Real-time dynamic prediction method and system in spandex fiber winding process

Abstract

The invention provides a real-time dynamic prediction method and system in a spandex fiber winding process. The method comprises the following steps: constructing a state space model, an interlayer extrusion effect model and an air entrainment effect model which take the winding layer number of a spandex spinning cake as a time index; based on the real-time state vector of the spandex fiber winding system, the air entrainment effect model is used for correcting the current equivalent winding radius by constructing the relation between the air layer equivalent thickness and the equivalent winding radius; on the basis of a real-time state vector of a spandex fiber winding system, the interlayer extrusion effect model is used for correcting the current layer density by constructing a nonlinear relation between the spandex spinning cake layer pressure and the layer density; and based on the corrected current layer density and the corrected current equivalent winding radius, utilizing the state space model to predict the state of the next layer and performing error correction to obtain the corrected prediction state of the next layer. The method and the device are used for adapting to strong nonlinearity and long-time-history time-varying characteristics in the winding process, and the response capability to high-frequency interference is improved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of polyurethane production, and in particular to a real-time dynamic prediction method and system in a spandex fiber winding process. BACKGROUND

[0002] In the spandex dry spinning process, the fiber winding speed is generally 200-600 m / min, but can be as high as 1000 m / min. The high-speed winding speed is a core link for determining the quality and production efficiency of the final product. The fundamental automatic control problem of the spandex fiber winding lies in that, in a long-time, strong nonlinear and time-varying process in which the diameter of a bobbin increases from several centimeters to tens of centimeters, two types of properties that are quite different in nature must be simultaneously coped with.

[0003] In industrial practice, it is found that in the spandex fiber winding process, the following specific physical properties and benign expectations exist: during system operation, the system will continuously be subjected to high-frequency interference such as mechanical resonance and external environmental vibration, and the frequency of the interference can be as high as several hundred hertz, so the benign expectation of the control system is to quickly and accurately respond and suppress such disturbances; during system operation, sudden faults such as broken filaments and tangled filaments can occur, so the benign expectation of the control system is to quickly and accurately respond and suppress such disturbances; there are important nonlinear physical factors such as interlayer extrusion and high-speed air entrainment between the filaments on the bobbin, thereby causing model mismatch and severely restricting the improvement space of the control performance.

[0004] Traditional controllers are linear, time-invariant simplified models. Firstly, they cannot adapt to the strong nonlinearity and long-term time-varying characteristics of the winding process: during the winding process, the diameter of the bobbin continuously grows from a few centimeters to dozens of centimeters, which is a typical long-term time-varying process. Meanwhile, the interaction between the layers of the wire and the dynamic changes of the process parameters (such as tension and angular velocity) exhibit strong nonlinearity. However, the traditional simplified model assumes that the system parameters are fixed and the characteristics are linear, which cannot match the complex changes in actual production, resulting in large fluctuations in winding tension, uneven tightness of the bobbin, and serious lack of control accuracy. Secondly, the response ability to high-frequency interference is weak: during the production process, the system is continuously subjected to high-frequency interference such as mechanical resonance (frequency up to hundreds of hertz) and external environmental vibration. Due to the inherent simplification of the structure of the traditional simplified model, it lacks a fast response mechanism and cannot timely capture these high-frequency disturbances or accurately suppress their negative effects on winding tension and bobbin formation, which easily leads to fiber breakage and reduced winding accuracy. Thirdly, it is difficult to respond to sudden failures: during the winding process, sudden failures such as fiber breakage and entangled wire occur, which require the control system to respond quickly to reduce losses. However, the fixed control logic of the traditional model lacks flexibility and cannot quickly and accurately respond to such sudden situations, often resulting in long downtime for troubleshooting and a reduction in production efficiency of more than 30%. Finally, the traditional model does not consider two key physical phenomena in the winding process: the extrusion effect between the layers of the bobbin: the outer wire compacts the inner wire, directly affecting the density and tension distribution of the bobbin; the air entrainment effect during high-speed winding: an air layer is formed between the wire layers and the guide wheel, changing the actual winding radius. Ignoring these two factors will cause a serious mismatch between the model and the actual working conditions, with a density calculation deviation of the bobbin of more than 5%, significantly limiting the improvement of control performance. SUMMARY

[0005] To solve the technical problems that the existing method cannot adapt to the strong nonlinearity and long-term time-varying characteristics of the winding process and has weak response ability to high-frequency interference, the present application provides a real-time dynamic prediction method in the process of spandex fiber winding, which can quickly and accurately respond to interference such as mechanical resonance and external environmental vibration in the process of spandex fiber winding, as well as sudden failures such as fiber breakage and entangled wire.

[0006] To achieve the above-mentioned purpose, the technical scheme of the present application is as follows:

[0007] A real-time dynamic prediction method in the process of spandex fiber winding, comprising the steps of:

[0008] S1: constructing a state space model indexed by the number of winding layers of the spandex bobbin, an interlayer extrusion effect model, and an air entrainment effect model;

[0009] S2: based on the real-time state vector of the spandex fiber winding system, using the air entrainment effect model to correct the current equivalent winding radius by constructing the relationship between the equivalent thickness of the air layer and the equivalent winding radius;

[0010] S3: based on the real-time state vector of the spandex fiber winding system, using the interlayer extrusion effect model to correct the current layer density by constructing the nonlinear relationship between the spandex cake layer pressure and the layer density;

[0011] S4: based on the corrected current layer density and the corrected current equivalent winding radius, using the state space model to predict the next layer state and correct the error to obtain the corrected next layer predicted state.

[0012] Further, the state space model indexed by the winding layer number of the spandex cake is:

[0013]

[0014] Wherein, k represents discrete time, corresponding to the winding layer of the spandex cake, represents the state vector of the spandex fiber winding system at time k; represents the control input applied by the spandex fiber winding system at time k; represents the process noise or disturbance, represents the state transition function of the spandex fiber winding process, is the observation noise, is the observation function, is the system predicted output state.

[0015] Further, the state vector includes winding tension, equivalent winding radius of the cake with air entrainment effect, cake density, winding motor angular velocity, guide wire motor position; control input includes winding motor torque set value, guide wire motor speed set value.

[0016] Further, the air entrainment effect model is:

[0017]

[0018]

[0019] Wherein, is the equivalent thickness of the air layer, is the equivalent winding radius, is the real-time detection radius, is the air viscosity, is the fiber linear velocity, is the winding motor angular velocity, is the guide wheel radius, is the single layer filament thickness, is an empirical coefficient.

[0020] Further, the interlayer extrusion effect model is:

[0021]

[0022]

[0023] Further, wherein, is the current layer pressure of the spool of elastane filament, is the current layer tension of the spool of elastane filament, is a coefficient of the radial component of tension, is the current layer winding radius of the spool of elastane filament, represented by the equivalent is the real-time detected density of the previous layer, is related to the modulus of elasticity of the elastane.

[0024] Further, the method for correcting the current equivalent winding radius by constructing the relationship between the equivalent thickness of the air layer and the equivalent winding radius using the air entrainment effect model is:

[0025] inputting a winding motor torque set value , a godet motor speed set value , obtaining a current layer state vector of the current elastane fiber winding system , wherein, is the current layer tension of the spool of elastane filament, is the angular velocity of the current layer winding motor, is the position of the godet motor of the current layer;

[0026] obtaining the current layer tension of the spool of elastane filament and the angular velocity of the current layer winding motor , the air viscosity , the guide wheel radius , the empirical coefficient , the single layer filament thickness , calculating the corrected current layer equivalent winding radius according to the calculation formula of the air entrainment effect model.

[0027] Further, the method for correcting the current layer density by constructing the nonlinear relationship between the layer pressure and the layer density of the spool of elastane filament using the interlayer extrusion effect model is:

[0028] obtaining the current layer tension of the spool of elastane filament and the corrected current layer equivalent winding radius of the spool of elastane filament , and the width of the spool of elastane filament , the coefficient of the radial component of tension​ Hertz contact coefficient calculating the corrected current layer density according to a calculation formula of the interlayer extrusion effect model .

[0029] Further, predicting the next layer state using the state space model comprises: obtaining a corrected current layer state vector of the spandex fiber winding system as an actual output state , calculating a current predicted output state , and calculating a difference between the current predicted output state and the actual output state as process noise feedback to a state transition function formula of the state space model , and obtaining a current control input , updating the predicted next layer state .

[0030] Further, the error correction comprises: correcting the error using a calculation formula of the state space model , and outputting a predicted state .

[0031] A real-time dynamic prediction system in a spandex fiber winding process comprises:

[0032] An obtaining module is configured to obtain a real-time state of the spandex fiber winding process.

[0033] A prediction processing module comprises:

[0034] A state space model with the number of winding layers of the spandex cake as a time index is configured to predict a next layer state using the state space model based on a corrected current layer density and a corrected current equivalent winding radius, and correct the error to obtain a corrected next layer predicted state.

[0035] An interlayer extrusion effect model is configured to correct the current layer density by constructing a nonlinear relationship between the layer pressure and the layer density of the spandex cake using the interlayer extrusion effect model.

[0036] An air entrainment effect model is configured to correct the current equivalent winding radius by constructing a relationship between the equivalent thickness of the air layer and the equivalent winding radius using the air entrainment effect model.

[0037] The present application has the following advantages:

[0038] By using the discrete time index and the multi-model coupling means, the long-term time-varying process of the cake diameter from a few centimeters to a few tens of centimeters is accurately adapted, and the strong nonlinear physical effects such as interlayer extrusion and air entrainment are completed, and the problem of insufficient control accuracy caused by the linear time-invariant assumption of the traditional model is solved.

[0039] By means of real-time feedback of process noise and full-dimensional state vector, high-frequency interference (frequency up to hundreds of hertz) such as mechanical resonance and external vibration is captured in real time, and the winding state is quickly corrected, reducing the response time of the system to interference, effectively suppressing the influence of interference on winding tension, reducing the filament breakage rate, and solving the problem that the traditional model cannot cope with high-frequency interference.

[0040] By means of full-dimensional state output and dynamic correction of process noise, a second-level response to sudden faults such as filament breakage and filament entanglement is realized.

[0041] By means of interlayer extrusion effect model (Hertz contact theory) and aerodynamics model (dynamic pressure lubrication theory), the density change caused by interlayer extrusion of the filament cake and the equivalent radius change caused by high-speed winding air layer are quantified, solving the problem of large calculation deviation of filament cake density and radius caused by ignoring interlayer extrusion and air entrainment effect in the traditional model. BRIEF DESCRIPTION OF DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings described below are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative labor.

[0043] Figure 1 The method flowchart of the present application.

[0044] Figure 2 The structural schematic diagram of the present application. DETAILED DESCRIPTION

[0045] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0046] The spandex fiber winding system generally comprises: a driving execution module, mainly a winding motor, which provides winding power to drive the rotation of the bobbin; a tension control module, mainly a tension sensor and a tension regulator, which monitors and stabilizes the winding tension in real time; a guide wire module, mainly a guide wire motor, which controls the uniform arrangement of the fiber in the axial direction of the bobbin; a bobbin bearing module, mainly a bobbin spindle, which is used to fix the spandex bobbin and ensure the coaxiality of the bobbin during winding; a detection sensor module, mainly a radius sensor, a tension sensor and a density sensor, which collects key parameters such as the geometric radius (calculates the equivalent radius), tension and layer density of the bobbin in real time to provide raw data for model calculation; and a control hub module, including an industrial controller and an algorithm module, which comprises three models as follows.

[0047] Embodiment 1

[0048] A real-time dynamic prediction method in the spandex fiber winding process, as shown in Figure 1 The method comprises the following steps:

[0049] S1: constructing a state space model indexed by the winding layer number of the spandex bobbin, an interlayer extrusion effect model and an air entrainment effect model.

[0050] In the embodiment, the state space model indexed by the winding layer number of the spandex bobbin is mainly used to describe the dynamic changes of the winding system and is the core of the prediction. The state space model is used to describe the winding process of the spandex fiber, and has the advantages of strong universality and expression ability, can describe nonlinear, time-varying and multivariable systems. At the same time, it can be explicitly controlled, and is convenient for describing the controlled system,

[0051] The calculation formula of the state space model is as follows:

[0052]

[0053] Wherein, k represents the discrete time, corresponding to the winding layer of the spandex bobbin, represents the state vector of the spandex fiber winding system at time k, which specifically includes the winding tension, the equivalent winding radius of the bobbin (including the air entrainment effect), the current layer density of the bobbin, the angular velocity of the winding motor and the position of the guide wire motor; represents the control input applied to the spandex fiber winding system at time k; specifically, the torque set value of the winding motor and the speed set value of the guide wire motor; represents the process noise or disturbance, such as high-frequency random interference inside / outside the system, represents the state transition function of the spandex fiber winding process, which is based on the mechanical and kinematic mechanism of the system, and maps the current state, control input and process disturbance into the mathematical relationship of the next time state, to observe noise, such as random errors in sensor measurements, such as tension sensor errors, to observe functions, to predict the output state of the system.

[0054] The following five state transition functions are given in the embodiments of the present application as examples, but different state transition functions can be selected according to specific application scenarios in actual applications, which are not limited to the state transition functions of the present application, and the observation function is taken as an example of direct addition with observation noise.

[0055] State transition function of winding tension:

[0056]

[0057] where B is the motor damping coefficient, is the rotational inertia of the winding motor, is the discrete sampling period.

[0058] State transition function of the equivalent winding radius of the bobbin:

[0059]

[0060] where d is the diameter of the spandex fiber.

[0061] State transition function of the current layer density of the bobbin:

[0062] .

[0063] State transition function of the angular velocity of the winding motor:

[0064] .

[0065] State transition function of the guide wire motor position:

[0066] .

[0067] In the embodiments of the present application, an air entrainment effect model based on the dynamic pressure lubrication theory is constructed, which is mainly used to calculate the air layer thickness of high-speed winding and correct the equivalent winding radius parameter of the state space model.

[0068] The calculation formula of the air entrainment effect model includes:

[0069]

[0070]

[0071] where, is the equivalent thickness of the air layer, is the equivalent winding radius, for real-time detection of the radius, for air viscosity, for fiber linear velocity, for winding motor angular velocity, for guide wheel radius, for single-layer filament thickness, for an empirical coefficient, calibrated by experiments.

[0072] In the embodiments of the present application, based on the fact that the spool of spandex fibers is not a rigid body, the inner layer of filament lines will be compacted under the great pressure from the outer layer of filament lines. Therefore, based on the Hertz contact theory, an interlayer extrusion effect model is established, which is mainly used to calculate the density change caused by the interlayer extrusion of the spool, and correct the layer density parameters of the state space model.

[0073] The calculation formula of the interlayer extrusion effect model includes:

[0074]

[0075]

[0076] wherein, is the current layer pressure of the spool of spandex fibers, is the current layer tension of the spool of spandex fibers, is a tension radial component coefficient, i.e., the angle between the tension and the radial direction, is the current layer winding radius of the spool of spandex fibers, which is equivalent to represents, is the real-time detected density of the previous layer, is the width of the spool of spandex fibers, is the current layer density of the spool of spandex fibers, is the Hertz contact coefficient, which is related to the elastic modulus of spandex.

[0077] S2: based on the current layer real-time state vector of the spandex fiber winding system, the current equivalent winding radius is corrected by constructing the relationship between the equivalent thickness and the equivalent winding radius of the air layer by using the air entrainment effect model.

[0078] In the embodiments of the present application, first, the current control input is obtained, wherein, represents the winding motor torque set value, which is set by the motor controller and determines the tension of the winding, represents the guide motor speed set value, which is set by the guide module and determines the arrangement density of the fiber on the spool; T represents transposition.

[0079] Further, the current layer state vector of the current spandex fiber winding system is obtained wherein, is the current layer tension of the spool of spandex fibers, the current layer winding motor angular velocity, the current layer guide wire motor position (i.e. the rotation angle of the motor rotor).

[0080] Further, the current layer tension of the spandex cake and the current layer winding motor angular velocity , the air viscosity , the guide wheel radius , the empirical coefficient , the single layer wire thickness , the corrected current layer equivalent winding radius is calculated according to the calculation formula of the air entrainment effect model .

[0081] S3: Based on the current layer real-time state vector of the spandex fiber winding system, the current layer density is corrected by constructing the nonlinear relationship between the layer pressure and the layer density of the spandex cake using the interlayer extrusion effect model.

[0082] In the embodiment of the application, the current layer tension of the spandex cake and the corrected current layer equivalent winding radius of the spandex cake , and the spandex cake width , the tension radial component coefficient , the Hertz contact coefficient , the corrected current layer density is calculated according to the calculation formula of the interlayer extrusion effect model .

[0083] S4: Based on the corrected current layer density and the corrected current equivalent winding radius, the next layer state is predicted and error correction is performed using the state space model to obtain the corrected next layer predicted state.

[0084] In the embodiment of the application, the state space model is used to predict the next layer state, which includes: obtaining the corrected current layer state vector of the spandex fiber winding system as the actual output state , calculating the difference between the current predicted output state and the actual output state as the process noise feedback to the state transition function formula of the state space model , and obtaining the current control input , updating the predicted next layer state .

[0085] Error correction includes: using the calculation formula of the state space model to perform error correction, and outputting the predicted state .

[0086] Embodiment 2

[0087] A real-time dynamic prediction system in a spandex fiber winding process, comprising:

[0088] an acquisition module, configured to acquire a real-time state of the spandex fiber winding process;

[0089] a prediction processing module, comprising:

[0090] a state space model with the number of wound layers of the spandex cake as a time index, configured to predict a next layer state and perform error correction by using the state space model based on a corrected current layer density and a corrected current equivalent winding radius, to obtain a corrected next layer predicted state;

[0091] an interlayer extrusion effect model, configured to correct the current layer density by constructing a nonlinear relationship between the layer pressure and the layer density of the spandex cake by using the interlayer extrusion effect model;

[0092] an air entrainment effect model, configured to correct the current equivalent winding radius by constructing a relationship between the equivalent thickness of the air layer and the equivalent winding radius by using the air entrainment effect model.

[0093] The foregoing describes the preferred embodiments of the present application in detail. It should be understood that those skilled in the art can make various modifications and changes without creative labor based on the concept of the present application. Therefore, any technical solution obtained by logical analysis, reasoning or limited experiment based on the prior art according to the concept of the present application shall be within the protection scope defined by the claims.

Claims

1. A real-time dynamic prediction method for the winding process of spandex fibers, characterized in that, Including the following steps: S1: Construct a state-space model, an interlayer extrusion effect model, and an air entrainment effect model with the number of winding layers of spandex yarn cake as the time index; S2: Based on the real-time state vector of the spandex fiber winding system, the current equivalent winding radius is corrected by constructing the relationship between the equivalent thickness of the air layer and the equivalent winding radius using the air entrainment effect model. S3: Based on the real-time state vector of the spandex fiber winding system, the current layer density is corrected by constructing a nonlinear relationship between the layer pressure and layer density of the spandex filament cake using the interlayer extrusion effect model; S4: Based on the corrected current layer density and the corrected current equivalent winding radius, the state space model is used to predict the state of the next layer and perform error correction to obtain the corrected predicted state of the next layer.

2. The real-time dynamic prediction method for the spandex fiber winding process according to claim 1, characterized in that, The state-space model with the number of layers of spandex yarn as the time index is as follows: Where k represents discrete time, corresponding to the winding layer of the spandex yarn cake. This represents the state vector of the spandex fiber winding system at time k; This represents the control input applied to the spandex fiber winding system at time k; This indicates process noise or disturbance. This represents the state transition function during the winding process of spandex fibers. To observe the noise, For the observation function, Predict the output state for the system.

3. The real-time dynamic prediction method for the spandex fiber winding process according to claim 2, characterized in that, State vector Including winding tension, equivalent winding radius of the yarn cake with air entrainment effect, yarn cake density, winding motor angular velocity, and guide motor position; control inputs. This includes the torque setting value of the winding motor and the speed setting value of the guide motor.

4. The real-time dynamic prediction method for the spandex fiber winding process according to any one of claims 1-3, characterized in that, The air entrainment effect model is as follows: in, The equivalent thickness of the air layer. For the equivalent winding radius, For real-time detection radius, For air viscosity, For fiber linear velocity, The angular velocity of the winding motor. Where is the radius of the guide wheel. For single-layer filament thickness, This is an empirical coefficient.

5. The real-time dynamic prediction method for the spandex fiber winding process according to any one of claims 4, characterized in that, The interlayer squeezing effect model is as follows: in, For the current layer pressure of the spandex filament cake, The tension of the current layer of the spandex filament cake, The radial component coefficient of the tension. The current layer winding radius of the spandex filament cake is given by the equivalent express, This represents the real-time detection density of the previous layer. It is related to the elastic modulus of spandex.

6. The real-time dynamic prediction method for the spandex fiber winding process according to claim 5, characterized in that, The method for correcting the current equivalent winding radius by constructing the relationship between the equivalent thickness of the air layer and the equivalent winding radius using the air entrainment effect model is as follows: Input winding motor torque setting value Guide wire motor speed setting value Obtain the current layer state vector of the current spandex fiber winding system. ,in, The tension of the current layer of the spandex filament cake, The angular velocity of the current layer winding motor. This indicates the current position of the guide wire motor. Get the current layer tension of the spandex yarn cake and the current layer winding motor angular velocity air viscosity Guide wheel radius empirical coefficient Single layer filament thickness The corrected equivalent winding radius of the current layer is calculated according to the calculation formula of the air entrainment effect model. .

7. The real-time dynamic prediction method for the spandex fiber winding process according to claim 6, characterized in that, The method for correcting the current layer density by constructing a nonlinear relationship between the layer pressure and layer density of spandex filament cake using the interlayer extrusion effect model is as follows: Get the current layer tension of the spandex yarn cake The equivalent winding radius of the current layer, corrected for the spandex filament cake. and the width of the spandex yarn cake Tension radial component coefficient Hertzian contact coefficient The corrected current layer density is calculated based on the calculation formula of the interlayer squeezing effect model. .

8. The real-time dynamic prediction method for the spandex fiber winding process according to claim 7, characterized in that, Predicting the next layer state using the state-space model includes: obtaining the corrected current layer state vector of the spandex fiber winding system. As the actual output state Calculate the current predicted output state Compared with the actual output state The difference is fed back as process noise to the state transition function formula of the state space model. and obtain the current control input. Update the prediction of the next layer state. .

9. The real-time dynamic prediction method for the spandex fiber winding process according to claim 8, characterized in that, Error correction includes: using the calculation formula of the state-space model. Perform error correction and output the predicted state. .

10. A real-time dynamic prediction system for the spandex fiber winding process, employing the real-time dynamic prediction method for the spandex fiber winding process as described in any one of claims 1-9, characterized in that, include: The acquisition module is used to acquire the real-time status of the spandex fiber winding process; The prediction processing module includes: A state-space model with the number of winding layers of spandex yarn cake as the time index is used to predict the state of the next layer and perform error correction based on the corrected current layer density and the corrected current equivalent winding radius, so as to obtain the corrected predicted state of the next layer. The interlayer squeezing effect model is used to correct the current layer density by constructing a nonlinear relationship between the layer pressure and layer density of the spandex filament cake. An air entrainment effect model is used to correct the current equivalent winding radius by constructing the relationship between the equivalent thickness of the air layer and the equivalent winding radius.

Images