preliminarily feature extracted; Multi-scale fusion analysis and pattern recognition: the preprocessed feature data is sent into the digital twin neural network model in the data center; the digital twin neural network model identifies the spatial pattern of vibration voiceprint spectrum and thermal field distribution, and compares it with the standard winding mode library in real time, while correlating the readings of different sensors in space and time; Predictive analysis and defect early warning, based on the fused features, the long short-term memory network layer in the model is used to deduce the future time series data, predict the winding state trend and evaluate the defect probability; when the predicted defect probability exceeds the preset safety threshold, the system starts the early warning; Self-adaptive closed-loop control and optimization, at the same time of the early warning, the optimal tension adjustment strategy is calculated through the reinforcement learning algorithm, and the control instruction is generated and sent to the adjustable magneto-rheological damper tension control system (200) to complete the fine adjustment of the tension of the corresponding steel wire.

7. A method for winding a steel wire for a tire according to claim 6, wherein In the multi-scale fusion analysis and pattern recognition step, the digital twin neural network model is based on the steel wire winding voiceprint resonance theory, which quantifies the inconsistency of tension by identifying the beat frequency phenomenon formed by the vibration interference caused by the inconsistency of tension between multiple steel wires.

8. A method for winding a steel wire for a tire according to claim 6, wherein Before the self-adaptive closed-loop control and optimization step is started, it also includes a cross-validation step; wherein, the cross-validation step identifies the hot spots generated by friction through the thermal field imaging system (402), confirms the early warning issued by the vibration voiceprint spectrum identification, to exclude false positives.

9. A method for winding a steel wire for a tire according to claim 6, wherein It also includes monitoring of internal damage of materials; wherein, the high-frequency stress wave signals released by the micro-crack propagation in the internal material of the steel wire are captured by the wideband acoustic emission monitoring system (403), and an alarm is issued at the early stage of irreversible damage of the material.

10. A method for winding a steel wire for a tire according to claim 6, wherein It also includes a holographic visualization and quality traceability step; wherein, the whole winding process is reconstructed into a three-dimensional digital twin model in real time, and after the winding is completed, a holographic data report containing the quality state of the steel wire is generated.

Images