Intelligent tension regulation and control system of multi-station self-adaptive winding machine

Abstract

The invention relates to the technical field of industrial automation control, and discloses a tension intelligent regulation and control system of a multi-station self-adaptive winding machine, which comprises an operation data acquisition and inertia identification module used for acquiring operation data of a winding motor in real time; the friction dynamics state observation module is used for receiving the mechanical angular velocity; the impedance parameter dynamic reconstruction module is used for receiving the microcosmic elastic deformation state quantity and the change rate of the microcosmic elastic deformation state quantity; and the variable impedance compliant control module is used for constructing an impedance control law based on the target stiffness coefficient and the target damping coefficient, and generating a final motor torque instruction in combination with the total rotational inertia of the system and the total friction torque estimation value. According to the method, the friction microcosmic deformation state is mapped into the impedance control parameter, parasitic rigidity is stripped in real time, dynamic damping is injected, nonlinear friction interference is effectively restrained, and the tension control precision and stability in the multi-station winding process are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial automation control, in particular to a tension intelligent regulation system of a multi-station self-adaptive winding machine. BACKGROUND

[0002] In the production and manufacturing process of precision materials such as lithium battery separators, optical films and metal foils, a multi-station winding machine is a key equipment in the post-process. Such equipment needs to complete station switching and roll material connection without stopping, and must maintain the constancy of material tension in the large range of roll diameter from empty roll to full roll. The accuracy of tension control directly determines the end face neatness and internal stress distribution of the finished roll material, and then affects the final good product rate of the material.

[0003] The existing winding tension control scheme usually adopts an open-loop or semi-closed-loop control strategy based on torque mode. The conventional method is to estimate the system inertia by real-time calculation of the roll diameter, and to generate a motor torque instruction by superimposing a linear friction compensation item based on speed. However, the mechanical transmission chain of the multi-station winding machine usually includes complex components such as slip winding shaft, gear reduction box and rotary air joint, and these components have nonlinear friction characteristics, i.e. Stribeck effect, during operation. Especially in the device start-up, low-speed peristalsis or zero-speed roll change stage, the mechanical contact surface is in the critical transition zone of static friction and dynamic friction, and the friction torque will fluctuate dramatically and discontinuously with speed and contact state.

[0004] The traditional control method usually feeds forward compensates the friction force as a macroscopic disturbance quantity related only to speed, ignoring the elastic deformation behavior of the friction contact surface at the microscopic level. This microscopic deformation physically behaves as a kind of energy storage characteristic similar to a spring, and when the control system fails to identify and handle this physical characteristic, the elastic potential energy generated by the friction contact surface will be misjudged by the controller as the tension feedback of the material. This results in the system showing extremely high false stiffness in the low-speed high-friction working condition, i.e. the so-called transmission chain parasitic stiffness.

[0005] In addition, the existing variable parameter impedance control or PID control strategy usually only adjusts the control parameters (such as stiffness coefficient and damping coefficient) according to the roll diameter or target speed by simple table lookup. This parameter adjustment method is disconnected from the actual friction dynamics state. When the mechanical system shows high stiffness characteristics due to static friction "locking", if the controller still maintains a high reference stiffness, the superposition of the two will cause the system to have too large equivalent stiffness, which is easy to cause high-frequency oscillation or cause material breakage at the start-up moment; on the contrary, in the sliding region where the friction force suddenly decreases, if there is a lack of sufficient damping injection, the system will also have response overshoot. SUMMARY

[0006] In view of the deficiencies of the prior art, the present application provides a tension intelligent regulation system of a multi-station adaptive winding machine, which solves the problem of poor tension control stability in low-speed and nonlinear friction working conditions caused by the fact that the existing winding machine tension control system ignores the parasitic stiffness interference caused by the micro elastic deformation of the friction contact surface and the impedance control parameters cannot be adaptively adjusted according to the friction dynamics state.

[0007] To achieve the above object, the present application is implemented by the following technical solutions:

[0008] The present application provides a tension intelligent regulation system of a multi-station adaptive winding machine, which comprises:

[0009] The running data acquisition and inertia identification module is used for real-time acquisition of the running data of the winding motor, and the running data includes the mechanical angular velocity, and the current total rotational inertia of the system is calculated based on the real-time acquired winding diameter.

[0010] The friction dynamics state observation module is used for receiving the mechanical angular velocity, real-time iterative calculation of the micro elastic deformation state quantity and its rate of change of the friction contact surface based on the dynamic friction model containing the internal state variable, and synchronous output of the total friction torque estimation value.

[0011] The impedance parameter dynamic reconstruction module is used for receiving the micro elastic deformation state quantity and its rate of change, mapping the micro elastic deformation state quantity into the parasitic stiffness of the transmission chain to correct the reference stiffness of the system, and mapping the rate of change into the nonlinear damping term to correct the reference damping of the system, so as to output the corrected target stiffness coefficient and target damping coefficient.

[0012] The variable impedance compliance control module is used for simultaneously receiving the total rotational inertia of the system, the total friction torque estimation value, the target stiffness coefficient and the target damping coefficient, constructing the impedance control law based on the target stiffness coefficient and the target damping coefficient, and generating the final motor torque instruction in combination with the total rotational inertia of the system and the total friction torque estimation value.

[0013] Preferably, the running data acquisition and inertia identification module comprises a data synchronous acquisition unit, a basic inertia storage unit and a variable inertia calculation unit. The data synchronous acquisition unit is connected to the driver interface of the winding motor to synchronously read the mechanical angular velocity and the stator current of the motor at a preset sampling frequency, and the real-time winding diameter is obtained through an external diameter measuring sensor or a number of turns cumulative algorithm. The basic inertia storage unit pre-stores the fixed inertia values of the winding motor rotor, the speed reduction mechanism and the winding shaft under no load. The variable inertia calculation unit calculates the variable inertia value generated by the winding of the winding material according to the winding diameter, the density parameter of the winding material and the width parameter of the winding material, adds the fixed inertia value and the variable inertia value to obtain the total rotational inertia of the system.

[0014] In one specific embodiment, the variable inertia calculation unit, in calculating the variable inertia value, performs operation logic based on the hollow cylinder inertia formula, specifically: calculating the difference between the fourth power of the real-time winding roll diameter and the fourth power of the winding core radius, multiplying the difference with the width parameter of the winding material, the density parameter of the winding material, and the constant of the circle, and finally multiplying by a preset proportionality coefficient to obtain the variable inertia value.

[0015] Preferably, the friction dynamics state observation module includes a steady-state friction characteristic calculation unit, a micro-state iterative solution unit, and a friction torque synthesis output unit. The steady-state friction characteristic calculation unit is used to receive the mechanical angular velocity, calculate the Stribeck effect function value describing the change of the friction contact surface with speed in the steady state according to the preset Coulomb friction torque parameter, the maximum static friction torque parameter, and the Stribeck speed parameter. The micro-state iterative solution unit constructs a state differential equation describing the dynamic evolution of the internal state variable, updates the micro-elastic deformation state quantity in real time through a discrete iterative algorithm according to the ratio relationship between the mechanical angular velocity and the Stribeck effect function value, and calculates the change rate of the micro-elastic deformation state quantity. The friction torque synthesis output unit multiplies the micro-elastic deformation state quantity by a preset contact stiffness coefficient to obtain an elastic friction component, multiplies the change rate by a preset micro-damping coefficient to obtain a micro-damping component, multiplies the mechanical angular velocity by a preset macro-viscous friction coefficient to obtain a viscous friction component, and adds the elastic friction component, the micro-damping component, and the viscous friction component to obtain a total friction torque estimation value.

[0016] Further preferably, the micro-state iterative solution unit further includes a state saturation limiting logic. This logic detects the numerical absolute value of the micro-elastic deformation state quantity after each iteration update, and if the numerical absolute value exceeds the physical boundary value determined by the maximum static friction torque parameter divided by the contact stiffness coefficient, the micro-elastic deformation state quantity is forced to be clamped to the physical boundary value.

[0017] Preferably, the impedance parameter dynamic reconstruction module includes a stiffness decoupling correction unit and a damping injection correction unit. The stiffness decoupling correction unit calculates a reference stiffness that decreases with the increase of the winding roll diameter, simultaneously receives the micro-elastic deformation state quantity, and multiplies it by the contact stiffness coefficient to calculate the parasitic stiffness of the transmission chain. By subtracting the gain-adjusted parasitic stiffness from the reference stiffness, the target stiffness coefficient after correction is obtained. The damping injection correction unit calculates a reference damping according to the winding roll diameter, simultaneously receives the change rate of the micro-elastic deformation state quantity, calculates a nonlinear damping term proportional to the absolute value of the change rate, and obtains the target damping coefficient after correction by superimposing the nonlinear damping term on the reference damping.

[0018] In one specific embodiment, the damping injection correction unit introduces a velocity attenuation factor related to the mechanical angular velocity when calculating the nonlinear damping term. The damping injection correction unit constructs a negative exponential function with the absolute value of the mechanical angular velocity as the independent variable as the velocity attenuation factor, and performs a multiplication operation on the absolute value of the change rate of the micro-elastic deformation state quantity of the elastic contact surface, the preset damping injection gain and the velocity attenuation factor to obtain the nonlinear damping term, so that the nonlinear damping term decays exponentially with the increase of the mechanical angular velocity.

[0019] Preferably, the variable impedance compliance control module includes an error calculation unit, an impedance torque generation unit and a comprehensive instruction synthesis unit. The error calculation unit receives the externally input reference position trajectory and reference speed trajectory, and compares them with the measured motor mechanical angle and mechanical angular velocity respectively to obtain the position error and speed error. The impedance torque generation unit multiplies the position error by the target stiffness coefficient, multiplies the speed error by the target damping coefficient, and adds them to obtain the variable impedance feedback torque. The comprehensive instruction synthesis unit calculates the inertia feedforward torque according to the total inertia of the system, calculates the tension reference torque according to the target tension, and sums the inertia feedforward torque, the tension reference torque, the variable impedance feedback torque and the total friction torque estimation value to generate the final motor torque instruction.

[0020] Further preferably, the comprehensive instruction synthesis unit is internally configured with a differential operation logic for differential processing of the reference speed trajectory to obtain the reference angular acceleration. The comprehensive instruction synthesis unit multiplies the received real-time changing total inertia of the system with the reference angular acceleration to obtain the inertia feedforward torque, and multiplies the externally set target tension value with the real-time winding diameter to obtain the tension reference torque.

[0021] The second aspect of the present application provides a tension intelligent control method for a multi-station adaptive winding machine, which comprises the following steps:

[0022] S1, real-time read the mechanical angular velocity of the winding motor by running the data acquisition and inertia identification module, simultaneously acquire the real-time winding diameter, and according to the winding diameter, the density parameter of the winding material and the width parameter of the winding material, solve the current total inertia of the system;

[0023] S2, use the friction dynamics state observation module to receive the mechanical angular velocity, based on the dynamic friction model, iteratively solve the micro-elastic deformation state quantity and its change rate of the elastic contact surface in real time, and synthesize and output the total friction torque estimation value;

[0024] S3, receiving the micro-elastic deformation state quantity and its change rate through the impedance parameter dynamic reconstruction module, calculating the parasitic stiffness of the transmission chain and stripping it from the reference stiffness of the system to obtain the corrected target stiffness coefficient, and calculating the nonlinear damping term and injecting it into the reference damping of the system to obtain the corrected target damping coefficient;

[0025] S4, constructing an impedance control law using a variable impedance compliance control module, calculating position error and velocity error according to the externally input reference trajectory, and combining them with the target stiffness coefficient and the target damping coefficient respectively to generate a variable impedance feedback torque;

[0026] S5, summing the inertia feedforward torque generated by the total rotational inertia of the system, the tension reference torque generated based on the target tension, the variable impedance feedback torque, and the total friction torque estimation value to generate the final motor torque instruction to drive the motor to operate.

[0027] The application provides a tension intelligent regulation and control system of a multi-station adaptive winding machine.

[0028] 1. The application solves the problem of interference of transmission chain parasitic stiffness in tension control by establishing a direct mapping mechanism between friction micro-deformation state and stiffness parameters of the control system. The system uses the LuGre model to calculate the elastic deformation potential energy of the friction contact surface, which is quantified as parasitic stiffness and is stripped from the target stiffness of the controller in real time. This stiffness decoupling strategy enables the motor to distinguish between the real tension generated by material stretching and the false torque generated by the friction engagement of the slip mechanism, effectively suppressing the sudden change in tension caused by the sudden release of friction potential energy during low-speed creep or start-up.

[0029] 2. The application uses the change rate of the micro-state variable to construct a nonlinear damping injection logic, which improves the stability of the system in the Stribeck nonlinear region. When the winding motor crosses zero speed or is in the critical state of static-dynamic friction transition, the system can detect the sharp fluctuations of the micro-deformation and actively increase the virtual damping to absorb the shock energy. This not only suppresses the common "creep" phenomenon in mechanical transmission, but also avoids the defect of system response delay caused by traditional fixed high damping parameters at high speed, achieving dynamic stability in the full speed range.

[0030] 3. The application constructs a parameter-level strong coupling between the observer and the impedance controller, realizing the adaptive reconstruction of the physical characteristics of the winding process. Unlike traditional methods that only use friction observation as feedforward compensation, the solution directly uses the observation results to modify the physical parameters of the impedance control law. This design enables the control system to automatically switch seamlessly between "high stiffness and low damping" and "low stiffness and high damping" characteristics according to the current friction contact state, ensuring the tension control accuracy of the multi-station winding machine under the impact of a large range of roll diameter changes and station switching. BRIEF DESCRIPTION OF DRAWINGS

[0031] Figure 1 is a system architecture diagram of the present application;

[0032] Figure 2 is a method flow diagram of the present application.

[0033] Wherein, 10, running data acquisition and inertia identification module; 20, friction dynamics state observation module; 30, impedance parameter dynamic reconstruction module; 40, variable impedance compliance control module. DETAILED DESCRIPTION

[0034] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the specification of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0035] Reference Figure 1 , Figure 1 is a whole architecture diagram of a tension intelligent control system of a multi-station adaptive winding machine according to an embodiment of the present application. The present application provides a tension intelligent control system of a multi-station adaptive winding machine, which is applied to a winding device including a winding motor, a speed reduction transmission mechanism and a slip winding shaft, and the system runs in a computing platform such as a digital signal processor, a servo driver or a programmable logic controller.

[0036] The tension intelligent control system of the multi-station adaptive winding machine includes a running data acquisition and inertia identification module 10, a friction dynamics state observation module 20, an impedance parameter dynamic reconstruction module 30 and a variable impedance compliance control module 40. The modules interact with each other in real time through internal data bus or memory sharing mechanism.

[0037] The running data acquisition and inertia identification module 10, as the input stage of the system, is physically connected to the encoder feedback interface and current sampling circuit of the winding motor. The running data acquisition and inertia identification module 10 is configured to collect the running state data of the winding motor in real time at a preset control period, and the running state data includes the mechanical angular velocity and stator current of the motor.

[0038] The running data acquisition and inertia identification module 10 is also connected with an external diameter measuring sensor or configured with an internal winding diameter calculation logic for real-time acquisition of the current winding diameter. Based on the winding diameter, the preset material density and width parameters, the running data acquisition and inertia identification module 10 calculates the current total rotational inertia of the winding system in real time and sends the total rotational inertia to the variable impedance compliance control module 40.

[0039] A friction dynamics state observer module 20 is connected to the operation data acquisition and inertia identification module 10 to receive real-time mechanical angular velocity. A nonlinear state observer based on the LuGre model is built in the friction dynamics state observer module 20 to dynamically observe the nonlinear friction characteristics in the slip winding shaft and transmission mechanism.

[0040] The friction dynamics state observer module 20 is configured to solve the micro-elastic deformation state quantity of the friction contact surface and the change rate of the micro-elastic deformation state quantity in real time through an iterative algorithm. At the same time, the friction dynamics state observer module 20 synthesizes and outputs the total friction torque estimation value. The micro-elastic deformation state quantity and the change rate thereof are sent to the impedance parameter dynamic reconstruction module 30, and the total friction torque estimation value is sent to the variable impedance compliant control module 40.

[0041] The impedance parameter dynamic reconstruction module 30 is connected to the friction dynamics state observer module 20. The impedance parameter dynamic reconstruction module 30 is used to receive the micro-elastic deformation state quantity and the change rate thereof, and to correct the impedance parameters of the control system in real time according to the preset mapping logic.

[0042] Specifically, the impedance parameter dynamic reconstruction module 30 maps the received micro-elastic deformation state quantity into the parasitic stiffness of the transmission chain, and deducts it from the reference stiffness of the system, thereby outputting the corrected target stiffness coefficient. At the same time, the impedance parameter dynamic reconstruction module 30 maps the change rate of the received micro-elastic deformation state quantity into the nonlinear damping term, and superimposes it on the reference damping of the system, thereby outputting the corrected target damping coefficient. The target stiffness coefficient and the target damping coefficient are sent to the variable impedance compliant control module 40 in real time.

[0043] The variable impedance compliant control module 40, as the output execution stage of the system, is connected to the operation data acquisition and inertia identification module 10, the friction dynamics state observer module 20, and the impedance parameter dynamic reconstruction module 30, respectively.

[0044] The variable impedance compliant control module 40 is configured to receive the externally given reference position trajectory and reference speed trajectory, and to receive the total rotational inertia of the system, the total friction torque estimation value, the target stiffness coefficient, and the target damping coefficient. The variable impedance compliant control module 40 constructs a variable-parameter impedance control law based on the target stiffness coefficient and the target damping coefficient, and performs feedforward compensation combined with the total rotational inertia and disturbance compensation combined with the total friction torque estimation value, to finally generate a motor torque instruction.

[0045] The motor torque instruction is converted into a voltage or current signal through a driver interface, and drives the winding motor to perform corresponding actions, thereby realizing adaptive compliant control of the coiled material tension in the multi-station winding process.

[0046] Referring to Figure 1 , the running data acquisition and inertia identification module 10 specifically comprises a data synchronous acquisition unit, a basic inertia storage unit, and a variable inertia calculation unit.

[0047] The data synchronous acquisition unit is connected to the underlying data interface of the winding motor driver and is configured to synchronously read the real-time running data of the winding motor at a fixed sampling period (e.g., 1 millisecond). The acquired data objects include the mechanical angular velocity and the stator axial current of the motor.

[0048] For the acquisition of the winding diameter , the embodiment adopts real-time calculation logic based on line speed integration. The data synchronous acquisition unit receives the real-time line speed of the main traction roller, combines the mechanical angular velocity , and calculates the real-time winding diameter through the following logic:

[0049] ;

[0050] In the formula, is the real-time winding diameter at time ; is the real-time line speed; is the mechanical angular velocity; is the mechanical reduction ratio between the motor output shaft and the winding shaft. In the case of equipment startup or shutdown, etc. near zero, the data synchronous acquisition unit automatically switches to the thickness accumulation mode, i.e., updates the winding diameter according to the product of the material thickness and the winding turns : .

[0051] The basic inertia storage unit is configured as a non-volatile memory for presetting the fixed physical parameters of the winding mechanical system. The preset parameters include the basic rotational inertia , the core radius , the material width , and the material density . The basic rotational inertia includes the rotational inertia of the winding motor rotor itself, the rotational inertia of the reduction gear box converted to the motor side, and the rotational inertia of the winding air inflation shaft under no load.

[0052] The variable inertia calculation unit is used to solve the total system rotational inertia The variable inertia calculation unit regards the material winding in the winding process as a hollow cylinder with a growing outer diameter, and according to the rigid body dynamics principle, first calculates the variable moment of inertia of the material winding part :

[0053] ;

[0054] Then, the variable inertia calculation unit superimposes the basic moment of inertia and the variable moment of inertia to obtain the current total moment of inertia of the system :

[0055] ;

[0056] wherein is the total moment of inertia of the system converted to the motor shaft side at the moment ; is the basic moment of inertia; is the constant of pi; is the bulk density of the material being wound; is the width of the material being wound; is the real-time winding diameter at the moment ; is the initial outer radius of the winding core.

[0057] The variable inertia calculation unit performs the above operation once every control period, and sends the updated total moment of inertia of the system to the variable impedance compliance control module 40 in real time as the physical basis for inertia feedforward torque calculation.

[0058] Referring to Figure 1 , the friction dynamics state observation module 20 specifically includes a steady-state friction characteristic calculation unit, a microstate iterative solution unit, and a friction torque synthesis output unit. Based on the LuGre dynamic friction model, the friction dynamics state observation module 20 physically equates the friction contact between the two contact surfaces to the random contact behavior of a large number of micro elastic bristles.

[0059] The steady-state friction characteristic calculation unit is used to quantify the steady-state friction level of the slip mechanism at different rotational speeds. It receives the real-time mechanical angular velocity , and uses a Gaussian exponential function to fit the Stribeck effect curve to calculate the average characteristic function of the micro contact points in the steady state. The calculation formula is as follows:

[0060] ;

[0061] wherein is the Coulomb friction torque; it characterizes the dry friction resistance of the system at high speed steady state; is the maximum static friction torque, which characterizes the maximum static friction force that needs to be overcome at the moment of system startup; is the Stribeck velocity, which characterizes the critical velocity feature of the friction force transition from static friction to Coulomb friction; is the contact stiffness coefficient, which characterizes the stiffness property of the micro-brush elastic deformation; is the real-time mechanical angular velocity.

[0062] The micro-state iterative solving unit is the core of the observer, which is used to reconstruct the internal state variables that cannot be directly measured . The state variables represent the average elastic deformation length of the micro-brush of the friction contact surface in the physical sense. This unit constructs a state observer according to the differential equation of dynamic friction:

[0063] ;

[0064] wherein, is the real-time mechanical angular velocity; is the average characteristic function; is the micro-elastic deformation state variable output by the micro-state iterative solving unit; is the change rate of the micro-elastic deformation state variable.

[0065] In the digital control system, the micro-state iterative solving unit uses the Euler method or the Runge-Kutta method to discretize and solve the above differential equation. With a sampling period as the step size, the micro-elastic deformation state variable at the current time and its change rate are calculated.

[0066] In order to prevent the calculation from diverging due to model parameter mismatching or sensor noise in the numerical integration process, the unit is embedded with state saturation limiting logic. This logic monitors the absolute value of in real time. Physically, the deformation of the brush cannot increase indefinitely, and its upper limit depends on the ratio of the maximum static friction force to the contact stiffness. Therefore, when the calculated exceeds the physical boundary value , the unit forces to be clamped to , ensuring the bounded input and output stability of the observer.

[0067] The friction torque synthesis output unit decomposes the friction torque into elastic force component, micro-damping component and macro-viscous component according to the physical definition of the LuGre model, and synthesizes them. The calculation formula of the total friction torque estimation value is:

[0068] ;

[0069] wherein, is the contact stiffness coefficient; is the micro-damping coefficient; is the macro-viscous friction coefficient, representing the viscous resistance proportional to the speed generated by the bearing lubricating oil film, etc.; is the micro-elastic deformation state quantity output by the micro-state iterative solving unit; is the change rate of the micro-elastic deformation state quantity; is the real-time mechanical angular velocity.

[0070] The friction torque synthesis output unit outputs the calculated to the subsequent control module in real time, for feedforward compensation of the torque loss caused by friction.

[0071] Referring to Figure 1 , the impedance parameter dynamic reconstruction module 30 specifically includes a stiffness decoupling correction unit and a damping injection correction unit. The core function of the impedance parameter dynamic reconstruction module 30 is to map the micro-physical state quantity output by the friction dynamics state observation module 20 to the impedance parameter in the control domain, so as to realize the deep coupling between the physical layer and the control layer.

[0072] The stiffness decoupling correction unit aims to eliminate the false enhancement effect of the elastic energy storage of the friction contact surface on the tension control stiffness. The stiffness decoupling correction unit first calculates the reference stiffness of the system according to the real-time winding diameter . Since the winding process increases the winding diameter, the force arm becomes longer, and in order to maintain the stability of the tension control, the physical stiffness requirement of the system usually decays inversely with the square of the winding diameter. The calculation of the reference stiffness follows the following logic:

[0073] ;

[0074] wherein, and respectively correspond to the preset stiffness boundary values in the empty roll and full roll states; is the core radius; is the reference stiffness under the current winding diameter; is the winding diameter.

[0075] At the same time, the stiffness decoupling correction unit receives the micro-elastic deformation state quantity . At the physical level, the micro-deformation of the friction contact surface behaves like a spring energy storage behavior, which will behave as a "parasitic stiffness" in the closed-loop control, causing the equivalent stiffness of the actual system to be higher than the stiffness set by the controller, and thus causing the tension overshoot. The unit calculates the parasitic stiffness to be stripped through the following formula and output the corrected target stiffness coefficient. :

[0076] ;

[0077] In the formula, This is the reference stiffness for the current roll diameter; This is the contact stiffness coefficient (consistent with the parameters in the observer); It is a dimensionless stiffness decoupling gain coefficient (its value is usually between 0 and 1). Let be the modulus of the microscopic elastic deformation state quantity. If the calculated... If the stiffness is below the minimum safety stiffness allowed by the system, the unit clamps it to the minimum safety stiffness value.

[0078] The damping injection correction unit aims to resolve system oscillations caused by drastic changes in frictional characteristics in low-speed or zero-crossing speed regions. The damping injection correction unit first considers the total rotational inertia of the system. and reference stiffness Calculate the reference damping It is usually set to a certain proportion of the critical damping state to ensure the stability of the system's basic response.

[0079] Based on this, the damping injection correction unit introduces a nonlinear damping injection mechanism based on the rate of change of microstate. When the friction contact surface undergoes drastic changes (i.e., When the friction value is very high, it means that the system is in the transition zone between static and dynamic friction or is experiencing a "creeping" phenomenon. At this time, additional damping is needed to dissipate the oscillation energy.

[0080] To avoid the impact of high damping on the dynamic response during high-speed operation, a velocity attenuation factor is introduced into this unit. Nonlinear damping term. The calculation formula is as follows:

[0081] ;

[0082] The final output target damping coefficient for:

[0083] ;

[0084] In the formula, This is a nonlinear damping term; The target damping coefficient; As the reference stiffness; Injecting gain into the damping is used to adjust the intensity of the nonlinear damping effect; It is the absolute value of the rate of change of the microscopic elastic deformation state quantity, which directly reflects the instability of the friction state; This is the absolute value of the mechanical angular velocity; is the speed decay constant, the velocity bandwidth of nonlinear damping is defined. When the motor speed is much larger than , the exponential term tends to zero, making automatically invalid, and the system reverts to the baseline damping control.

[0085] Through the above logic, the impedance parameter dynamic reconstruction module 30 can adjust the and in real time according to the micro-friction state, providing an adaptive parameter basis for subsequent compliant control.

[0086] Referring to Figure 1 , the variable impedance compliant control module 40 as the execution center of the entire system mainly includes an error calculation unit, an impedance torque generation unit, and a comprehensive instruction synthesis unit. Based on the impedance control theory, the variable impedance compliant control module 40 equivalent position closed-loop control to the mass-spring-damper system at the physical level, and the spring stiffness and damping coefficient of the system are dynamically assigned in real time by the aforementioned module.

[0087] The error calculation unit is responsible for constructing the tracking error space of the system. It receives the reference position trajectory and the reference speed trajectory from the upper motion planner, and receives the real-time motor mechanical angle and the mechanical angular velocity feedback by the encoder. The error calculation unit compares the above signals synchronously in each control period, calculates the position error and the speed error . In order to eliminate the high-frequency interference caused by sensor noise, the unit will be preprocessed by a low-pass filter with adjustable cutoff frequency before outputting the error signal.

[0088] The impedance torque generation unit is the generator of compliant control characteristics. Unlike the fixed gain of traditional PID controller, the impedance torque generation unit receives the target stiffness coefficient and the target damping coefficient output by the impedance parameter dynamic reconstruction module 30 after physical layer correction.

[0089] According to the second-order mechanical impedance model, the unit maps the position error to the elastic restoring torque and the speed error to the damping dissipation torque, and calculates the variable impedance feedback torque :

[0090] ;

[0091] In the formula, is the variable impedance feedback torque; is the target stiffness coefficient; is the position error; Target damping coefficient; Velocity error.

[0092] By introducing dynamic parameters determined by friction state and , the feedback torque can automatically exhibit "soft" characteristics to absorb impact when the friction is large or changes drastically, and exhibit "hard" characteristics to ensure accuracy when running stably.

[0093] The comprehensive command synthesis unit is responsible for integrating feedforward control, feedback control and disturbance compensation to generate the final torque command sent to the motor driver. The comprehensive command synthesis unit includes three parallel calculation branches:

[0094] Inertia feedforward branch: The comprehensive command synthesis unit is internally configured with a differential operation logic that performs first-order discrete differentiation on the input reference speed trajectory to obtain the reference angular acceleration . Combined with the system total moment of inertia real-time sent by the running data acquisition and inertia identification module 10, the inertia feedforward torque is calculated. The calculation logic follows Newton's second law: . This term is used to overcome the acceleration and deceleration inertia of the winding roller and the material itself.

[0095] Tension reference branch: The comprehensive command synthesis unit receives the externally set target tension value (typically set by the process operator, unit: Newton), and combines the real-time winding diameter to calculate the static reference torque required to maintain the current tension. The calculation formula is: .

[0096] Friction compensation branch: Directly receives the total friction torque estimate value output by the friction dynamics state observation module 20 as direct compensation for the system's internal physical loss.

[0097] Finally, the comprehensive command synthesis unit linearly superimposes all the torque components mentioned above to generate the final motor torque command :

[0098] ;

[0099] where, is the static reference torque; is the inertia feedforward torque; is the final torque command output to the motor driver; is the real-time updated system total moment of inertia; Use angular acceleration as a reference. The preset target tension value; For real-time winding diameter; It is a variable impedance feedback torque, which includes stiffness and damping characteristics corrected for friction conditions; This is an estimate of the total frictional torque observed based on the LuGre model.

[0100] Through this synthesis method, the control system achieves a control architecture of "feedforward dominance, feedback fine-tuning, and disturbance decoupling", ensuring that the torque output by the motor can accurately cover the inertial force, target tension, frictional resistance, and the correction force required for dynamic error.

[0101] Reference Figure 2 , Figure 2 This is a flowchart of a tension intelligent control method for a multi-station adaptive winding machine according to an embodiment of the present invention. The present invention provides a tension intelligent control method for a multi-station adaptive winding machine, which is periodically executed by a digital processor and specifically includes the following steps:

[0102] S100: Synchronous acquisition of operating data and real-time identification of system inertia. At the beginning of each control cycle, the system first synchronously reads the mechanical angular velocity and stator current of the winding motor through the underlying drive interface. At the same time, the system obtains the real-time winding diameter through external sensor feedback or internal integration algorithm.

[0103] Based on this, the system executes variable inertia calculation logic. The system treats the roll material as a hollow cylinder with an ever-increasing outer diameter. First, it calculates the difference between the fourth power of the real-time winding diameter and the fourth power of the core radius. Then, it performs a continuous multiplication operation with preset roll material width parameters, roll material density parameters, and the constant pi to obtain the variable inertia of the roll material. The system adds this variable inertia to the pre-stored basic fixed inertia, including the motor rotor and reduction mechanism, to calculate the precise total rotational inertia of the system at the current moment, providing a physical basis for subsequent dynamic feedforward.

[0104] S200, iterative observation of the microscopic state of tribodynamics: The system uses the received mechanical angular velocity to drive the internal LuGre dynamic friction observer. First, based on the current rotational speed and combined with preset Coulomb friction, maximum static friction, and Stribeck velocity parameters, the system calculates the Stribeck effect function value describing the steady-state friction characteristics.

[0105] Subsequently, the system enters the microstate discretization iteration process. According to the ratio of the mechanical angular velocity to the Stribeck effect function value, the system updates the micro-elastic deformation state quantity of the friction contact surface by using the difference equation. After each update, the system immediately performs state saturation detection: if the absolute value of the micro-elastic deformation state quantity exceeds the physical limit determined by the maximum static friction torque and the contact stiffness, the system forcibly limits the state quantity within the physical limit range to prevent numerical calculation divergence.

[0106] The system uses the updated micro-elastic deformation state quantity and its rate of change to multiply the contact stiffness coefficient, the micro-damping coefficient, and the macro-viscous friction coefficient respectively, superimposes the calculated elastic component, micro-damping component, and viscous component, and synthesizes the output total friction torque estimate at the current time.

[0107] S300, impedance parameter dynamic reconstruction based on microstate, this S300 is the core link of the method, aiming at adjusting the characteristics of the controller according to the observed physical state. The system first calculates the reference stiffness which decays with the increase of the winding diameter according to the real-time winding diameter, and the reference damping based on the system inertia.

[0108] Then, the system performs parameter mapping and correction. On the one hand, the system calculates the product of the micro-elastic deformation state quantity and the contact stiffness coefficient to obtain the parasitic stiffness of the transmission chain, and subtracts the parasitic stiffness from the reference stiffness to obtain the corrected target stiffness coefficient, thereby eliminating the false high stiffness phenomenon caused by the friction dead zone. On the other hand, the system calculates the absolute value of the rate of change of the micro-elastic deformation state quantity and multiplies it by a negative exponential decay factor based on the absolute value of the mechanical angular velocity to obtain a nonlinear damping term; the system superimposes the nonlinear damping term on the reference damping to obtain the corrected target damping coefficient, thereby actively increasing the system damping to suppress oscillation when the speed is low or the friction state changes dramatically.

[0109] S400, construction and execution of variable impedance compliance control law, the system receives the reference position trajectory and reference speed trajectory generated by the external motion planning, and compares them with the measured motor mechanical angle and mechanical angular velocity to generate position error and speed error.

[0110] The system calls the corrected target stiffness coefficient and target damping coefficient generated in step S300 to execute the impedance control algorithm. Specifically, the system multiplies the position error by the target stiffness coefficient to obtain the elastic restoring torque, multiplies the speed error by the target damping coefficient to obtain the damping dissipation torque, and adds them to generate a variable impedance feedback torque containing the adaptive characteristics of the physical layer.

[0111] S500, synthesis of the synthesis of the torque command and driving, the system finally executes the synthesis of multi-dimensional torque. The system differentiates the reference speed trajectory to obtain the reference angular acceleration, multiplies it by the total rotational inertia of the system obtained in step S100 to obtain the inertia feedforward torque; at the same time, the set target tension is multiplied by the real-time winding diameter to obtain the tension reference torque.

[0112] The system linearly sums the inertia feedforward torque, the tension reference torque, the variable impedance feedback torque generated in step S400, and the total friction torque estimation value output in step S200. The sum result is sent to the motor driver as the final motor torque command, and the winding motor outputs accurate electromagnetic torque, so as to compensate for system inertia, friction loss and dynamic disturbance while realizing constant tension soft winding control.

Claims

1. A tension intelligent control system for a multi-station adaptive winding machine, characterized in that, include: The operation data acquisition and inertia identification module is used to collect the operation data of the winding motor in real time. The operation data includes mechanical angular velocity, and the current total rotational inertia of the system is calculated based on the real-time acquired winding diameter. The friction dynamics state observation module is used to receive the mechanical angular velocity, and based on the dynamic friction model containing internal state variables, iteratively calculate the micro-elastic deformation state quantities and their rate of change of the friction contact surface in real time, and synchronously output the estimated value of the total friction torque. The impedance parameter dynamic reconstruction module is used to receive the micro-elastic deformation state quantity and its rate of change, map the micro-elastic deformation state quantity to the parasitic stiffness of the transmission chain to correct the reference stiffness of the system, and map the rate of change to a nonlinear damping term to correct the reference damping of the system, thereby outputting the corrected target stiffness coefficient and target damping coefficient. The variable impedance compliant control module is used to simultaneously receive the total rotational inertia of the system, the estimated total frictional torque, the target stiffness coefficient, and the target damping coefficient. Based on the target stiffness coefficient and the target damping coefficient, it constructs an impedance control law and generates the final motor torque command by combining the total rotational inertia of the system and the estimated total frictional torque.

2. The tension intelligent control system for a multi-station adaptive winding machine according to claim 1, characterized in that, The operational data acquisition and inertia identification module includes: The data synchronization acquisition unit is used to connect to the driver interface of the winding motor, synchronously read the mechanical angular velocity and the stator current of the motor at a preset sampling frequency, and obtain the real-time winding roll diameter through an external diameter measuring sensor or a turn count accumulation algorithm. The basic inertia storage unit is used to pre-store the fixed inertia values ​​of the take-up motor rotor, reduction mechanism, and take-up shaft when unloaded. The variable inertia calculation unit is used to calculate the variable inertia value generated by the winding of the roll material based on the winding diameter, the density parameter of the roll material and the width parameter of the roll material, add the fixed inertia value and the variable inertia value to obtain the total rotational inertia of the system, and send the value to the variable impedance compliance control module in real time.

3. The tension intelligent control system for a multi-station adaptive winding machine according to claim 2, characterized in that, When calculating the variable inertia value, the variable inertia calculation unit executes the calculation logic based on the inertia formula of a hollow cylinder. That is, it first calculates the difference between the fourth power of the real-time winding diameter and the fourth power of the core radius, then multiplies the difference with the width parameter of the roll material, the density parameter of the roll material, and the pi constant, and finally multiplies it by a preset scaling factor to obtain the variable inertia value.

4. The tension intelligent control system for a multi-station adaptive winding machine according to claim 1, characterized in that, The tribodynamic state observation module includes: The steady-state friction characteristic calculation unit is used to receive the mechanical angular velocity and calculate the Stribeck effect function value describing the change of the friction contact surface with velocity in steady state, based on the preset Coulomb friction torque parameter, maximum static friction torque parameter and Stribeck velocity parameter. The micro-state iterative solution unit is used to construct the state differential equation describing the dynamic evolution of the internal state variables, and update the micro-elastic deformation state quantity in real time through a discretization iterative algorithm based on the ratio of the mechanical angular velocity to the Stribeck effect function value, and calculate the rate of change of the micro-elastic deformation state quantity. The friction torque synthesis output unit is used to multiply the micro-elastic deformation state quantity by a preset contact stiffness coefficient to obtain the elastic friction component, multiply the rate of change by a preset micro-damping coefficient to obtain the micro-damping component, multiply the mechanical angular velocity by a preset macro-viscous friction coefficient to obtain the viscous friction component, and add the elastic friction component, the micro-damping component, and the viscous friction component to obtain the estimated value of the total friction torque.

5. The tension intelligent control system for a multi-station adaptive winding machine according to claim 4, characterized in that, The micro-state iterative solution unit also includes state saturation limiting logic, which is used to detect the absolute value of the micro-elastic deformation state quantity after each iteration update; if the absolute value exceeds the physical boundary value determined by dividing the maximum static friction torque parameter by the contact stiffness coefficient, the micro-elastic deformation state quantity is forcibly clamped to the physical boundary value to prevent observation divergence caused by model parameter deviation.

6. The tension intelligent control system for a multi-station adaptive winding machine according to claim 1, characterized in that, The impedance parameter dynamic reconstruction module includes: The stiffness decoupling correction unit is used to calculate the reference stiffness that decreases as the winding diameter increases based on the winding diameter, and at the same time receive the micro-elastic deformation state quantity, multiply it by the contact stiffness coefficient to calculate the parasitic stiffness of the transmission chain, and obtain the corrected target stiffness coefficient by subtracting the parasitic stiffness after gain adjustment from the reference stiffness. The damping injection correction unit is used to calculate the reference damping based on the winding diameter, and simultaneously receive the rate of change of the micro-elastic deformation state quantity, calculate the nonlinear damping term that is proportional to the absolute value of the rate of change, and obtain the corrected target damping coefficient by superimposing the nonlinear damping term onto the reference damping.

7. The tension intelligent control system for a multi-station adaptive winding machine according to claim 6, characterized in that, When calculating the nonlinear damping term, the damping injection correction unit introduces a velocity attenuation factor related to the mechanical angular velocity. The damping injection correction unit constructs a negative exponential function with the absolute value of the mechanical angular velocity as the independent variable as the velocity attenuation factor, and performs a multiplication operation on the absolute value of the rate of change of the micro-elastic deformation state quantity, the preset damping injection gain, and the velocity attenuation factor to obtain the nonlinear damping term.

8. The tension intelligent control system for a multi-station adaptive winding machine according to claim 1, characterized in that, The variable impedance compliance control module includes: The error calculation unit is used to receive the externally input reference position trajectory and reference velocity trajectory, and compare them with the measured motor mechanical angle and the mechanical angular velocity to obtain the position error and velocity error. The impedance torque generation unit is used to multiply the position error by the target stiffness coefficient, multiply the velocity error by the target damping coefficient, and add the two together to obtain the variable impedance feedback torque. The integrated command synthesis unit is used to calculate the inertia feedforward torque based on the total rotational inertia of the system, calculate the tension reference torque based on the target tension, and sum the inertia feedforward torque, the tension reference torque, the variable impedance feedback torque, and the estimated total friction torque to generate the final motor torque command.

9. The tension intelligent control system for a multi-station adaptive winding machine according to claim 8, characterized in that, The integrated instruction synthesis unit has differential operation logic inside, which is used to perform differential processing on the reference velocity trajectory to obtain the reference angular acceleration; the integrated instruction synthesis unit multiplies the received real-time changing total rotational inertia of the system with the reference angular acceleration to obtain the inertia feedforward torque, and at the same time multiplies the externally set target tension value with the real-time winding diameter to obtain the tension reference torque.

10. A method for intelligent tension control of a multi-station adaptive winding machine, comprising an intelligent tension control system for a multi-station adaptive winding machine according to any one of claims 1-9, characterized in that, Includes the following steps: S1. The mechanical angular velocity of the winding motor is read in real time by running the data acquisition and inertia identification module, and the real-time winding diameter is obtained. Based on the winding diameter, the density parameter of the roll material and the width parameter of the roll material, the current total rotational inertia of the system is calculated. S2. The mechanical angular velocity is received by the tribodynamic state observation module, and the micro-elastic deformation state quantity and its rate of change of the friction contact surface are calculated in real time based on the dynamic friction model, and the total friction torque is estimated and output. S3. Receive the micro-elastic deformation state quantity and its rate of change through the impedance parameter dynamic reconstruction module, calculate the parasitic stiffness of the transmission chain and separate it from the system's reference stiffness to obtain the corrected target stiffness coefficient, and at the same time calculate the nonlinear damping term and inject it into the system's reference damping to obtain the corrected target damping coefficient. S4. Construct an impedance control law using a variable impedance compliant control module, calculate the position error and velocity error based on the externally input reference trajectory, and combine them with the target stiffness coefficient and the target damping coefficient to generate a variable impedance feedback torque. S5. The inertia feedforward torque generated by the total rotational inertia of the system, the tension reference torque generated based on the target tension, the variable impedance feedback torque, and the estimated total friction torque are summed to generate the final motor torque command to drive the motor.

Images