Multi-directional pre-deformation device for tire bead and method thereof

Abstract

The application discloses a multi-directional pre-deformation device and method for a tire steel ring, and belongs to the technical field of intelligent manufacturing. The device comprises a digital twin system, which is used for processing data and planning control instructions; an annular segmented electromagnetic deformation chamber, which is used for multi-directional deformation of the steel ring; a sensor array, which is used for real-time monitoring of the state of the steel ring and transmission of data to the digital twin system; a non-magnetic support and positioning mechanism, which is arranged at the center of the annular segmented electromagnetic deformation chamber and is used for placing the steel ring; and a controller, which is connected to control the work of the digital twin system, the annular segmented electromagnetic deformation chamber and the sensor array. The application eliminates the influence of material differences on product accuracy and significantly reduces the waste rate.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing, specifically to a multi-directional pre-deformation device and method for tire steel wire rings. Background Technology

[0002] In the tire wire bead manufacturing industry, pre-deformation of the welded wire bead is a critical process. However, traditional mechanical rolling or molding equipment has many drawbacks. Due to differences in the physical properties of steel wire materials between batches, such as elastic modulus and yield strength, and the use of fixed displacement or pressure programs in traditional equipment, the geometric accuracy of the final product is inconsistent, resulting in defects such as non-compliance in cone angle, diameter, and roundness, leading to a high scrap rate. Traditional equipment is mostly unidirectional or two-dimensional deformation, making it difficult to complete complex three-dimensional contour forming in one go. This usually requires multiple processes and various equipment, which not only increases the production cycle time but also raises equipment investment costs. In addition, changing to different specifications of wire bead products requires a lot of time for mechanical mold changing and debugging, making it difficult to adapt to the current flexible production needs of small batches and multiple varieties. Traditional automated equipment lacks the ability to perceive and adjust the deformation process in real time, and cannot fundamentally solve the problem of accuracy loss caused by material inconsistency.

[0003] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to provide a multi-directional pre-deformation device and method for tire steel wire rings to solve the problems mentioned in the background art.

[0005] The technical solution of the present invention includes:

[0006] A digital twin system is used to process data and plan control commands;

[0007] An annular segmented electromagnetic deformation chamber is used for multi-directional deformation of steel wire coils;

[0008] A sensor array is used to monitor the status of the wire coil in real time and transmit data to the digital twin system;

[0009] A non-magnetic support and positioning mechanism is located at the center of the annular segmented electromagnetic deformation chamber and is used to place the steel wire ring.

[0010] The controller is connected to control the operation of the digital twin system, the annular segmented electromagnetic deformation chamber, and the sensor array.

[0011] Preferably, the annular segmented electromagnetic deformation chamber includes:

[0012] The annular base includes two parts, upper and lower, on which multiple modular electromagnetic actuation units are distributed radially and axially.

[0013] An electromagnetic actuator array, which consists of multiple high-power electromagnetic coils, includes a radial actuation module, an axial actuation module, and a torsional actuation module.

[0014] Preferably, the radial actuation modules are evenly distributed along the circumference of the chamber and are used to apply radial force to the wire coil.

[0015] Preferably, the axial actuation module is arranged on the two annular bases for applying axial force to the wire ring.

[0016] Preferably, the torsion actuation module is used to apply a torsional force to correct local twisting of the wire coil.

[0017] Preferably, the sensor array includes:

[0018] A line laser 3D contour scanner is used to perform three-dimensional contour scanning of the wire ring;

[0019] A distributed strain sensor is used to monitor the strain and stress distribution of the wire coil during deformation.

[0020] A Hall effect sensor is used to measure the strength of the magnetic field generated by the electromagnetic actuator.

[0021] An infrared temperature sensor is used to measure the real-time temperature of the wire coil.

[0022] Preferably, the digital twin system includes:

[0023] A springback prediction and model predictive control algorithm is used to predict the springback error and plan the timing and intensity sequence of multi-directional electromagnetic force application for the multi-directional deformation.

[0024] A residual correction neural network is used to learn historical data of the instructions, the deformation, and the rebound, and to correct and optimize the rebound prediction and model prediction control algorithms.

[0025] A method for controlling a multi-directional pre-deformation device for tire steel wire rings includes:

[0026] S1. Place the steel wire ring on the non-magnetic support and positioning mechanism, acquire the initial data of the steel wire ring through the sensor array, and transmit it to the digital twin system;

[0027] S2. The digital twin system predicts the target intermediate geometry required to counteract elastic rebound based on the initial data and target model, and plans the optimal timing and intensity sequence of multi-directional electromagnetic force application.

[0028] S3. The annular segmented electromagnetic deformation chamber applies electromagnetic force to the steel wire ring according to the timing and intensity sequence;

[0029] S4. The sensor array continuously monitors the real-time shape and stress of the wire ring and feeds the data back to the digital twin system in real time. The digital twin system dynamically adjusts the subsequent force application command based on the feedback data.

[0030] S5. After deformation is completed, the digital twin system unloads the electromagnetic force and scans again to calculate the actual rebound error, and uses the data as a new sample for retraining the residual correction neural network.

[0031] This invention provides an improved multi-directional pre-deformation device and method for tire steel wire rings, which has the following improvements and advantages compared with the prior art:

[0032] 1. This solution addresses this issue by introducing a digital twin system and a sensor array. The springback prediction and model predictive control algorithms within the digital twin system predict the target intermediate geometry required to counteract elastic springback based on the input material batch number and real-time acquired initial state data. Based on this, it plans the optimal timing and intensity sequence for multi-directional electromagnetic force application. During deformation, the line laser 3D contour scanner and distributed strain sensors in the sensor array continuously monitor the real-time shape and stress of the wire coil and feed the data back to the controller in real time. The digital twin system dynamically adjusts subsequent force application commands based on the feedback data, and the controller executes these commands, correcting deviations in real time and ensuring the geometric accuracy consistency of the final product. This closed-loop control based on model prediction and real-time feedback eliminates the impact of material differences on product accuracy and significantly reduces the scrap rate.

[0033] 2. The core innovation of this solution lies in the annular segmented electromagnetic deformation chamber, which realizes the transformation from mechanical contact rigid forming to electromagnetic non-contact flexible shaping. This chamber is composed of radial actuation modules, axial actuation modules, and torsional actuation modules. These modules can work together to apply non-contact, distributed, and multi-directional forces to the wire ring by controlling the intensity and direction of the electromagnetic field in different areas at high frequency and high precision. The radial actuation module is used to precisely control the diameter and roundness; the axial actuation module is used to precisely control the height and cone angle, and form complex non-planar contours; the torsional actuation module is used to correct local distortions. This design allows complex three-dimensional contours to be completed in one flexible shaping process, eliminating the limitations of traditional physical molds, thereby reducing production steps, shortening the production cycle, and reducing equipment investment costs.

[0034] 3. The moldless manufacturing concept of this solution solves this pain point; since the deformation process is realized through a programmable electromagnetic actuator array, when it is necessary to change steel wire ring products of different specifications, the operator only needs to input the new target 3D model in the software interface, and the system can quickly load the corresponding model and replan the deformation path without any mechanical mold change or debugging; this greatly shortens the product changeover time, enabling the production line to respond quickly to market demands and improve the level of production flexibility;

[0035] 4. This solution achieves system self-optimization and continuous improvement by introducing a residual correction neural network. After deformation, the system performs a fine scan again to calculate the actual springback error. This error data, along with the entire process data of this task, serves as a new sample for retraining the residual correction neural network. The network corrects and optimizes the physical model by learning from historical data of command-deformation-springback. This continuous learning and iterative process makes the system increasingly adaptable to batch differences in materials and environmental changes, ensuring that the processing accuracy continuously improves with the increase in production frequency, achieving the beneficial effect of becoming more accurate with use. Attached Figure Description

[0036] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0037] Figure 1 This is a schematic diagram of the overall structure of the device;

[0038] Figure 2 This is a schematic diagram of the structure of a digital twin system, sensor array, and controller;

[0039] Figure 3 This is a schematic diagram of the process flow of the method of the present invention;

[0040] In the diagram: 100, Digital Twin System; 200, Annular Segmented Electromagnetic Deformation Chamber; 210, Annular Base; 221, Radial Actuation Module; 222, Axial Actuation Module; 223, Torsional Actuation Module; 300, Sensor Array; 400, Non-Magnetic Support and Positioning Mechanism; 5, Controller. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0042] Example 1

[0043] Please see Figure 1-2 This invention provides a multi-directional pre-deformation device for tire steel wire rings, comprising:

[0044] Digital twin system 100, which is used to process data and plan control commands;

[0045] Annular segmented electromagnetic deformation chamber 200, which is used for multi-directional deformation of steel wire ring;

[0046] Sensor array 300 is used to monitor the status of the wire loop in real time and transmit data to digital twin system 100;

[0047] A non-magnetic support and positioning mechanism 400 is located at the center of the annular segmented electromagnetic deformation chamber 200 and is used to place the steel wire ring.

[0048] Controller 5 is connected to control the operation of digital twin system 100, annular segmented electromagnetic deformation chamber 200 and sensor array 300;

[0049] The controller 5 is responsible for accurately sending the control commands generated by the digital twin system 100 to the electromagnetic actuator array and receiving data feedback from the sensor array 300 in real time, thereby enabling the coordinated operation of the entire device.

[0050] This embodiment discloses a multi-directional pre-deformation device for tire steel wire rings. This device aims to solve the pain points of traditional mechanical roller pressing or molding devices when pre-deforming welded steel wire rings, such as poor geometric accuracy consistency, high scrap rate, and low production efficiency caused by batch material differences, limited forming dimensions, and time-consuming mold changing and debugging. The device adopts a brand-new electromagnetic non-contact flexible shaping technology, which realizes one-time shaping of complex three-dimensional contours and real-time self-adaptation to material properties, thereby significantly improving product accuracy and production flexibility.

[0051] The non-magnetic support and positioning mechanism 400 can be specifically composed of a three-jaw or multi-jaw chuck made of ceramic or non-magnetic polymer material. The chuck can be precisely clamped and positioned by pneumatic or electric means. To ensure positioning accuracy, each jaw of the chuck can be integrated with a micro displacement sensor to provide real-time feedback on the initial position of the wire coil and compare it with the reference model in the digital twin system 100. The mechanism can also be designed to be adjustable in height and tilt angle to accommodate wire coils of different specifications and initial states.

[0052] This device mainly consists of the following core components:

[0053] Digital Twin System 100: As the brain of the entire device, it is not only the executor of preset instructions, but also an intelligent closed-loop system that can realize perception-prediction-decision-optimization. The system integrates advanced physical models, machine learning models and control algorithms, and can process real-time data from sensor array 300 and dynamically plan the optimal deformation path and control instructions according to the target model.

[0054] Annular segmented electromagnetic deformation chamber 200: This chamber is the core actuator of this device. Its main body is an annular segmented structure, and the interior is composed of multiple independent and controllable electromagnetic actuator arrays. This design realizes the transformation from traditional mechanical contact rigid forming to electromagnetic non-contact flexible forming. It can apply non-contact, distributed, and multi-directional electromagnetic forces to the wire ring, thereby completing the complex 3D contour forming in one go.

[0055] Sensor array 300: This array consists of a variety of high-precision sensors, which serve as the sensing organs of the system to monitor the initial state of the wire coil before and after deformation and the dynamic changes during the deformation process in real time and at high frequency. These sensors include a line laser 3D profile scanner, a distributed strain sensor, a Hall effect sensor, and an infrared temperature sensor, which provide comprehensive and accurate real-time data feedback for the digital twin system 100 and are the basis for realizing closed-loop control.

[0056] Non-magnetic support and positioning mechanism 400: This mechanism is located at the center of the annular segmented electromagnetic deformation chamber 200 and is used to stably place the steel wire ring to be processed; because it is made of non-magnetic material, it can ensure that the distribution of electromagnetic field will not be disturbed during electromagnetic deformation, thereby ensuring the accuracy of the deformation force application.

[0057] The mechanism can be made of a three-jaw or multi-jaw chuck made of ceramic or non-magnetic polymer materials. To ensure positioning accuracy, each jaw of the chuck can be integrated with a micro displacement sensor. These sensors and their connections are made of non-magnetic materials that are resistant to magnetic field interference or are treated with special electromagnetic shielding to ensure that the distribution of the electromagnetic field will not be interfered with during electromagnetic deformation, thereby ensuring the accuracy of the deformation force.

[0058] Controller 5: As the central hub connecting the digital twin system 100, the annular segmented electromagnetic deformation chamber 200 and the sensor array 300, the controller 5 is responsible for accurately sending the control commands generated by the digital twin system 100 to the electromagnetic actuator array and receiving data feedback from the sensor array 300 in real time, thereby realizing the coordinated operation of the entire device.

[0059] The device operates within a tight closed loop of perception, cognition, action, and learning: the sensor array 300 senses the initial state of the wire loop; the digital twin system 100 combines the model to perform cognition and decision-making, planning the optimal deformation path; the controller 5 drives the electromagnetic chamber to precisely execute the deformation command; the sensors scan again and provide feedback data for the digital twin system 100 to perform self-optimization learning; this collaborative working mode solves the precision and flexibility problems existing in traditional processes, realizing moldless intelligent manufacturing.

[0060] The annular segmented electromagnetic deformation chamber 200 includes:

[0061] The annular base 210 includes two parts, upper and lower, and multiple modular electromagnetic actuation units are distributed radially and axially on the annular base 210.

[0062] The electromagnetic actuator array consists of multiple high-power electromagnetic coils and includes a radial actuation module 221, an axial actuation module 222, and a torsional actuation module 223.

[0063] This embodiment further describes in detail the annular segmented electromagnetic deformation chamber 200 mentioned above; this chamber is the core actuator of this device, and its innovation lies in replacing the traditional fixed mold with a flexible one.

[0064] The structure of the chamber includes:

[0065] Annular base 210: Composed of upper and lower annular bases 210, these two annular bases 210 provide stable structural support for the entire chamber and serve as a platform for installing the electromagnetic actuator array; on the annular base 210, modular electromagnetic actuator units are arranged radially and axially to jointly construct annular or polygonal deformation fields; this segmented and modular design makes the maintenance and expansion of the equipment more convenient;

[0066] Electromagnetic actuator array: This array consists of multiple high-power electromagnetic coils, each module driven by an independent, high-speed-response power supply. This allows each module to generate a pulsed magnetic field with precisely controllable intensity and direction, thereby applying precise electromagnetic force to the wire coil. Based on its function and installation location, the array is further divided into three modules that work together to achieve complex three-dimensional molding:

[0067] The electromagnetic actuator array consists of multiple high-power, high-speed response electromagnetic coils. These coils can be wound with Litz wire to reduce losses caused by the skin effect and proximity effect. The drive power supply can be an IGBT-based half-bridge or full-bridge circuit. Through high-frequency pulse width modulation control, the pulse current of the electromagnetic actuator can be precisely controlled, thereby accurately adjusting the strength and timing of the electromagnetic force.

[0068] Radial actuation modules 221: These modules are evenly distributed along the circumference of the chamber and are used to apply radial forces to the wire coil, either pointing towards or away from the central axis; by precisely controlling the magnitude and timing of these radial forces, micron-level precise control of the wire coil diameter, roundness and local curvature can be achieved.

[0069] Axial actuation module 222: These modules are arranged on the upper and lower annular bases 210 and are used to apply an axial force perpendicular to the plane of the wire ring. Through these axial forces, the height and cone angle of the wire ring can be precisely controlled, and a complex three-dimensional contour that is not planar can be formed at one time, which completely solves the limitations of traditional two-dimensional deformation technology.

[0070] Torsional actuation module 223: This module can be realized by alternating energization of coil arrays of some radial or axial modules to generate a local rotating magnetic field; this magnetic field can apply a small torsional torque to the wire coil to correct local torsion defects caused by welding or uneven material stress;

[0071] The coordinated operation of these three actuation modules enables the device to achieve multi-dimensional, high-precision, one-time flexible shaping, thereby eliminating the dependence on physical molds, significantly shortening product changeover time, and significantly improving production flexibility.

[0072] Radial actuation modules 221 are evenly distributed along the circumference of the chamber and are used to apply radial force to the wire coil;

[0073] This embodiment describes in detail the working principle of the radial actuation module 221 described above; this module is the key to achieving precise control of the diameter and roundness of the wire coil.

[0074] Traditional mechanical rolling methods typically apply radial deformation force through rollers with fixed geometry, making it difficult to finely adjust the local curvature of the wire coil and unable to adapt to the elasticity differences between different batches of materials, which can easily lead to the final product having unqualified diameter and roundness.

[0075] In this embodiment, the radial actuation modules 221 are uniformly distributed along the circumference of the annular cavity. Each radial actuation module 221 includes a pair of high-power electromagnetic coils, the central axis of which is tangent to the radial plane of the wire coil. By controlling the pair of coils to be supplied with high-frequency pulse current in the same or opposite directions, a precisely controllable gradient magnetic field is generated. When this gradient magnetic field acts on the wire coil placed at the center of the cavity, induced eddy currents are generated inside the wire coil according to the principle of electromagnetic induction. According to the Lorentz force law, the gradient magnetic field and the induced eddy currents interact to generate a radial electromagnetic force pointing towards or away from the central axis.

[0076] Implementation process:

[0077] The digital twin system 100 calculates the radial electromagnetic force distribution required to achieve the target diameter and roundness based on the target model and real-time sensor data;

[0078] The controller 5 converts these instructions into corresponding high-frequency pulse current sequences and sends them to the radial actuation modules 221 distributed along the circumference.

[0079] Each module generates pulsed electromagnetic force with precise controllable strength and direction according to instructions, and pushes or pulls the steel wire coil in a non-contact, distributed manner in the circumference.

[0080] During the deformation process, the 3D contour scanner continuously monitors the real-time diameter and roundness of the wire ring and feeds the data back to the digital twin system 100.

[0081] The system dynamically adjusts subsequent pulse currents based on feedback deviations, corrects radial force in real time, and ensures final diameter and roundness accuracy.

[0082] The radial actuation module 221 can apply precise and adjustable radial force to any circumferential position of the wire coil in a non-contact manner, achieving sub-millimeter or even micrometer-level precise control over diameter and roundness. This flexible shaping method completely eliminates the geometric limitations of physical rollers, significantly improves product accuracy and yield, and can adapt to the elasticity differences of different material batches.

[0083] Axial actuation module 222 is arranged on two annular bases 210 for applying axial force to the wire coil;

[0084] This embodiment describes in detail the working principle of the above-mentioned central axis actuation module 222; this module is the key to realizing the shaping of the wire ring height, cone angle, and non-planar complex contours.

[0085] Traditional molding or rolling devices are mostly unidirectional or two-dimensional deformation devices, which are difficult to complete complex three-dimensional contour forming in one go. For example, to form a wire coil with a specific cone angle or wavy contour, multiple processes and multiple machines are usually required, which not only increases the production cycle time, but also increases the equipment investment cost.

[0086] The axial actuation module 222 in this embodiment is composed of several annular or segmented electromagnetic coils, the central axis of which is parallel to the central axis of the wire ring; these modules are arranged on the upper and lower annular bases 210, and can apply an axial electromagnetic force perpendicular to its plane to the wire ring.

[0087] Implementation process:

[0088] During the predictive deformation path planning phase, the digital twin system 100 calculates the timing of the axial electromagnetic force distribution required to achieve the desired height, cone angle, or complex contour based on the target 3D model.

[0089] During the closed-loop real-time control execution phase, the controller 5 drives the axial actuation modules 222 arranged at different heights and positions;

[0090] By precisely controlling the energizing timing and intensity of these coils, a distributed and controllable axial electromagnetic force can be applied to the wire coil, thereby achieving precise control over its height and cone angle.

[0091] When it is necessary to form a complex non-planar contour, the system can apply forces of different magnitudes and directions to the axial actuation modules 222 in different regions to achieve local upward or downward deformation, thereby completing the forming of a complex three-dimensional contour in one go.

[0092] During the deformation process, the 3D contour scanner continuously monitors the real-time three-dimensional shape of the wire ring and feeds the data back to the controller 5, which is used to dynamically adjust the subsequent axial force application command to correct the deviation in real time.

[0093] The axial actuation module 222 overcomes the limitations of traditional devices in unidirectional or two-dimensional deformation by applying multi-directional force in a non-contact manner. It makes it possible to shape complex three-dimensional contours in one go, greatly reducing the number of production steps and equipment, significantly improving production efficiency, and reducing equipment investment costs.

[0094] The torsional actuation module 223 is used to apply torsional force to correct local twisting of the wire coil.

[0095] This embodiment describes in detail the working principle of the torsion actuation module 223 described above; this module is an auxiliary functional module used to correct local torsion defects in the wire coil;

[0096] During the manufacturing process of wire rings, especially in the welding and pre-deformation stages, uneven internal stress and unbalanced force of the material can easily lead to local twisting defects in the wire rings. Traditional straightening methods usually require additional processes and equipment, are cumbersome to operate, and have limited accuracy.

[0097] In this embodiment, the torsional actuation module 223 can be implemented by alternating energization of a coil array of a portion of the radial actuation module 221; through a specific current control strategy, these coils can generate a local rotating magnetic field, thereby applying a small torsional torque to the wire coil.

[0098] Implementation process:

[0099] During the real-time monitoring phase of the initial calibration or deformation process, a line laser 3D profile scanner or distributed strain sensor detects a torsion defect in a local area of ​​the wire ring.

[0100] The digital twin system 100 calculates the magnitude and area of ​​the torsional moment required to counteract this distortion;

[0101] The controller 5 sends instructions to the radial actuation module 221 at the corresponding position, so that it alternately energizes according to a preset timing sequence;

[0102] The local rotating magnetic field generated by these modules interacts with the eddy current induced by the wire coil, producing a small torsional torque, thereby performing non-contact, precise reverse torsional correction on the tortuous area.

[0103] During the correction process, the sensor continuously monitors the torsion state and feeds the data back to the controller 5 in real time until the torsion is effectively corrected.

[0104] The torsion actuation module 223 enables the correction of torsional defects to be completed simultaneously in the main deformation process without the need for additional processes and equipment, thus simplifying the production process. This non-contact, flexible correction method can achieve precise and controllable correction of local torsional distortion, further improving the geometric accuracy and quality consistency of the final product.

[0105] Sensor array 300 includes:

[0106] Line laser 3D contour scanner, used for three-dimensional contour scanning of wire loops;

[0107] Distributed strain sensors are used to monitor the strain and stress distribution of a wire coil during deformation.

[0108] Hall effect sensor, used to measure the strength of the magnetic field generated by an electromagnetic actuator;

[0109] Infrared temperature sensor, used to measure the real-time temperature of the wire coil;

[0110] This embodiment describes in detail the structure and function of the sensor array 300 described above; the sensor array 300 is the sensing basis for this device to realize intelligent closed-loop control and self-optimization;

[0111] Traditional automated devices lack the ability to perceive the deformation process in real time and cannot acquire dynamic change data of the physical state and geometric shape of the material. Therefore, they cannot fundamentally solve the problem of accuracy loss caused by material inconsistency.

[0112] The sensor array 300 in this embodiment consists of various types of sensors that work together to provide comprehensive, high-frequency real-time data to the digital twin system 100, thereby building a powerful sensing system.

[0113] Line laser 3D contour scanner: This sensor is deployed above and below the deformation chamber to perform high-speed, full three-dimensional contour scanning of the wire ring before, during and after deformation; it can acquire the initial shape, real-time deformation and final geometry of the wire ring with extremely high accuracy; these precise shape data are the basis of closed-loop control and an important basis for calculating springback error;

[0114] Distributed strain sensor: This sensor can be a non-contact eddy current sensor or integrated into the support mechanism; it is used to monitor the strain and stress distribution of the wire ring in real time during the deformation process; by obtaining internal stress feedback, the digital twin system 100 can determine whether the material is in a dangerous plastic deformation state, thereby preventing excessive stretching or stress concentration of the material and ensuring the structural integrity of the product.

[0115] Hall effect sensor: This sensor is integrated near each electromagnetic actuator to accurately measure the actual magnetic field strength generated; it serves as a self-testing tool for the actuator, ensuring that the commands issued by the controller 5 match the actual force field generated, and providing crucial feedback data for precise control;

[0116] Infrared temperature sensor: This sensor is used to measure the surface temperature of the steel wire coil in real time. Since temperature affects the mechanical properties of materials such as elastic modulus and yield strength, the data provided by this sensor is a key environmental parameter for the physical model to make accurate predictions, ensuring the stability of the system under different ambient temperatures.

[0117] Technical benefits: The multi-functional sensor array 300 provides the digital twin system 100 with real-time data on shape, internal stress, external magnetic field, temperature, and other dimensions, enabling comprehensive and accurate monitoring of the entire deformation process. This powerful sensing capability allows the system to correct deviations in real time and provides high-quality data samples for subsequent self-optimization learning, fundamentally solving the problem of accuracy loss caused by material inconsistency.

[0118] Digital twin system 100 includes:

[0119] A springback prediction and model predictive control algorithm is used to predict the springback error and plan the timing and intensity sequence of multi-directional electromagnetic force application for the multi-directional deformation.

[0120] The algorithm takes the initial geometry, material batch number, and final target model as input and uses a physics-machine learning hybrid model to predict the springback after deformation; this model characterizes the springback properties of the material under different stress states. Based on the predicted springback, the algorithm inversely calculates the overshoot intermediate target shape and generates the electromagnetic force sequence required to achieve this intermediate shape.

[0121] The residual correction neural network is used to learn historical data of commands, deformation, and rebound, and to correct and optimize the rebound prediction and model predictive control algorithms.

[0122] The neural network receives historical data samples, including control commands, deformation process data, environmental parameters, and the final springback residual vector as input. It uses deep learning to identify the complex nonlinear mapping relationship between material properties, environmental changes, and actual springback error. The output of the network is a correction factor, which is used to correct and optimize the physical model parameters in the MPC algorithm in real time, thereby ensuring that the MPC algorithm can make more accurate predictions and plans when facing new material batches or environmental changes.

[0123] Target intermediate geometry: This is the intermediate target of dynamic programming, used to compensate for material elastic rebound. This shape is obtained by inversely superimposing the final target geometry with the rebound amount predicted by the rebound prediction and model predictive control algorithm based on material properties. After the rebound prediction and model predictive control algorithm calculates this intermediate shape, it will serve as the control target of the model predictive control algorithm to ensure that the final product can achieve the required geometric accuracy.

[0124] This embodiment describes in detail the core software modules and functions of the digital twin system 100 described above; this system is the brain that enables intelligent decision-making and self-optimization in this device;

[0125] Traditional control systems typically execute tasks based on preset fixed programs, lacking the ability to predict and compensate for the elastic rebound of materials, and are unable to learn from and adapt to new variables in the production process; when encountering different batches of materials, the system cannot make adaptive adjustments, resulting in large fluctuations in product accuracy;

[0126] The digital twin system 100 in this embodiment constructs an intelligent closed loop capable of prediction-control-learning by deeply integrating advanced control algorithms and machine learning models.

[0127] Springback Prediction and Model Predictive Control Algorithm: This is the core decision-making algorithm of the system. Utilizing a springback prediction algorithm that combines a physical model with a machine learning correction model, the shape and material parameters of the initial wire coil are analyzed to predict the amount of springback due to elasticity after deformation. To counteract this springback, the system calculates the intermediate geometric shape that needs overshoot. The MPC algorithm uses this intermediate shape as the target and, combined with real-time sensor data, dynamically plans the current / voltage sequence required by each electromagnetic actuation module over a future period, i.e., the timing and intensity sequence of multi-directional electromagnetic force application. This algorithm can adjust control commands in real time to ensure that the target is achieved with maximum efficiency without damaging the material.

[0128] Residual Correction Neural Network: This is the key machine learning optimization loop, enabling the system to self-evolve and become more accurate with each use. After a processing task is completed, the system accurately scans the final formed wire coil and compares its shape with the result of the springback prediction model to obtain the springback residual vector. This vector, along with all the process data of this task, including material batch number, temperature, control command history, sensor readings, etc., is input into the RCNN as a complete data sample. This neural network can learn the complex nonlinear mapping relationship between different input features, such as material properties, ambient temperature, control path, and springback residual. Through periodic retraining, RCNN can continuously correct and optimize the physical model, making the system more adaptable to new material batches or environmental changes.

[0129] Through predictive control of the MPC algorithm and self-optimizing learning of RCNN, this digital twin system 100 achieves real-time adaptation to material differences and environmental changes. It can not only accurately deform the wire coil to the required intermediate shape to offset elastic rebound, but also continuously improve processing accuracy by learning production data, thus solving the accuracy fluctuation problem in traditional solutions.

[0130] Example 2

[0131] Please see Figure 3 A method for controlling a multi-directional pre-deformation device for tire steel wire rings, comprising:

[0132] S1. Place the wire ring on the non-magnetic support and positioning mechanism 400, and acquire the initial data of the wire ring through the sensor array 300 and transmit it to the digital twin system 100.

[0133] S2, the digital twin system 100 predicts the intermediate geometry of the target required to counteract elastic rebound based on the initial data and target model, and plans the optimal timing and intensity sequence of multi-directional electromagnetic force application;

[0134] S3, the annular segmented electromagnetic deformation chamber 200 applies electromagnetic force to the steel wire ring according to the timing and intensity sequence;

[0135] S4. The sensor array 300 continuously monitors the real-time shape and stress of the wire coil and feeds the data back to the digital twin system 100 in real time. The digital twin system 100 dynamically adjusts the subsequent force application commands based on the feedback data.

[0136] S5. After deformation, the digital twin system 100 unloads the electromagnetic force and scans again to calculate the actual rebound error, and uses the data as new samples for retraining the residual correction neural network.

[0137] Springback error calculation and data processing flow:

[0138] Input source: The input to this process is the actual final three-dimensional contour data obtained by the line laser 3D contour scanner after deformation, and the ideal final target model planned by the digital twin system 100 in step S2;

[0139] Logical steps:

[0140] Data alignment: Aligning and matching the actual final 3D contour data obtained from the scan with the ideal final target model in 3D coordinate system;

[0141] Error calculation: After alignment, the distance between all corresponding points of the two models is calculated to obtain the spatial position deviation of each point;

[0142] Vector generation: Collect the spatial deviation data of all points to form a rebound residual vector, which contains error information in each direction;

[0143] Output and flow: The final output of the process is the springback residual vector. This vector will be packaged with the full process data of this task, including material batch number, ambient temperature, all control command sequences and sensor readings, to form a complete data sample. This sample will then be stored in the historical database and used for the retraining of the residual correction neural network.

[0144] Springback residual vector: This is a dynamic index of quantification error, used to characterize the deviation between the actual geometric shape of the wire coil after processing and the model-predicted shape. This vector is obtained by comparing and calculating the final geometric shape data obtained by the line laser 3D contour scanner after deformation with the ideal target geometric shape data initially calculated by the springback prediction and model prediction control algorithm in three-dimensional space. This vector serves as the basis for core logical judgment and will be used as new sample data to be input into the residual correction neural network for the system's self-optimization learning.

[0145] This embodiment describes in detail the control method applied to the above-mentioned multi-directional pre-deformation device; the method is a complete initialization-prediction-execution-feedback-learning closed-loop process, which aims to achieve intelligent, high-precision, and adaptive pre-deformation of tire steel wire rings;

[0146] Traditional control methods are mostly open-loop control or simple PID closed-loop control, which cannot predictively compensate for differences in material properties before processing, nor can they dynamically correct deformation in real time and in multiple dimensions during processing; this makes the accuracy of the final product heavily dependent on the initial settings and material consistency.

[0147] The method in this embodiment deeply integrates the digital twin system 100 with intelligent control algorithms to form a self-evolving intelligent manufacturing control process; the method includes the following key steps:

[0148] Step S1, Initialization and Data Acquisition: The operator inputs the 3D model of the target steel wire ring and the current material batch number into the user interface; the steel wire ring to be processed is placed on the non-magnetic support and positioning mechanism 400 located in the center of the electromagnetic chamber; the system activates the sensor array 300, including a 3D contour scanner and an infrared temperature sensor, to acquire the initial shape, temperature, and other reference data of the steel wire ring, and transmits them to the digital twin system 100; these data are used to create an instance of this processing task in the digital twin system 100, providing initial state information for subsequent prediction and control;

[0149] Step S2, Predictive Path Planning: The digital twin system 100 is the core of this step; it calls a springback prediction algorithm that combines a physical model and a machine learning correction model; based on the initial data obtained in S1 and the target model set by the user, this algorithm predicts the amount of springback that the wire coil will generate due to its inherent elasticity after deformation; to counteract this springback, the system calculates an intermediate geometry that needs to be overtuned as a control target; based on this target, the MPC algorithm plans the optimal timing and intensity sequence of multi-directional electromagnetic force application; this planning process takes into account the material properties of the wire coil, ensuring that the target is achieved efficiently without damaging the material;

[0150] Step S3, Deformation: The system drives the ring electromagnetic actuator array to apply time-varying, multi-directional, and distributed electromagnetic forces to the steel wire ring according to the timing and intensity sequence planned in step S2, thereby achieving its deformation;

[0151] Step S4, Closed-loop real-time correction: During the deformation process, the sensor array 300, including a 3D scanner and strain sensors, continuously monitors the real-time shape and stress of the wire ring; these data are fed back to the MPC controller 5 in the digital twin system 100 in high frequency and real time; the controller 5 dynamically adjusts the control commands that have not yet been executed according to the feedback data, so as to correct any deformation deviations that may be caused by slight differences in materials or changes in the environment in real time.

[0152] Step S5, Self-Optimization Learning: After the electromagnetic force is completely unloaded, the system performs a fine scan on the final formed wire coil to accurately calculate the actual springback error of this task. This error data, along with all the data from the entire task, including material parameters, temperature, control commands, and sensor readings, is stored as a new data sample in the historical database. The system will periodically use the accumulated data samples to retrain the residual correction neural network and automatically update the model parameters. Through this learning and iteration, the system can continuously adapt to new material batches or environmental changes, achieving continuous improvement in processing accuracy.

[0153] This control method deeply integrates predictive compensation, real-time closed-loop control, and self-optimizing learning, solving the accuracy fluctuation problem in traditional processes and making the entire manufacturing process more intelligent, flexible, and reliable.

[0154] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A multi-directional pre-deformation device for tire steel wire rings, characterized in that, include: A digital twin system (100) is used to process data and plan control commands; An annular segmented electromagnetic deformation chamber (200) is used for multi-directional deformation of a wire coil; A sensor array (300) is used to monitor the status of the wire loop in real time and transmit data to the digital twin system (100). A non-magnetic support and positioning mechanism (400) is disposed at the center of the annular segmented electromagnetic deformation chamber (200) for placing the wire ring; The controller (5) is connected to control the operation of the digital twin system (100), the annular segmented electromagnetic deformation chamber (200), and the sensor array (300); The digital twin system (100) includes: A springback prediction and model predictive control algorithm is used to predict the springback error and plan the timing and intensity sequence of multi-directional electromagnetic force application for the multi-directional deformation. A residual correction neural network is used to learn historical data of the command, the deformation, and the rebound, and to correct and optimize the rebound prediction and model prediction control algorithm. The control method applied to the multi-directional pre-deformation device of the tire bead includes: S1. Place the steel wire ring on the non-magnetic support and positioning mechanism (400), and obtain the initial data of the steel wire ring through the sensor array (300) and transmit it to the digital twin system (100). S2. The digital twin system (100) predicts the target intermediate geometry required to counteract elastic rebound based on the initial data and target model, and plans the optimal multi-directional electromagnetic force application timing and intensity sequence. S3. The annular segmented electromagnetic deformation chamber (200) applies electromagnetic force to the wire ring according to the multi-directional electromagnetic force application timing and intensity sequence; S4. The sensor array (300) continuously monitors the real-time shape and stress of the wire ring and feeds the data back to the digital twin system (100) in real time. The digital twin system (100) dynamically adjusts the subsequent force application command based on the feedback data. S5. After the deformation is completed, the digital twin system (100) unloads the electromagnetic force and scans again to calculate the actual rebound error, and uses the data as a new sample for the retraining of the residual correction neural network.

2. The multi-directional pre-deformation device for tire steel wire rings according to claim 1, characterized in that, The annular segmented electromagnetic deformation chamber (200) includes: The annular base (210) includes two parts, upper and lower, and multiple modular electromagnetic actuation units are radially and axially distributed on the annular base (210); An electromagnetic actuator array, which consists of multiple high-power electromagnetic coils, includes a radial actuation module (221), an axial actuation module (222), and a torsional actuation module (223).

3. The multi-directional pre-deformation device for tire steel wire rings according to claim 2, characterized in that, The radial actuation module (221) is evenly distributed along the circumference of the chamber and is used to apply radial force to the wire coil.

4. The multi-directional pre-deformation device for tire steel wire rings according to claim 2, characterized in that, The axial actuation module (222) is arranged on the two annular bases (210) for applying axial force to the wire ring.

5. A multi-directional pre-deformation device for tire steel wire rings according to claim 2, characterized in that, The torsion actuation module (223) is used to apply a torsional force to correct local twisting of the wire coil.

6. The multi-directional pre-deformation device for tire steel wire rings according to claim 1, characterized in that, The sensor array (300) includes: A line laser 3D contour scanner is used to perform three-dimensional contour scanning of the wire ring; A distributed strain sensor is used to monitor the strain and stress distribution of the wire coil during deformation. A Hall effect sensor is used to measure the strength of the magnetic field generated by the electromagnetic actuator. An infrared temperature sensor is used to measure the real-time temperature of the wire coil.

Images