Multi-axis synchronous tension control method of vertical winding machine

Abstract

The present application relates to the technical field of automatic winding equipment control, and discloses a multi-axis synchronous tension control method of a vertical winding machine, which collects cumulative rotation angles of each axis through a servo drive module, uses a layering accumulation algorithm containing a downward rounding logic to estimate a current physical roll diameter in real time, and realizes soft measurement of the layering and stacking behavior of wire rods. The system calculates the mean value of roll diameters of all working axes as a dynamic consistency benchmark, monitors single-axis deviation to identify abnormal states. At the same time, the system calculates a gain correction coefficient according to the estimated roll diameter, adaptively adjusts the proportional, integral and derivative gains of a PID controller to offset system gain attenuation caused by nonlinear increase of load moment of inertia. A main control unit finally calculates a torque compensation amount in combination with tension feedback, drives a servo motor to adjust output torque. The present application effectively solves the problem of tension fluctuation under large inertia working conditions and improves the consistency of multi-axis winding.

Description

Technical Field

[0001] This invention relates to the field of automated winding equipment control technology, and in particular to a multi-axis synchronous tension control method for a vertical winding machine. Background Technology

[0002] In the coil winding process, the stability of the winding tension is a key factor determining the coil forming quality, dimensional accuracy, and electrical performance. For vertical winding machines, ensuring that the tension is maintained within the allowable range during high-speed winding is fundamental to achieving high-quality production.

[0003] Existing vertical winding machines generally adopt a single-axis independent control architecture, meaning that each winding axis is equipped with its own tension control system. Under this architecture, each axis performs closed-loop adjustment based solely on its own sensor feedback, and the axes do not interfere with each other. This control method has a relatively simple structure and can basically meet production needs when handling single-axis winding tasks or simple winding tasks where consistency requirements are not high.

[0004] However, with increasing demands for efficiency and consistency in industrial production, multi-axis parallel synchronous operation has become the primary working mode for vertical winding machines. Under existing independent control systems, the lack of a unified coordination mechanism and state interaction between the winding axes prevents the system from compensating for dynamic differences during multi-axis operation. When the equipment performs multi-axis synchronous winding, the tension between axes often fluctuates and deviates to varying degrees due to slight differences in the mechanical characteristics of each axis or external disturbances. This discrete control method struggles to guarantee consistent tension across all axes in mass production, resulting in inconsistent coil quality and consequently affecting the overall production pass rate and operational efficiency, making it difficult to meet the demands of high-precision, high-consistency modern manufacturing. Summary of the Invention

[0005] The purpose of this invention is to provide a multi-axis synchronous tension control method for vertical winding machines, which solves the problems of poor multi-axis tension consistency and low dynamic control accuracy caused by the nonlinear increase of the winding diameter and the drastic change of the load rotational inertia in multi-axis synchronous operation of vertical winding machines.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a multi-axis synchronous tension control method for a vertical winding machine. This method is based on a main control unit module, a servo drive module, a tension detection module, and a human-machine interface module, and solves the tension fluctuation problem caused by nonlinear changes in winding diameter and increased rotational inertia during multi-axis winding.

[0007] The method first performs initialization parameter configuration. The main control unit module reads the basic physical parameters of the winding process, including wire diameter, effective bobbin width, and initial bobbin diameter, and uses these parameters as static constants for subsequent calculations. During system operation, the servo drive module drives the servo motor unit to rotate and acquires the mechanical position of the motor shaft at microsecond intervals using a built-in high-resolution absolute encoder. The servo drive module packages the acquired position data into cumulative rotation angles and synchronously transmits them to the main control unit module via a bus communication protocol.

[0008] After acquiring real-time motion data, this invention performs real-time roll diameter estimation based on a soft-sensor model. This step employs a layer-accumulation algorithm to simulate the roll diameter growth process as a linear step function that varies with the number of layers. The system first normalizes the cumulative rotation angle, and then, combining the effective width of the skeleton and the wire diameter, calculates the ratio of the total wire length to the maximum capacity of a single layer. During this process, the system applies floor logic to filter out non-integer layers, ensuring that the layer count is only updated stepwise when the wiring mechanism completes a single layer of filling and folds back to the previous layer. This achieves discretization of the physical folding and stacking behavior of the wire at the algorithm level. The system multiplies the calculated integer value of the layer number by twice the wire diameter thickness of a single layer, and then adds it to the initial diameter of the skeleton to obtain the estimated physical roll diameter at the current moment.

[0009] Based on the estimated roll diameter data, this invention further performs multi-axis consistency verification and anomaly monitoring. The main control unit module iterates through and reads the estimated roll diameter of all axes in operation, calculates the average roll diameter of the group using the arithmetic mean method, and defines this average as the dynamic consistency benchmark for the current production batch. The system calculates the absolute value of the difference between the single-axis roll diameter and the average roll diameter of the group, and monitors whether this difference exceeds the allowable deviation threshold. When the deviation is within the allowable range, the system determines that the winding status is normal; once the difference exceeds the threshold, the system immediately locks the abnormal axis, generates a stop interruption request and sends an emergency braking command, while outputting a fault code and the corresponding axis number.

[0010] To overcome the control challenges caused by the time-varying system model parameters during winding, this invention implements adaptive adjustment of control parameters based on the winding diameter state. The physical basis of this adjustment mechanism lies in constructing a variable-gain adaptive control law to counteract the gain decay trend of the controlled object. During vertical winding, the load lever arm increases proportionally with the winding diameter, while the load moment of inertia increases proportionally with the fourth power of the winding diameter, causing the open-loop gain of the system to exhibit a nonlinear decay as the winding diameter increases. To maintain the consistency of the closed-loop system's pole distribution and dynamic response performance, the system calculates the ratio of the estimated physical winding diameter to the initial diameter of the skeleton at the current moment, using it as a gain correction coefficient. Using this coefficient, the system performs a linear multiplication of the basic proportional gain, basic integral gain, and basic derivative gain stored internally in the controller, generating real-time control parameters adapted to the current high-inertia operating condition.

[0011] Finally, the main control unit module uses the updated control parameter set, combined with the deviation between the real-time tension feedback value collected by the tension detection module and the target tension value, to calculate the torque compensation required to eliminate the deviation through a discrete PID control algorithm. This torque compensation is converted into a current command and sent to the servo drive module. The servo drive module adjusts the inverter circuit output through pulse width modulation technology to control the amplitude and phase of the current flowing into the stator winding of the servo motor, thereby achieving constant tension control across the entire winding diameter range.

[0012] In summary, the present invention has at least one of the following beneficial technical effects:

[0013] 1. This invention employs a layered accumulation algorithm based on motor encoder feedback data for roll diameter estimation. It converts continuous cumulative rotation angles into discrete layer indices through a rounding-down logical operation. This soft measurement method eliminates the hardware cost and installation space requirements of external ranging sensors such as lasers or ultrasonic sensors. Furthermore, the algorithm model directly simulates the stepped stacking behavior of the wire on the frame, ensuring that the roll diameter value is updated only when the physical layer number changes. This effectively filters out measurement noise caused by microscopic unevenness in the wire arrangement, providing the control system with stable roll diameter parameters that conform to actual working conditions.

[0014] 2. This invention constructs a variable gain adaptive control mechanism based on the winding diameter state. It calculates the gain correction coefficient using the real-time estimated winding diameter and dynamically linearly corrects the parameters of the PID controller. This method addresses the physical characteristic that the load rotational inertia of a vertical winding machine increases with the fourth power of the winding diameter. It actively compensates for the nonlinear decay of the system's open-loop gain, thereby ensuring that the closed-loop dynamic response characteristics of the control system remain consistent throughout the entire production cycle from small to large winding diameters, solving the problem of hysteresis in tension response during the large inertia stage.

[0015] 3. This invention implements a multi-axis consistency verification strategy based on the group mean. By calculating the real-time average roll diameter of all working axes as a dynamic benchmark, the deviation of the single-axis roll diameter from this benchmark is monitored. This mechanism utilizes the data redundancy characteristics of multi-axis parallel production, and can effectively identify roll diameter deviations caused by slippage, broken threads, or abnormal wiring on individual axes in the absence of absolute external references. When the deviation exceeds a threshold, it promptly triggers a shutdown protection, preventing batch product scrapping caused by single-point failures. Attached Figure Description

[0016] Figure 1 This is a block diagram illustrating the hardware architecture of the present invention.

[0017] Figure 2 This is a flowchart of the control method of the present invention. Detailed Implementation

[0018] The following is in conjunction with the appendix Figure 1 and attached Figure 2 The present invention will be further described in detail below.

[0019] This invention provides a multi-axis synchronous tension control method for a vertical winding machine, such as... Figure 1 As shown, this method relies on a control system with a specific architecture. The control system mainly includes a main control unit module, a servo drive module, a tension detection module, and a human-machine interface module.

[0020] The main control unit module, serving as the system's computational core, establishes a bidirectional data connection with the servo drive module via an industrial fieldbus. The main control unit module is equipped with a high-speed floating-point processor for executing the winding diameter soft measurement algorithm and variable gain control logic. The servo drive module contains several servo motor units corresponding to different winding axes. Each servo motor unit integrates a high-resolution absolute encoder for acquiring real-time position and speed information of the motor rotor. The tension detection module contains several tension sensor units (such as tension lever potentiometers or thin-film pressure sensors) corresponding to each winding axis. The output of this tension detection module is connected to the analog acquisition interface or high-speed communication interface of the main control unit module for real-time acquisition of actual tension feedback data of the wire during the winding process. The human-machine interface module is connected to the main control unit module for inputting the geometric parameters and initial process values ​​of the winding skeleton.

[0021] like Figure 2 As shown, the specific execution flow of the adaptive tension control implementation method under the above hardware architecture includes steps S1 to S6.

[0022] S1. The system executes initialization parameter configuration. The main control unit module reads the basic physical parameters of the winding process through the human-machine interface module. These basic physical parameters include the wire diameter. Effective width of the skeleton and the initial diameter of the skeleton These parameters are stored in the registers of the main control unit module as static constant inputs for the subsequent roll diameter estimation model. At this time, the system has not yet started the winding action, and each servo motor unit is in a standby locked state.

[0023] S2. The system starts up and collects real-time kinematic data. As the winding operation begins, the servo drive module drives each servo motor unit to rotate along a predetermined trajectory. During this process, the encoder inside each servo motor unit collects its cumulative rotation angle in real time. ,in Indicates the first Index number of each winding shaft. Cumulative rotation angle. It is synchronously transmitted to the main control unit module via industrial fieldbus, serving as the dynamic input variable for the roll diameter observer.

[0024] S3. Perform real-time estimation of the roll diameter based on the soft measurement model. The main control unit module uses the static constants obtained in step S1. And the dynamic variables obtained in step S2 are used to calculate the current time step using a preset cascading accumulation algorithm. Physical volume of each winding spool index number This step does not rely on measurement data from external lasers or contact sensors. Instead, it indirectly obtains the roll diameter value by calculating the volumetric stacking effect of the wire in space, thereby constructing a controlled object model of the system.

[0025] S4. Perform multi-axis consistency verification and anomaly monitoring. Since the vertical winding machine operates with multiple axes working in tandem, the main control unit module aggregates the estimated physical roll diameters of all working axes within each calculation cycle. Calculate the average group volume diameter Subsequently, the system will determine the physical roll diameter of the single axis. with the mean Perform a comparison. If the absolute value of the deviation exceeds the preset threshold... The system determines that the axis has an abnormality of overlapping or broken wiring and generates a shutdown protection command; if the deviation is within the allowable range, the program proceeds to the next step.

[0026] S5. Perform adaptive adjustment of control parameters based on roll diameter status. The main control unit module adjusts the control parameters according to the current estimated roll diameter. relative to the initial diameter of the skeleton The ratio relationship is used to analyze the change in the system's physical gain under the current operating condition. Based on this rate of change, the main control unit module adjusts the proportional gain of the internal PID controller. Integral gain and differential gain Perform linear correction to generate a set of control parameters adapted to the current load arm state.

[0027] S6. Generate and output closed-loop control commands. The main control unit module uses the updated control parameter set from step S5, combined with the deviation between the real-time tension feedback value collected by the tension detection module and the set tension target value and the actual feedback value, to calculate the required torque compensation amount at the current moment. This torque compensation amount is converted into a current command. The signal is sent to the servo drive module via the bus. The servo drive module then adjusts the output torque of the corresponding servo motor unit to maintain constant wire tension. The system then returns to step S2 to begin the next control cycle.

[0028] Step S1 is fundamental to the entire control cycle and aims to establish the static physical model required for roll diameter estimation. This step specifically includes the following sub-steps:

[0029] S101. Human-Machine Interaction and Recipe Loading. After the system is powered on, the main control unit module first establishes communication with the human-machine interface module. The operator selects the corresponding coil specification recipe through the human-machine interface. The main control unit reads the preset basic physical parameters in the recipe, specifically including: wire diameter. Effective width of the skeleton and the initial diameter of the skeleton .

[0030] S102. Parameter validity verification. To prevent input errors from causing subsequent control divergence, the main control unit module performs logical verification on the read parameters. System check. Is it greater than zero? Is it greater than ,as well as Check if it is within the allowable range of the mechanical structure. If the verification fails, the system will trigger an alarm and prevent startup; if the verification passes, proceed to the next step.

[0031] S103, Observer static constant mapping. After verification, the main control unit module will... The system locks the written non-volatile memory area and maps these parameters to static constant inputs for subsequent steps. Simultaneously, the system sends an enable preparation command to the servo drive module, and each servo motor unit enters a position-locked state, ready to begin winding.

[0032] Step S2 is integral to the entire winding process and is responsible for providing kinematic data with high immediacy and synchronization. This step specifically includes the following sub-steps:

[0033] S201, Trajectory Interpolation and Servo Drive. The main control unit module generates target position commands for each axis based on a preset winding and wiring process (such as a precision wiring algorithm), and sends them to the servo drive module via an industrial fieldbus. The servo drive module drives each servo motor unit to rotate according to a specified speed and acceleration curve, thereby driving the frame to perform the winding operation.

[0034] S202, High-frequency encoder sampling. While the servo motor unit rotates, its built-in high-resolution absolute encoder samples the mechanical position of the motor shaft at microsecond intervals. The encoder records the pulse count value relative to zero in real time, reflecting the cumulative physical rotation of the servo motor since startup.

[0035] S203, Bus Synchronization Data Return. The servo drive module packages the position data sampled by each axis through the PDO channel of the bus communication protocol. The system ensures that the data of all axes has the same timestamp attribute and defines this data as cumulative rotation angles. This data is uploaded in real-time to the high-speed input register of the main control unit module. This is the only dynamic input variable for the roll diameter estimation algorithm in subsequent steps.

[0036] This invention utilizes kinematic data from a servo drive system to construct a soft-measurement observer for the coil diameter. Its core logic lies in establishing a deterministic mapping relationship between the motor's rotational speed and the physical stacking state of the wire. During the operation of the vertical winding equipment, the servo motor drives the winding frame to rotate, and the wire, under tension, is tightly wound within the effective winding area of ​​the frame. Unlike the continuous level changes of a fluid medium, the radial stacking of solid wire on the frame exhibits a discrete, stepped characteristic. That is, the wire can only be folded back and stacked onto the next layer after filling the axial width of the current layer, resulting in abrupt changes in the coil diameter value.

[0037] To accurately characterize this physical process, a mathematical model based on discrete layer indexing was established within the main control unit. This model uses the cumulative rotation angle of the servo motor as the input variable, and the geometric parameters of the skeleton and wire specifications as system constants. It derives the current physical roll diameter by calculating the equivalent number of layers of wire in space. Specifically, this physical model assumes that the wire is tightly packed within each layer, ignores the small errors caused by interlayer gaps and wire cross-sectional deformation, and treats the increase in roll diameter as a linear step function that varies with the number of layers.

[0038] Based on the above physical mechanism, the mathematical model expression for volume diameter estimation is defined as follows:

[0039] ;

[0040] in, Indicates the first The estimated physical roll diameter of the axis at the current moment, whose value directly corresponds to the length of the load arm, in millimeters; The base diameter (unwound diameter) of the winding skeleton is the initial radial dimension when no wire is wound around it, in millimeters. This parameter represents the nominal diameter of the wire input from the human-computer interaction interface, in millimeters. It determines the radial increment of the coil diameter after a single layer of winding is completed. Indicates the first The cumulative rotation angle of the axis servo motor from the start of winding (zero point) to the current time is fed back by the servo encoder and processed by the main control unit into a continuously increasing radian value, in radians; This parameter, expressed in millimeters, represents the effective axial width of the bobbin used to accommodate the wire. It limits the maximum number of turns a single coil can hold. Pi (π) is used to perform the geometric transformation between angular displacement and linear displacement.

[0041] In this model, the item Physically, it refers to the cumulative number of revolutions of the motor. Physically, it represents the maximum theoretical number of turns that a single layer of the skeleton can accommodate. The quotient obtained by dividing the two reflects the ratio of the total length of the currently wound wire to the capacity of a single layer. Operator For the floor operation, it accurately reproduces the stacking effect in mathematical logic. That is, no matter how many turns the wire is wound in the current layer, as long as the layer change condition is not met, its contribution to the roll diameter remains unchanged until a layer jump occurs.

[0042] By constructing this physical model, the control system transforms the continuously changing motor angle signal into a stepped-changing roll diameter signal. This processing method not only eliminates the computational noise caused by incomplete layers in traditional analog quantity estimation, but also improves the accuracy of the estimated physical roll diameter. It can accurately reflect the nonlinear characteristics of lever arm changes in the physical system, thereby supporting the discretization switching of variable gain parameters in subsequent control stages and ensuring precise matching of actual tension and torque characteristics. The technology for acquiring and transmitting servo motor data is well-known in this field and will not be elaborated upon here.

[0043] In step S3, the main control unit transforms the physical winding model into a computer-executable operation sequence through specific algorithm logic. This sequence specifically includes the following steps.

[0044] S301. Perform normalization processing of the cumulative rotation amount. The raw position data uploaded by the servo drive module is the encoder pulse count value. The main control unit first converts this pulse count into a standard radian cumulative rotation angle according to the encoder resolution. This step ensures the monotonically increasing property of the location data, that is... The data recorded is the total winding distance of the wire on the skeleton from zero point, rather than the relative position within a single turn. The conversion of encoder pulse signals to angle values ​​involves basic digital signal processing logic, which is well-known in this field and will not be elaborated upon here.

[0045] S302. Calculate the complete layer index of the current winding. The main control unit uses the cumulative rotation angle obtained in step S301. Combined with the pre-stored effective width of the skeleton and wire diameter First, the theoretical layer ratio of the current cable position is calculated. This ratio represents the multiple relationship between the total cable length and the maximum capacity of a single layer. Then, the system performs a floor function on this ratio, filtering out the decimal part, to obtain the integer value of the number of layers that have been fully covered and wound. This process achieves a discretized simulation of the physical behavior of wire folding and stacking at the algorithm level. Specifically, the layer count only jumps when the wire fills a layer and is guided to the next layer by the wiring mechanism. The specific calculation expression is as follows:

[0046] ;

[0047] in, The integer value representing the current complete layer number. This indicates the integer-down operator; the remaining symbol definitions are the same as described above. This step uses the integer-down operation to strip away the transitional row data that has begun to wrap but has not yet filled a layer, ensuring that the level counter is updated only when a physical layer of stacking is completed.

[0048] S303, Generate the final physical volume size estimate. Based on the complete number of layers obtained in step S302. The main control unit performs a linear superposition operation. The system will then calculate the number of layers. Multiply by double the wire diameter thickness of a single layer The total radial increment of the roll diameter is obtained, and this increment is added to the initial diameter of the skeleton. Up, thus synthesizing the current moment's first... Shaft diameter The operational logic for this step is shown in the following formula:

[0049] ;

[0050] After the calculation is completed, the main control unit will output the calculated result. The roll diameter state variable, updated in the internal register, serves as the basis for anisotropic parameters in the tension compensation calculation of the next control cycle, and also as input data for multi-axis consistency verification. Through this step-by-step calculation strategy, the system decouples the nonlinear physical model into three consecutive computational steps: angle normalization, layer discretization, and roll diameter linear synthesis. This reduces the single-step computational load on the controller while ensuring the real-time performance and accuracy of the roll diameter data updates in the time domain.

[0051] Step S4 fully utilizes the cluster characteristics of multi-axis parallel operation of vertical winding machines to achieve online monitoring of winding quality by constructing a logical observation network that does not rely on external physical sensors. As the data foundation for subsequent anomaly detection and protection logic, the collaborative monitoring principle based on group data is implemented through the following sub-steps.

[0052] S401, Constructing and aggregating data for collaborative monitoring of the execution group. The system first defines the key state variables within the current control cycle: setting... The index number of the winding axis that is currently servo enabled and performing a winding task; This represents the total number of axes currently running synchronously in the system. The first step is calculated in real time based on the soft measurement model in step S3. The physical roll diameter of the axis at the current moment.

[0053] The main control unit module uses its internal high-speed bus to traverse and read the register values ​​of all valid working axes at the beginning of each operation cycle, constructing a set of roll diameter states for the current moment. For axes that are in a downtime state due to fault reports or manual blocking, the system will automatically remove them during the data aggregation stage to ensure the validity of the statistical sample.

[0054] Subsequently, the main control unit module calculates the real-time average roll diameter of all working axes based on the aforementioned set. The average diameter of this roll This is defined as the dynamic consistency benchmark for the current production batch. The technical significance of introducing this dynamic benchmark lies in eliminating the impact of systematic deviations on monitoring accuracy. In actual production, the diameter of the wire raw material... There are often batch tolerances. If only an absolute threshold based on the theoretical design value is used for judgment, when the diameter of the entire batch of wires is too large, the estimated roll diameter of all shafts will simultaneously exceed the theoretical range, thus triggering a large number of false alarms.

[0055] By establishing With this relative evaluation system at its core, the system can automatically filter out such common-mode interference. That is, as long as the changing trend of each axis is consistent with the real-time average roll diameter... By maintaining synchronization, the system determines that the batch operation is under control. This approach shifts the monitoring focus from whether the data conforms to theoretical values ​​to whether the inter-axis differences are significant, thus providing adaptive reference coordinates for the anomaly detection algorithm in subsequent steps. The specific implementation of data acquisition and mean calculation can be accomplished using the floating-point unit built into the main control unit, a conventional technique that can be implemented by those skilled in the art through programming.

[0056] S402, Perform logic judgment for single-axis roll diameter deviation. The main control unit module calls the internal comparator logic to determine the deviation of each axis in the working state. Estimated roll diameter of the shaft The average group roll diameter calculated in step S401 Perform the difference calculation. The system determines whether the current axis winding state is within the normal threshold range based on the following inequality:

[0057] ;

[0058] in, Representing the The current real-time estimated roll diameter of the axis; This represents the arithmetic mean of the roll diameters of all working axes at the current moment; This represents the preset allowable deviation threshold. The setting is based on the wire diameter. The multiplier, for example, set to Or 2.0 To tolerate normal mechanical vibration errors.

[0059] When the above inequality holds, the first inequality is determined to be true. The shaft has experienced a physical anomaly. At a physical level, if... Greater than This indicates that the shaft has experienced a cable overlap fault, causing abnormal accumulation of wires at the same location, resulting in a rapid increase in the lever arm. Less than If the value remains unchanged while other axes are growing, it indicates that the axis has experienced a fault such as wire breakage, slippage, or servo encoder data loss.

[0060] S403. Execute abnormal response and shutdown protection. Once the judgment logic output in step S402 is true, the main control unit module immediately locks the index number of the abnormal axis. The system generates a highest-priority shutdown interruption request. It sends an emergency braking command to the servo drive module, cutting off the torque output of the servo motor unit or putting it into position-holding mode to prevent damage to the wiring mechanism due to wire overlap or waste accumulation due to wire breakage. Simultaneously, the main control unit outputs specific fault codes and corresponding axis numbers through the human-machine interface module, prompting operators to handle the situation on-site. The braking control of the servo motor and the alarm display on the human-machine interface are well-known technologies and will not be elaborated upon here.

[0061] In this embodiment, after performing multi-axis consistency verification and before performing adaptive parameter adjustment based on roll diameter, a torque characteristic analysis basis for the variable load object is established in advance. This analysis constitutes the physical basis for the subsequent variable gain control strategy.

[0062] During the continuous winding process, the load driven by the servo motor unit exhibits time-varying characteristics. The controlled system is not a constant linear model, but a nonlinear system whose parameters dynamically evolve over time. Specifically, as the wire is continuously layered on the bobbin, the estimated winding diameter... The continuous increase directly leads to changes in two key physical quantities in the mechanical transfer function: the load arm and the moment of inertia.

[0063] First, regarding the change in load arm. The essence of tension control is to maintain a constant tension on the wire. Based on the lever principle, the servo motor, in order to maintain a constant load arm at a radius of... A constant tangential tension is generated at the winding point, and the electromagnetic torque that must be output is proportional to the current winding diameter. This means that, with the input error signal and control parameters remaining unchanged, as the winding diameter increases, the actual tension effect acting on the wire will change due to the increase in lever arm, causing a factorial shift in the system's open-loop gain.

[0064] Secondly, regarding the change in moment of inertia. The moment of inertia of the winding shaft is a physical quantity that measures an object's ability to resist changes in its rotational motion. During the winding process, as the volume of the formed coil increases, the moment of inertia of the entire rotor system exhibits a high-order nonlinear growth trend relative to the winding diameter (related to the fourth power of the radius). If a traditional PID controller with fixed gain is used, the calibrated parameters are sufficient to maintain stability in the initial no-load stage (small inertia, short lever arm); however, when entering the full-load stage (large inertia, long lever arm), the huge inertia slows down the motor's response to torque commands. If the controller maintains its original low-gain output, it will not be able to provide sufficient acceleration to correct the tension error caused by speed fluctuations, thus leading to increased steady-state error or sluggish dynamic response. Conversely, if high-gain parameters are calibrated according to the full-load condition, excessive control action will be applied to a slight load in the no-load start-up stage, easily causing system overshoot, oscillation, or even wire breakage.

[0065] Based on the analysis of the torque and inertia characteristics described above, this method excludes control schemes with a single fixed parameter. The system logic clarifies the need for a compensation mechanism that can sense the current physical roll diameter in real time and adjust the controller's output strength accordingly. Specifically, this mechanism is constructed as a variable-gain adaptive control law, aiming to offset the gain change trend of the controller with the gain decay trend of the controlled object, thereby maintaining the distribution of the closed-loop system poles essentially unchanged across the entire roll diameter range and ensuring the consistency of tension control performance. This physical model is constructed using the correction coefficients in step S5. The definition and linear scaling transformation of PID parameters provide a theoretical basis and a necessity for implementation.

[0066] After completing multi-axis consistency verification and anomaly monitoring, the system enters step S5, which is the adaptive adjustment stage of control parameters based on the roll diameter state. This stage aims to solve the nonlinear drift problem of the controlled object model caused by changes in roll diameter. Through a gain scheduling strategy, the time-varying physical system is mapped into a linear control loop with constant performance.

[0067] S501, The main control unit module estimates the current physical volume size based on the previous step S3. Calculate the gain correction coefficient for each control axis. This coefficient represents the change in physical scaling of the current operating condition relative to the initial setup of the system. The calculation formula is as follows:

[0068] ;

[0069] in, For the current moment Estimated physical roll diameter of the shaft Let be the initial diameter of the bobbin. Since the load lever arm during winding is proportional to the winding diameter, and the load moment of inertia is proportional to the fourth power of the winding diameter, the open-loop gain of the system essentially exhibits a non-linear decreasing trend with increasing winding diameter. This is addressed by introducing a dimensionless gain correction coefficient. The system quantifies the degree to which the load characteristics deviate from the initial calibration state at the current moment, providing a quantitative benchmark for the dynamic reconfiguration of the subsequent control law.

[0070] S502, the main control unit module performs dynamic mapping and updating of PID control parameters. To achieve accurate response to variable parameter systems, the controller internally has a set of basic PID parameters pre-set under no-load conditions, which are the basic proportional gain. Basic integral gain and fundamental differential gain Within each control cycle, the system utilizes the gain correction coefficient calculated in step S501. Generate the current time-adapted to the following linear transformation rule. Real-time control parameters of the axis:

[0071] ;

[0072] in, , , These are the th times within the current control cycle. The proportional gain, integral gain, and derivative gain actually applied by the axis controller. This evolutionary model is based on the pole placement principle, which makes the controller gain inversely related to the physical gain decay characteristics of the controlled object, i.e., as the roll diameter increases... The increased synchronous boost of the controller's output gain offsets the decrease in torque transmission efficiency caused by the increased winding diameter. This end-to-end parameter adaptive mechanism ensures that the open-loop gain of the closed-loop control system remains relatively constant throughout the entire process cycle, from no-load start-up to full-wind stop, guaranteeing that the system's dynamic response speed and steady-state accuracy do not degrade with the winding process. The specific tuning methods for the PID parameters are well-known in this field and will not be elaborated upon here.

[0073] After completing the adaptive update of the control parameters, the system executes step S6, which is the generation and execution stage of closed-loop control commands. This stage transforms the convolution observation model and variable gain strategy built in the previous steps into specific physical drive signals to complete the torque control of the servo motor unit.

[0074] S601, the main control unit module performs real-time calculation of tension deviation. The system acquires real-time tension values ​​from the aforementioned tension detection module via a high-speed communication interface or analog input channel. The main control unit module compares the preset process target tension value with the real-time feedback value to determine the tension deviation within the current control cycle. This tension deviation It reflects the polarity and amplitude of the error between the actual stress state of the wire at the current moment and the process requirements, and is the direct input of the closed-loop control system.

[0075] S602, the main control unit module performs adaptive PID control law calculation based on the updated parameter set. The main control unit module calls the real-time control parameters corrected by the physical ratio in step S5, i.e., the proportional gain. Integral gain and differential gain The tension deviation obtained in step S601 Discrete PID control is performed. During this process, the controller uses the corrected high-gain parameters to compensate for the decrease in load sensitivity caused by the increase in roll diameter, and calculates the current time step to eliminate tension deviation. The required electromagnetic torque compensation amount is then calculated. Subsequently, the main control unit module converts the calculated torque compensation amount into a corresponding digital or analog current command. This current command It directly corresponds to the current component required by the servo motor and is used to generate an electromagnetic force in the air gap of the motor to resist the load resistance torque.

[0076] S603, the system executes the underlying driver and closed-loop cycle. The main control unit module will encapsulate the current command via the bus protocol. The command is sent to the servo drive module. The current loop controller inside the servo drive module responds to this command by adjusting the output of the inverter circuit using pulse width modulation (PWM) technology. This controls the amplitude and phase of the current flowing into the stator winding of the servo motor unit, thereby precisely outputting the target torque. As the motor torque is adjusted, tension fluctuations on the wire are suppressed, and the system completes one closed-loop adjustment. Subsequently, the control program logic jumps back to step S2, and when the next clock interrupt arrives, the cumulative rotation angle of the servo motor is collected again. The process then proceeds to the next cycle of winding diameter estimation and control until the winding task is completed. The specific vector control algorithm and current loop adjustment mechanism within the servo drive module are well-known technologies and will not be elaborated upon here.

Claims

1. A multi-axis synchronous tension control method for a vertical winding machine, characterized in that, Includes the following steps: S1. System executes initialization parameter configuration: The main control unit module reads the basic physical parameters of the winding process, which serve as the static constant input for the subsequent roll diameter estimation model; S2. The system starts running and collects real-time kinematic data: the servo drive module drives each servo motor unit to rotate and collects their respective cumulative rotation angles, which are synchronously transmitted to the main control unit module. S3. Perform real-time roll diameter estimation based on soft measurement model: The main control unit module uses the static constant and the cumulative rotation angle to calculate the physical roll diameter of the current winding axis index number through the stacked cumulative algorithm. S4. Perform multi-axis consistency verification and anomaly monitoring: The main control unit module collects the estimated roll diameters of all working axes, calculates the average roll diameter of the group through the mean statistical algorithm, and compares the single-axis roll diameter with the average roll diameter of the group. If the deviation is within the allowable range, the program proceeds to the next step. S5. Perform adaptive adjustment of control parameters based on roll diameter status: The main control unit module analyzes the changes in the system physical gain under the current operating condition, and calculates the proportional gain, integral gain and derivative gain of the corrected internal PID controller through the parameter linear correction algorithm. S6. Generate and output closed-loop control commands: The main control unit module uses the updated control parameter group in step S5, combined with the real-time tension feedback value collected by the tension detection module, to calculate the required torque compensation amount through the discrete PID control algorithm, and converts the torque compensation amount into a current command to be sent to the servo drive module. The servo drive module adjusts the output torque accordingly.

2. The multi-axis synchronous tension control method for a vertical winding machine according to claim 1, characterized in that, In step S3, the stacking accumulation algorithm treats the increase in volume diameter as a linear step function that varies with the number of layers. The specific calculation steps include: S301, Perform normalization processing of cumulative rotation: convert the raw position data uploaded by the servo drive module into a standard radian cumulative rotation angle; S302. Calculate the complete layer index of the current winding: Using the accumulated rotation angle, combined with the effective width of the skeleton and the wire diameter read in step S1, calculate the integer value of the number of layers that have been fully covered and filled using the layer index calculation formula containing the logic of rounding down. S303. Generate the final physical roll diameter estimate: Using the integer value of the number of layers, the double wire diameter thickness of a single layer, and the initial diameter of the skeleton, calculate the final physical roll diameter estimate using the roll diameter linear superposition calculation formula.

3. The multi-axis synchronous tension control method for a vertical winding machine according to claim 2, characterized in that, In step S302, the logic for calculating the integer value of the number of layers that have been fully covered and filled is as follows: The ratio calculation formula is used to calculate the multiple relationship between the total length of the wire and the maximum capacity of a single layer. The decimal part is filtered out by the floor function, so that the layer count only jumps when the wire fills a layer and is guided to the next layer by the wiring mechanism. This achieves discretization simulation of the physical behavior of wire folding and stacking at the algorithm level.

4. The multi-axis synchronous tension control method for a vertical winding machine according to claim 1, characterized in that, Step S4 includes the following specific execution process: The system first defines the key state variables within the current control cycle, iterates through and reads the register values ​​of all valid working axes, constructs the roll diameter state set at the current moment, and automatically removes axes that are in a stopped state. Based on the set of roll diameter states, the real-time average roll diameter of all working axes is calculated using the arithmetic mean calculation formula. The real-time average roll diameter is defined as the dynamic consistency benchmark for the current production batch. The absolute value of the difference between the estimated roll diameter of each shaft in operation and the average roll diameter of the group is calculated using the difference calculation formula, and the current winding status of the shaft is determined based on whether the absolute value of the difference exceeds the allowable deviation threshold.

5. The multi-axis synchronous tension control method for a vertical winding machine according to claim 4, characterized in that, Step S4 also includes steps for performing exception response and shutdown protection: When the absolute value of the difference exceeds the allowable deviation threshold, the main control unit module immediately locks the index number of the abnormal axis, generates a stop interruption request, sends an emergency braking command to the servo drive module, and outputs the fault code and the corresponding axis number through the human-machine interface module.

6. The multi-axis synchronous tension control method for a vertical winding machine according to claim 1, characterized in that, In step S5, the physical basis for performing adaptive adjustment of control parameters is: A variable gain adaptive control law is constructed to offset the gain change trend of the controller with the gain decay trend of the controlled object, thereby maintaining the distribution position of the closed-loop system poles basically unchanged over the entire roll diameter range. The gain decay trend of the controlled object refers to the load lever arm increasing proportionally to the roll diameter, and the load rotational inertia increasing proportionally to the fourth power of the roll diameter, resulting in the open-loop gain of the system decreasing nonlinearly with the increase of the roll diameter.

7. The multi-axis synchronous tension control method for a vertical winding machine according to claim 6, characterized in that, Step S5 includes the following specific execution process: S501. Based on the ratio between the estimated current physical roll diameter and the initial skeleton diameter read in step S1, calculate the gain correction coefficient of each control axis using the physical magnification calculation formula. S502. Perform dynamic mapping and updating of PID control parameters. Using the gain correction coefficient, calculate the real-time control parameters adapted to the winding shaft at the current moment using the parameter linear transformation calculation formula for the basic proportional gain, basic integral gain, and basic derivative gain stored in the controller.

8. The multi-axis synchronous tension control method for a vertical winding machine according to claim 1, characterized in that, Step S6 includes the following specific execution process: S601. Perform real-time calculation of tension deviation, compare the process target tension value with the real-time tension feedback value, and calculate the tension deviation in the current control cycle through a deviation comparison algorithm; S602. Call the updated real-time control parameters to process the tension deviation, calculate the electromagnetic torque compensation amount required to eliminate the deviation through the discrete PID control algorithm, and convert the electromagnetic torque compensation amount into the corresponding current command. S603, the servo drive module responds to the current command by adjusting the output of the inverter circuit through pulse width modulation technology, thereby controlling the amplitude and phase of the current flowing into the stator winding of the servo motor unit.

9. A multi-axis synchronous tension control method for a vertical winding machine according to claim 1, characterized in that, In step S2, the specific method for collecting real-time kinematic data is as follows: While the servo motor unit rotates, the built-in high-resolution absolute encoder samples the mechanical position of the motor shaft at microsecond intervals and records the pulse count value relative to zero at the current moment. The servo drive module packages the position data sampled from each axis through the bus communication protocol, defines these data as the cumulative rotation angle, and uploads them to the high-speed input register of the main control unit module in real time.

10. A multi-axis synchronous tension control method for a vertical winding machine according to claim 1, characterized in that, In step S1, the basic physical parameters include: The main control unit module reads the basic physical parameters and performs parameter validity checks, including whether the wire diameter is greater than zero, whether the effective width of the skeleton is greater than the initial diameter of the skeleton, and whether the initial diameter of the skeleton is within the allowable range of the mechanical structure.

Images