Winding device and winding method for tire steel wire

Abstract

The invention discloses a winding device and method for a tire steel wire, and relates to the technical field of tire steel wire winding. Comprising a multi-station rotary steel wire winding machine, an adjustable magnetorheological 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 a core working area of the multi-station rotary steel wire winding machine; an integrated vibration isolation winding wheel seat; the intelligent guide wheel set and the tension control system are arranged on the 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 processor is in data communication with the control center. According to the method, active closed-loop control is established through real-time multi-dimensional perception and predictive analysis, transformation from post-defect detection to pre-defect prevention is achieved, holographic quality data is generated, and the steel wire winding quality is comprehensively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of tire wire winding, and in particular to a winding device and a winding method for tire wire. Background Art

[0002] A tire's strength, durability, and driving stability depend largely on the quality of its internal skeleton material, the steel cord. Wire winding is a core and fundamental process in tire manufacturing. Its purpose is to tightly wind multiple independent strands of steel wire with precise, uniform tension and geometric arrangement into a wire coil for use in the subsequent cord calendering process. The quality of this process directly determines the overall performance of the final tire. Under ideal winding conditions, each steel wire should be arranged flatly and orderly under constant tension. Any uneven tension, overlapping, or excessive slack will pose potential quality risks. These risks can develop into structural damage when the tire is subjected to the harsh conditions of high loads and high speeds, seriously affecting driving safety and shortening the tire's service life.

[0003] However, existing wire winding monitoring technologies generally suffer from significant limitations in ensuring high-quality output. First, their monitoring methods suffer from significant lags. The industry generally relies on offline spot checks after production is completed, or on monitoring macro-process parameters (such as average tension and motor speed) during production. This approach fails to provide real-time online quality tracking of the entire length and volume of wire. Defects are often not discovered until after the entire coil of wire has been wound, resulting in significant waste of materials and labor time and potentially dangerous products being diverted to the next process. Second, existing technologies suffer from significant detection blind spots. Traditional sensors lack the effective means and sufficient sensitivity to detect early, root-cause defects, such as microscopic geometric inhomogeneities between winding layers, subtle sub-Newton tension differences between multiple parallel strands of wire, and microcracks that initiate due to stress concentration within the wire material. Finally, existing technologies are essentially passive defect detection logic, rather than proactive defect prevention. The entire monitoring system aims to detect existing physical defects, but cannot proactively intervene in and dynamically correct process parameters before the defects solidify. These problems together make it difficult for the existing production model to fundamentally eliminate winding quality problems, restricting the further improvement of the performance consistency of high-end tire products. Summary of the Invention

[0004] The purpose of the present invention is to provide a winding device and a winding method for tire wire, which solve the problems existing in the background technology.

[0005] To solve the above technical problems, the present invention provides a winding device for tire steel wire, comprising: a multi-station rotary steel wire winding machine, an adjustable magnetorheological damper tension control system, an intelligent guide wheel set, 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; Wherein, 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; The integrated vibration isolation winding wheel seat is used to support the main winding wheel, and a high-frequency vibration sensor array is installed inside the integrated vibration isolation winding wheel seat; The intelligent guide wheel group and the tension control system are arranged upstream of the main winding wheel; wherein, after the steel wire is drawn out from the source, it passes through the tension control system and the intelligent guide wheel group in sequence, and is finally wound around the main winding wheel; The 360-degree omnidirectional monitoring frame and the intelligent winding cavity are deployed around the outside of the main winding wheel, together forming a monitoring space; wherein, the 360-degree omnidirectional monitoring frame and the cavity are equipped with a thermal field imaging system, a broadband acoustic emission monitoring system and a laser displacement sensor group; The data processing and control center respectively communicates data with the high-frequency vibration sensor array, the intelligent guide wheel group, the thermal field imaging system, the broadband acoustic emission monitoring system and the laser displacement sensor group, and is electrically connected to the tension control system to form a closed loop of monitoring-analysis-prediction-control.

[0006] Preferably, the integrated vibration isolation winding wheel seat is isolated from interfering vibrations of the external environment by means of multi-layer damping materials and a suspension design; wherein, the high-frequency vibration sensor array is designed by bionics and covered with a graphene / piezoelectric ceramic composite nano-coating.

[0007] Preferably, the intelligent guide wheel group includes a plurality of independent guide wheels; wherein each of the guide wheels is independently equipped with a composite MEMS sensor chip, and the chip is used to synchronously measure the local vibration, temperature and strain of the steel wire at the contact point of the guide wheel.

[0008] Preferably, the data processing and control center is connected to a distributed edge computing node; wherein the distributed edge computing node is deployed near each sensor group, and is used to perform noise reduction and preliminary feature extraction on the raw data, and upload high-value feature information to the data processing and control center.

[0009] Preferably, the 360-degree all-round monitoring frame is also provided with a high-speed camera system, which cooperates with the laser displacement sensor group to scan the surface contour of the wound steel wire layer and the spatial position of the entering steel wire.

[0010] A method for winding tire wire is also provided, comprising: System initialization and benchmark library establishment, using standard process parameters and qualified steel wire for winding, synchronously collecting a full set of sensor data, establishing a standard winding pattern library containing ideal voice fingerprints and thermal field distribution, and performing initial training of the winding process digital twin neural network; Real-time data collection and edge preprocessing are performed, and the wrapping task is started. All sensors collect data synchronously. Distributed edge computing nodes deployed on the device side are used to reduce noise, filter, and perform preliminary feature extraction on the original signal. Multi-scale fusion analysis and pattern recognition: The pre-processed feature data is fed into the digital twin neural network model in the data center. The digital twin neural network model identifies the spatial patterns of the vibration soundprint and thermal field distribution, compares them with the standard winding pattern library in real time, and correlates the readings of different sensors in time and space. Predictive analysis and defect warning: Based on the fused features, the model's long-short-term memory network layer extrapolates future time series data to predict winding state trends and assess defect probability. When the predicted defect probability exceeds the preset safety threshold, the system initiates an alert. Adaptive closed-loop control and optimization. When the early warning is issued, the optimal tension adjustment strategy is calculated through the reinforcement learning algorithm, and the control instructions are generated and sent to the adjustable magnetorheological damper tension control system to complete the tension fine-tuning of the corresponding steel wire.

[0011] Preferably, in the multi-scale fusion analysis and pattern recognition step, the digital twin neural network model is based on the wire winding acoustic resonance theory, and quantifies the tension inconsistency by identifying the beat frequency phenomenon formed by the vibration interference caused by the inconsistent tension between multiple strands of steel wires.

[0012] Preferably, before the adaptive closed-loop control and optimization step is started, a cross-validation step is also included; wherein, the cross-validation step identifies the hot spots generated by friction through a thermal field imaging system, and confirms the warning issued by the vibration soundprint pattern recognition to eliminate false alarms.

[0013] Preferably, it also includes monitoring of internal damage of the material; wherein, the high-frequency stress wave signal released by the expansion of microcracks inside the steel wire material is captured by a broadband acoustic emission monitoring system, and an alarm is issued at the early stage of irreversible damage to the material.

[0014] Preferably, it also includes holographic visualization and quality traceability steps; wherein the entire 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 status of the steel wire is generated.

[0015] Compared with existing technologies, the present invention has the following beneficial effects: 1. By capturing and analyzing the microscopic physical signals generated during the winding process with high precision, the root causes of winding defects can be identified in advance, and early warnings can be issued before physical defects form, thereby upgrading quality control from traditional post-detection to pre-prevention, fundamentally avoiding the occurrence of quality risks.

[0016] 2. A comprehensive monitoring system with multiple physical dimensions and cross-validation has been constructed. It not only monitors traditional parameters, but also integrates multiple information sources such as vibration, heat distribution, and internal acoustic emission of materials to form a full-dimensional three-dimensional perception of the winding state. The various information sources verify each other, effectively eliminating the misjudgment that may be caused by a single monitoring method, greatly improving the system's recognition accuracy and reliability of complex working conditions, and ensuring comprehensive coverage of various potential problems.

[0017] 3. An adaptive closed-loop control system with rapid response and precise adjustment has been established. Once a potential quality deviation is predicted, it can immediately calculate the optimal process adjustment strategy and drive the control unit to perform fast and precise fine-tuning intervention. The response speed far exceeds the solidification speed of physical defects, ensuring that the corrective action is always completed before the defect is formed, realizing active and real-time optimization of the winding process and ensuring the high stability of the production process.

[0018] 4. By building a digital twin model of the physical process, transparent management of the entire production process and holographic traceability of quality are achieved. Digital archives containing the quality status of every detail can be generated, providing unprecedented detailed data support for quality management throughout the product life cycle. This not only provides a reliable basis for downstream processes, but also accumulates valuable insights for subsequent continuous process improvement and optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention, and those skilled in the art can derive other drawings based on these drawings without inventive effort. Figure 1 It is a schematic diagram of the overall structure of the device exterior; Figure 2 It is a structural diagram of the intelligent guide wheel group and the main winding wheel; Figure 3 It is a structural diagram of an adjustable magnetorheological damper tension control system; Figure 4 It is a structural diagram of the 360-degree all-round monitoring framework; Figure 5 is a flow chart of the method of the present invention.

[0020] 100. Multi-station rotary wire winding machine; 101. Main winding wheel; 102. Integrated vibration-isolated winding wheel seat; 200. Tension control system; 300. Intelligent guide wheel group; 301. Guide wheel; 302. Composite MEMS sensor chip; 400. 360-degree all-round monitoring frame; 401. Intelligent winding cavity; 402. Thermal field imaging system; 403. Broadband acoustic emission monitoring system; 404. Laser displacement sensor group; 405. High-speed camera system; 5. Data processing and control center. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0022] See also Figure 1-5The present invention provides a winding device for tire steel wire, comprising: 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 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 for winding steel wire; the integrated vibration isolation winding wheel seat 102 is used to support the main winding wheel 101, and a high-frequency vibration sensor array is installed inside the integrated vibration isolation winding wheel seat 102; the intelligent guide wheel group 300 and the tension control system 200 are arranged upstream of the main winding wheel 101; wherein, the steel wire is wound from the source After being led out, it passes through the tension control system 200 and the intelligent guide wheel group 300 in sequence, and is finally wound around the main winding wheel 101; the 360-degree omnidirectional monitoring frame 400 and the intelligent winding cavity 401 are deployed around the outside of the main winding wheel 101, together forming a monitoring space; wherein, the 360-degree omnidirectional monitoring frame 400 and the cavity are provided with a thermal field imaging system 402, a broadband acoustic emission monitoring system 403 and a laser displacement sensor group 404; the data processing and control center 5 respectively communicates data with the high-frequency vibration sensor array, the intelligent guide wheel group 300, the thermal field imaging system 402, the broadband acoustic emission monitoring system 403 and the laser displacement sensor group 404, and is electrically connected to the tension control system 200 to form a closed loop of monitoring-analysis-prediction-control; This embodiment provides a winding device for tire wire. Conventional winding monitoring methods suffer from technical issues such as monitoring lag, blind spots in detecting microscopic defects, and the ability to only perform passive intervention. This embodiment addresses these issues through an integrated multi-physical quantity sensing and control system. The device is based on a multi-station rotary steel wire winding machine 100. Along the steel wire's path, starting from the source, are an adjustable magnetorheological damper tension control system 200 and an intelligent guide wheel assembly 300. After passing through these components, the steel wire is ultimately wound onto a main winding wheel 101. This main winding wheel 101 is located in the core working area of ​​the winding machine and supported by an integrated vibration-isolating winding wheel base 102. This base houses a high-frequency vibration sensor array for collecting vibration signals during the winding process. A 360-degree omnidirectional monitoring frame 400 is deployed around the outside of the main winding wheel 101. This frame and the intelligent winding cavity 401 together form a three-dimensional monitoring space. Inside the space, a thermal imaging system 402, a broadband acoustic emission monitoring system 403, and a laser displacement sensor group 404 are installed to monitor the winding status from multiple physical dimensions. 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 broadband acoustic emission monitoring system 403 and the laser displacement sensor group 404, communicate 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 to the adjustable magnetorheological 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 occur.

[0023] The integrated vibration isolation winding wheel seat 102 is isolated from interfering vibrations from the external environment through multi-layer damping materials and a suspension design; wherein the high-frequency vibration sensor array is designed through bionics and coated with a graphene / piezoelectric ceramic composite nano-coating; This embodiment further defines the integrated vibration isolation winding wheel base 102 and the high-frequency vibration sensor array therein. To address the technical issue of environmental vibration (such as motor and frame vibration) interfering with high-sensitivity measurements, the integrated vibration isolation winding wheel base 102 employs a unique structural design. Specifically, its internal structure comprises multiple layers of damping materials with varying mechanical properties, combined with a suspension design. This design effectively isolates over 99% of external environmental interference vibrations, ensuring that the signals collected by the sensor array purely reflect the microscopic vibrations generated by the wire winding process itself. To address 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 biomimetic design, and its layout mimics the structure of a living organism's ability to sense vibration. In addition, the sensor's sensing surface is covered with a layer of graphene / piezoelectric ceramic composite nanocoating. This coating material has excellent piezoelectric effect and mechanical properties, and its application increases the sensor's detection sensitivity by three orders of magnitude, enabling the system to capture early vibration precursors caused by slight tension imbalances.

[0024] The intelligent guide wheel assembly 300 includes a plurality of independent guide wheels 301 ; wherein each guide wheel 301 is independently equipped with a composite MEMS sensor chip 302 , and the composite MEMS sensor chip 302 is used to synchronously measure the local vibration, temperature and strain of the steel wire at the contact point of the guide wheel 301 ; This embodiment further defines the structure of the intelligent guide wheel assembly 300. To address the technical issues of difficulty in tracing the source of faults and the inability to accurately locate the source of the abnormality, the intelligent guide wheel assembly 300 is composed of multiple guide wheels 301 that are independent of each other in structure and function. These guide wheels 301 are arranged along the path of the wire before it enters the main winding wheel 101. Each independent guide wheel 301 is equipped with a composite micro-electromechanical system (MEMS) sensor chip; the composite here means that the single chip integrates multiple sensing functions and can simultaneously measure three key physical quantities of the steel wire at the contact point with the guide wheel 301: local vibration, local temperature and local strain; when the data processing and control center 5 detects a systematic tension anomaly, it can compare and analyze the data of the sensor chips on each independent guide wheel 301 along the way to determine whether the abnormal state originated from a specific guide wheel 301 (such as a bearing failure) or existed before entering the intelligent guide wheel group 300; this distributed sensing arrangement significantly enhances the spatial resolution capability and maintainability of the system.

[0025] The data processing and control center 5 is connected to a distributed edge computing node; wherein the distributed edge computing node is deployed near each sensor group, and is used to perform noise reduction and preliminary feature extraction on the raw data, and upload high-value feature information to the data processing and control center 5; This embodiment further defines the data processing architecture; to address the technical issues of data congestion and network delays that may result from directly transmitting massive amounts of raw sensor data to the central processing unit, thereby affecting the real-time performance of the control system, this device adds distributed edge computing nodes between the data processing and control center 5 and each sensor group. The distributed edge computing nodes are small computing units that are physically deployed near each sensor group (such as a high-frequency vibration sensor array, an intelligent guide wheel group 300, etc.); the function of these nodes is to perform on-site preprocessing of the massive raw data generated by the sensors to which they are connected; the preprocessing process mainly includes two tasks: one is to reduce noise through a digital filtering algorithm to filter out irrelevant components in the signal; the other is to perform preliminary feature extraction on the noise-reduced signal, such as calculating the spectrum and time-domain statistical characteristics of the vibration signal; after preprocessing is completed, only the extracted feature data containing high-value status information is uploaded to the data processing and control center 5; this architecture significantly reduces the requirements for the central processing unit and data transmission bandwidth, ensuring the millisecond-level response capability of the entire closed-loop control system.

[0026] The 360-degree omnidirectional monitoring frame 400 is also provided with a high-speed camera system 405, which cooperates with the laser displacement sensor group 404 to scan the surface contour of the wound steel wire layer and the spatial position of the incoming steel wire; This embodiment enhances the functionality of the 360-degree monitoring framework 400. To address potential errors associated with a single measurement method and provide a high-precision geometric basis for training artificial intelligence models, a high-speed camera system 405 is added to the 360-degree monitoring framework 400. The high-speed camera system 405 works in conjunction with the existing laser displacement sensor group 404; the laser displacement sensor group 404, with its 0.01mm measurement accuracy, continuously scans the steel wire layer already wound on the main winding wheel 101 point by point to obtain its precise three-dimensional surface profile 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 work together to not only provide final geometric evidence of the winding quality, used to confirm whether there are physical defects such as intersections and depressions, but more importantly, the high-precision ground truth data they generate 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's predictions and avoiding the risk of erroneous control instructions due to the model being divorced from physical reality.

[0027] Example 2: See also Figure 5The present invention also provides a winding method for tire steel wire, including: system initialization and benchmark library establishment, winding with standard process parameters and qualified steel wire, synchronous collection of a full set of sensor data, establishment of a standard winding pattern library containing ideal voiceprint fingerprints and thermal field distribution, and initial training of the digital twin neural network of the winding process; real-time acquisition and edge preprocessing, starting the winding task, and all sensors synchronously collecting data; denoising, filtering and preliminary feature extraction of the original signal through the distributed edge computing nodes deployed on the device side; multi-scale fusion analysis and pattern recognition: sending the preprocessed feature data to the digital twin neural network model of the data center; the digital twin neural network The network model identifies the spatial patterns of vibration soundprints and thermal field distribution, and compares them with the standard winding pattern library in real time, while correlating the readings of different sensors in time and space. Predictive analysis and defect warning: Based on the fused features, the long-short-term memory network layer in the model is used to deduce future time series data, predict the winding state trend and evaluate the probability of defect occurrence. When the predicted defect probability exceeds the preset safety threshold, the system initiates an early warning. Adaptive closed-loop control and optimization: At the same time as the early warning is issued, the optimal tension adjustment strategy is calculated through the reinforcement learning algorithm, and a control instruction is generated and sent to the adjustable magnetorheological damper tension control system 200 to complete the tension fine-tuning of the corresponding steel wire. This embodiment provides a tire wire winding method applicable to any of the aforementioned devices, which converts the hardware capabilities of the device into a practical effect of preventing defects; Execute system initialization and benchmark library establishment steps; before formal production, use qualified steel wire to carry out winding operations under standard process parameters; during this process, the full set of sensors on the device (vibration, thermal field, acoustic emission, position, etc.) synchronously collect data; this data is used to establish a standard winding pattern library, which stores data patterns such as vibration soundprint fingers and thermal field spatial distribution under ideal winding conditions; at the same time, these benchmark data are also used to perform initial training of the digital twin neural network of the winding process in the data processing and control center 5.

[0028] The real-time acquisition and edge preprocessing steps begin. The wrapping task officially begins, and all sensors synchronously collect real-time data at a preset frequency (e.g., 1kHz). Distributed edge computing nodes deployed throughout the device perform real-time noise reduction, filtering, and preliminary feature extraction on the collected raw signals. Perform 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 this model performs spatial pattern recognition on the input vibration soundprint and thermal field distribution map, and compares them with the reference patterns in the standard winding pattern library in real time to identify deviations; at the same time, the model's spatiotemporal attention mechanism is responsible for correlating readings from different sensors at different time and spatial locations to construct a complete state causal chain; Perform predictive analysis and defect warning steps; based on the multi-dimensional features obtained from the previous fusion analysis, the long-short-term memory network layer in the model deduces the time series of the data, thereby being able 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 loosening in real time; when the predicted probability of any defect exceeds the preset safety threshold, the system immediately activates the warning mechanism; the preset safety threshold is a critical probability value determined through statistical analysis based on historical production data and experimental results; specifically, a large amount of sensor data corresponding to normal winding samples and known defect samples is collected, and machine learning models (such as logistic regression or support vector machines) are trained to find a probability point that can maximize the distinction between normal and abnormal states while balancing the omission rate and false alarm rate; for example, the threshold can be set so that when the model predicts that the probability of wire crossing within the next 60 seconds exceeds 5%, the system will initiate a warning; this threshold can also be adjusted online by technicians according to product quality level requirements; Adaptive closed-loop control and optimization steps are executed. When the warning is triggered, a reinforcement learning algorithm within the control system calculates the optimal tension adjustment strategy for correcting the potential defect in real time based on the current system state and the predicted defect type. This strategy is converted into specific control instructions and sent to the adjustable magnetorheological damper tension control system 200. Upon receiving the instructions, the target wire tension is precisely fine-tuned within 10 milliseconds. This series of steps forms a complete closed loop from prediction to execution, eliminating the root cause of physical defects (such as tension imbalance) before they actually form. Specifically, the reinforcement learning algorithm model is constructed as follows: State space: defined as a vector consisting of pre-processed characteristic data of all sensors at time t, including but not limited to the vibration beat frequency amplitude, local temperature, strain, thermal field distribution characteristics of each steel wire, and geometric profile deviation value obtained by laser scanning; Action space: defined as the voltage or current adjustment applied to each adjustable MR damper tension control system 200. This adjustment corresponds to the change in tension. Actions can be discrete (e.g., increase / decrease / hold) or continuous (precise adjustment within a certain range). Reward function: Designed to guide model learning; a positive reward is given when the system successfully avoids a predicted defect (i.e., after an early warning is issued, subsequent sensor data shows that the state has returned to normal); a negative reward is given when the defect still occurs after the early warning, or when the system makes unnecessary adjustments (cross-validation fails); 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 which tension adjustment action to take to obtain the maximum long-term cumulative reward under a given system state, that is, the optimal tension adjustment strategy.

[0029] In the multi-scale fusion analysis and pattern recognition step, the digital twin neural network model is based on the wire winding acoustic resonance theory and quantifies the tension inconsistency by identifying the beat frequency phenomenon formed by the vibration interference caused by the inconsistent tension between multiple strands of steel wire; This embodiment defines a specific technical approach in the multi-scale fusion analysis and pattern recognition step. To address the technical issue of being unable to quantitatively detect minute tension differences between multiple strands of steel wire, the digital twin neural network model applies the wire winding acoustic resonance theory for analysis in this step. The theory states that, under ideal conditions, a steel wire moving at a specific tension will produce a stable, baseline inherent vibration spectrum, namely, the voiceprint fingerprint. When the tension between multiple parallel-wound steel wires is inconsistent, their respective vibrations will interfere, thereby generating a specific beat frequency phenomenon in the superimposed vibration signal. The model specifically identifies this beat frequency phenomenon by performing spectrum analysis on the signals collected by the high-frequency vibration sensor array. Its theoretical basis is that if the main vibration frequencies of two adjacent steel wires are different due to tension differences, and , then the synthetic vibration will produce a frequency of The beat frequency signal is generated by the beat frequency. By calculating the frequency and amplitude of the beat frequency, the model can accurately quantify the degree of tension inconsistency between the strands of steel wire. This quantification result is the key input parameter for subsequent prediction of defect probability and accurate tension compensation.

[0030] Before the adaptive closed-loop control and optimization step is initiated, a cross-validation step is also included; wherein the cross-validation step uses the thermal field imaging system 402 to identify hot spots generated by friction and confirm the warning issued by the vibration soundprint pattern recognition to eliminate false alarms; This embodiment adds a cross-validation step to address the technical problem that a single sensing system may generate false alarms, leading to unnecessary downtime or incorrect adjustments. This cross-validation step is set after the predictive analysis and defect warning steps and before the adaptive closed-loop control and optimization steps are initiated. When the prediction system based on vibration signal analysis issues an early warning (for example, it predicts that uneven tension may cause steel wire crossing), the control system does not immediately perform tension adjustment; instead, it first starts a cross-validation step; in this step, the system calls the data of the thermal field imaging system 402 to analyze the thermal field distribution map of the area near the early warning position; due to abnormal friction of the steel wire (for example, the position where the crossing is about to occur), the local temperature will rise, forming a hot spot; if the thermal field imaging system 402 also identifies a hot spot that conforms to physical laws at the corresponding position of the early warning, the early warning is confirmed to be valid; only after this multi-physical quantity cross-validation is confirmed, the adaptive closed-loop control and optimization steps will be started; this can effectively eliminate false alarms caused by factors such as accidental vibration interference, thereby improving the reliability of the entire control system.

[0031] It also includes monitoring of internal damage of the material; wherein, the broadband acoustic emission monitoring system 403 captures the high-frequency stress wave signal released by the expansion of microcracks inside the steel wire material, and issues an alarm at the early stage of irreversible damage to the material; This embodiment adds a parallel monitoring dimension to address the technical problem that conventional methods cannot detect damage in the embryonic stage within the material. This step is achieved by a broadband acoustic emission monitoring system 403 provided within the intelligent winding cavity 401. Acoustic emission monitoring is a passive monitoring technology that works by monitoring stress waves released when microscopic damage occurs within a material. During the wire winding process, if the wire material itself has metallurgical defects or microcracks develop under tension, the initiation and expansion of these microcracks will release energy in the form of high-frequency stress waves. The broadband acoustic emission monitoring system 403 is specifically designed to capture these high-frequency, low-energy stress wave signals. When the system captures an acoustic emission signal that matches the characteristics of microcrack expansion, it immediately issues a material damage alarm that is independent of the defect warning. This monitoring enables the system to detect hidden dangers at an early stage before macroscopic irreversible damage such as material fracture occurs, which plays a decisive role in reducing the ultimate wire breakage rate. According to statistics, the wire breakage rate can be reduced by 85%.

[0032] It also includes holographic visualization and quality traceability steps; the entire winding process is reconstructed into a 3D digital twin model in real time, and after winding is completed, a holographic data report containing the wire quality status is generated; This embodiment adds a step that runs through the entire process and extends to the end of the process; in order to solve the technical problems of the winding process being opaque and the quality data being unable to be fully traced, the method includes holographic visualization and quality traceability; Throughout the winding process, the data processing and control center 5 uses all sensor input data (including laser scanning geometry, vibration data, thermal field data, etc.) to reconstruct a three-dimensional digital twin model that accurately corresponds to the physical equipment in real time. This model is visualized on the operation interface, allowing operators to observe the real-time winding status of the wire from any angle. Any warnings or alarms issued by the system are highlighted at the precise location on the three-dimensional model (with an accuracy of up to 0.1mm). After the winding of the entire reel of steel wire is completed, the system automatically performs data aggregation and generates a holographic data report; this report not only includes the conventional production parameters of the reel of steel wire, but more importantly, it records the key quality status data of each section of steel wire during the winding process (such as tension fluctuations, vibration spectrum, temperature changes, etc.) and all potential defect events that have been marked in units of meters; this report provides a detailed quality input basis for subsequent vulcanization and other processes, and realizes quality traceability throughout the product life cycle.

[0033] Compared with the prior art, the following beneficial effects are achieved: A fundamental shift from defect detection to defect prediction has been achieved, significantly improving the foresight of quality control. Existing technologies generally rely on offline spot checks or macro-parameter monitoring, which essentially involves passive identification after defects have formed. The integrated vibration isolation winding wheel seat 102 in this solution, with its multi-layer damping material and suspended design, provides an ultra-low-noise measurement environment for the internal high-frequency vibration sensor array. This environment ensures that the sensor array can capture pure, microscopic vibration signals generated by the wire winding process itself. The key is that the data processing and control center 5 can use these signals to identify vibration interference caused by sub-Newtonian tension inconsistencies between multiple strands of wire, namely the beat frequency phenomenon. By quantifying this phenomenon, the system can accurately predict the occurrence trend of physical defects such as crossovers or depressions tens of seconds before they form. This fundamentally solves the problem of blind spots in the detection of core defect precursors such as tension imbalance in existing technologies. A monitoring system with cross-validation of multiple physical quantities has been established, which has significantly improved the accuracy of early warnings and the dimension of defect identification; Existing technologies often rely on a single information source, which can easily lead to misjudgments. This solution uses a 360-degree all-round monitoring framework 400 to work in conjunction with multiple sensor systems within the intelligent winding cavity 401. When the high-frequency vibration sensor array issues an early warning, the system immediately calls the data from the thermal field imaging system 402 to cross-verify whether there are hot spots caused by abnormal friction at the warning location, effectively eliminating false alarms caused by accidental interference. At the same time, the broadband acoustic emission monitoring system 403 captures the high-frequency stress waves released by the expansion of microcracks within the steel wire material, achieving early warning of irreversible damage within the material, a dimension that traditional monitoring methods cannot reach. The laser displacement sensor group 404 and the high-speed camera system 405 jointly provide high-precision geometric contour and spatial position information, providing a data foundation for the final confirmation of the defect's shape and the precision calibration of the digital twin model. An active and adaptive closed-loop control circuit is established to achieve real-time intervention and optimization of the winding process; Existing intervention measures are passive, typically requiring manual intervention after a problem is discovered, resulting in significant lags. The operational logic of this solution is completely different. After confirming a problem through predictive analysis and defect warning, the data processing and control center 5 immediately calculates the optimal tension adjustment strategy using a reinforcement learning algorithm. This strategy is sent to the adjustable magnetorheological damper tension control system 200 in the form of control instructions. Leveraging the physical properties of magnetorheological fluid, this system can precisely fine-tune the tension of the target wire within 10 milliseconds. This complete closed loop of monitoring, analysis, prediction, and control responds much faster than the formation of physical defects, ensuring that corrective actions are always completed before defects solidify, transforming quality control from passive response to active avoidance. It realizes holographic traceability of the production process and provides data support for quality management and process optimization; The quality records of existing technologies are usually limited to batches and cannot be traced back to the quality status of specific sections within a single reel of steel wire. This solution uses holographic visualization and quality traceability steps to reconstruct the entire winding process into a three-dimensional digital twin model in real time. After winding is completed, the system will generate a holographic data report containing the quality status of each meter of steel wire, detailing key indicators such as tension fluctuations and vibration spectra. This not only provides a detailed basis for incoming material quality 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.

[0034] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A winding device for tire wire, characterized in that: include: A multi-station rotary wire winding machine (100), an adjustable magnetorheological damper tension control system (200), an intelligent guide wheel assembly (300), a main winding wheel (101), an integrated vibration isolation winding wheel seat (102), a 360-degree all-round monitoring frame (400), an intelligent winding cavity (401), and a data processing and control center (5); 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 steel wire; The integrated vibration isolation winding wheel seat (102) is used to support the main winding wheel (101), and a high-frequency vibration sensor array is installed inside the integrated vibration isolation winding wheel seat (102); The intelligent guide wheel group (300) and the tension control system (200) are arranged upstream of the main winding wheel (101); wherein, after the steel wire is drawn out from the source, it passes through the tension control system (200) and the intelligent guide wheel group (300) in sequence, and is finally wound around the main winding wheel (101); The 360-degree omnidirectional monitoring frame (400) and the intelligent winding cavity (401) are disposed around the outside of the main winding wheel (101), and together constitute a monitoring space; wherein the 360-degree omnidirectional monitoring frame (400) and the cavity are provided with a thermal field imaging system (402), a broadband acoustic emission monitoring system (403), and a laser displacement sensor group (404); The data processing and control center (5) respectively communicates with the high-frequency vibration sensor array, the intelligent guide wheel group (300), the thermal field imaging system (402), the broadband acoustic emission monitoring system (403) and the laser displacement sensor group (404), and is electrically connected to the tension control system (200) to form a closed loop of monitoring-analysis-prediction-control.

2. A winding device for tire wire according to claim 1, characterized in that: The integrated vibration isolation winding wheel seat (102) isolates interference vibrations from the external environment through multi-layer damping materials and a suspension design; wherein the high-frequency vibration sensor array is designed through bionics and is covered with a graphene / piezoelectric ceramic composite nanocoating.

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

4. A winding device for tire wire according to claim 1, characterized in that: The data processing and control center (5) is connected to a distributed edge computing node; wherein the distributed edge computing node is deployed near each sensor group, and is used to perform noise reduction and preliminary feature extraction on the raw data, and upload high-value feature information to the data processing and control center (5).

5. The winding device for tire wire according to claim 1, characterized in that: The 360-degree omnidirectional monitoring frame (400) is further provided with a high-speed camera system (405), which cooperates with the laser displacement sensor group (404) to scan the surface contour of the wound steel wire layer and the spatial position of the incoming steel wire.

6. A winding method for tire wire, applied to a winding device for tire wire according to any one of claims 1 to 5, characterized in that: include: System initialization and benchmark library establishment, using standard process parameters and qualified steel wire for winding, synchronously collecting a full set of sensor data, establishing a standard winding pattern library containing ideal voice fingerprints and thermal field distribution, and performing initial training of the winding process digital twin neural network; Real-time data collection and edge preprocessing are performed, and the wrapping task is started. All sensors collect data synchronously. Distributed edge computing nodes deployed on the device side are used to reduce noise, filter, and perform preliminary feature extraction on the original signal. Multi-scale fusion analysis and pattern recognition: The pre-processed feature data is fed into the digital twin neural network model in the data center. The digital twin neural network model identifies the spatial patterns of the vibration soundprint and thermal field distribution, compares them with the standard winding pattern library in real time, and correlates the readings of different sensors in time and space. Predictive analysis and defect warning: Based on the fused features, the model's long-short-term memory network layer extrapolates future time series data to predict winding state trends and assess defect probability. When the predicted defect probability exceeds the preset safety threshold, the system initiates an alert. Adaptive closed-loop control and optimization, when the warning is issued, the optimal tension adjustment strategy is calculated through the reinforcement learning algorithm, and a control instruction is generated and sent to the adjustable magnetorheological damper tension control system (200) to complete the tension fine-tuning of the corresponding steel wire.

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

8. The method for winding tire wire according to claim 6, characterized in that: Before the adaptive closed-loop control and optimization step is started, a cross-validation step is also included; wherein the cross-validation step identifies hot spots generated by friction through a thermal field imaging system (402) and confirms the warning issued by the vibration soundprint pattern recognition to eliminate false alarms.

9. A method for winding tire wire according to claim 6, characterized in that: It also includes monitoring of internal damage of the material; wherein, a broadband acoustic emission monitoring system (403) is used to capture high-frequency stress wave signals released due to the expansion of microcracks inside the steel wire material, and an alarm is issued at the early stage of irreversible damage to the material.

10. A winding method for tire wire according to claim 6, characterized in that: It also includes holographic visualization and quality traceability steps; among them, the entire 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 status of the steel wire is generated.

Images