A control method and system for the motion state of a spindle rod used in covered yarn spinning

Abstract

The present invention relates to the technical field of motion state control, and specifically provides a method and system for controlling the motion state of a spindle for covering yarn spinning. The yarn core tension, motor operation data, and environmental data are collected to construct a covering yarn spinning state vector, and the importance weights of each element in the covering yarn spinning state vector for the tension change are calculated. The covering yarn spinning state vector is weighted according to the importance weights to obtain a weighted state vector; based on the weighted state vector, the predicted yarn core tension at the next sampling moment is obtained through a tension prediction model; the tension error, predicted tension error, and tension error change rate are calculated and input into an adaptive controller to obtain the motor current control amount. The motor voltage control amount is calculated according to the motor current control amount, and the control of the spindle is realized according to the motor voltage control amount, and the parameters of the adaptive controller are adjusted in real time through an online parameter optimization method. The present invention can improve the control accuracy of the spindle for covering yarn spinning.

Description

Technical Field

[0001] The present invention relates to the technical field of motion state control, and specifically provides a method and system for controlling the motion state of a spindle for covered yarn spinning. Background Art

[0002] The textile industry, as an important part of the economy, plays a crucial role in multiple fields such as clothing, household textiles, and industrial fabrics. Among them, the yarn covering process is a core link in textile production, mainly used to enhance the physical properties of yarns, such as improving strength, feel, and abrasion resistance. The quality of yarn covering directly affects the performance and market competitiveness of the final product. Therefore, improving the uniformity and consistency of yarn covering, as well as optimizing production efficiency, has become an important goal in the research and development of textile machinery.

[0003] During the yarn covering process, as a key mechanical component, the control of the motion state of the spindle not only affects the appearance and feel of the covered yarn, but also may affect the overall strength and durability of the yarn, thereby affecting the quality and service life of the final product. Traditional spindle motion control methods mainly rely on the design of mechanical structures and simple control algorithms, and achieve the covering of yarns by adjusting the rotation speed and motion trajectory of the spindle. However, traditional control methods lack the ability of self - adaptation and optimization, which limits the control accuracy in complex and changeable production environments.

[0004] Therefore, a method and system for controlling the motion state of a spindle for covered yarn spinning are proposed. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for controlling the motion state of a spindle for covered yarn spinning. By collecting the core yarn tension, motor operation data, and environmental data to construct a covered yarn spinning state vector, and calculating the importance weights of each element in the covered yarn spinning state vector for tension change, the weighted state vector is obtained by weighting the covered yarn spinning state vector according to the importance weights; based on the weighted state vector, the predicted core yarn tension at the next sampling moment is obtained through a tension prediction model; the tension error, predicted tension error, and tension error change rate are calculated, and the motor current control amount is input to an adaptive controller. The motor voltage control amount is calculated according to the motor current control amount, and the control of the spindle is achieved according to the motor voltage control amount, and the parameters of the adaptive controller are adjusted in real time through an online parameter optimization method.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for controlling the motion state of a spindle for covered yarn spinning, comprising:

[0008] During the core-spun yarn spinning process, the core yarn tension, motor operation data, and environmental data are collected according to the sampling period; the motor operation data includes motor torque, motor speed, motor angular velocity, and motor current; the environmental data includes environmental temperature and environmental humidity;

[0009] Based on the core yarn tension, the motor operation data, and the environmental data, a core-spun yarn spinning state vector is constructed;

[0010] Calculate the importance weights of each element in the core-spun yarn spinning state vector with respect to the tension change to obtain an importance weight vector, and perform weighted processing on the core-spun yarn spinning state vector according to the importance weight vector to obtain a weighted state vector;

[0011] Based on the weighted state vector, the predicted core yarn tension at the next sampling moment is obtained through a tension prediction model;

[0012] Set the target core yarn tension, calculate the tension error at the current sampling moment according to the target core yarn tension and the core yarn tension at the current sampling moment, calculate the predicted tension error according to the predicted core yarn tension and the target core yarn tension, and calculate the tension error change rate according to the tension errors at the current sampling moment and the previous sampling moment;

[0013] Based on the tension error, the predicted tension error, and the tension error change rate, the motor current control amount is obtained through an adaptive controller, the motor voltage control amount is calculated according to the motor current control amount, the control of the spindle is realized according to the motor voltage control amount, and the parameters of the adaptive controller are adjusted in real time through an online parameter optimization method.

[0014] Preferably, the process of performing weighted processing on the core-spun yarn spinning state vector includes:

[0015] Calculate the importance weights of each element in the core-spun yarn spinning state vector with respect to the tension change, and the calculation formula is:

[0016] ;

[0017] Wherein, is the importance weight of the th element, is the natural exponential function, is the basic importance coefficient of the th element, is the number of elements in the core-spun yarn spinning state vector, is the correlation coefficient between the th element and the tension;

[0018] Take The importance weights form the importance weight vector. The weighted state vector is obtained by weighting the covered yarn spinning state vector according to the importance weight vector. The formula is:

[0019] ;

[0020] Among them, is the weighted state vector at the current sampling time , is the importance weight vector, is the covered yarn spinning state vector at the current sampling time , is the Hadamard product symbol.

[0021] Preferably, the process of the tension prediction model includes:

[0022] Extract the weighted state vectors at the nearest sampling times and construct a weighted state vector sequence;

[0023] Input the weighted state vector sequence into a gated recurrent unit neural network to predict the predicted core yarn tension at the next sampling time, including: calculating the forgetting coefficient and the retention coefficient based on the weighted state vector at the current sampling time and the hidden state at the previous sampling time; selectively forgetting the hidden state at the previous sampling time using the forgetting coefficient, and fusing it with the weighted state vector at the current sampling time to calculate the candidate hidden state at the current sampling time; using the retention coefficient to fuse the hidden state at the previous sampling time with the candidate hidden state at the current sampling time to obtain the hidden state at the current sampling time; obtaining the predicted core yarn tension at the next sampling time through a fully connected layer based on the hidden state at the current sampling time.

[0024] Preferably, the adaptive controller includes:

[0025] Set fuzzy sets for the tension error, the predicted tension error, and the tension error change rate respectively;

[0026] Construct an adaptive Gaussian membership function based on each fuzzy set and initialize the membership function parameters of each adaptive Gaussian membership function; the membership function parameters include the function center value and the function standard deviation;

[0027] Calculate the membership degrees of the tension error, the predicted tension error, and the tension error change rate with the corresponding fuzzy sets respectively through the adaptive Gaussian membership function;

[0028] Construct an initial fuzzy rule base based on fuzzy sets and initialize the rule parameters of each fuzzy rule; the initial fuzzy rule base includes fuzzy rules; the rule parameters include a first rule parameter, a second rule parameter, and a third rule parameter; the first rule parameter is related to the tension error, the second rule parameter is related to the predicted tension error, and the third rule parameter is related to the rate of change of the tension error;

[0029] Calculate the motor current adjustment amount under each fuzzy rule according to the rule parameters in combination with the tension error, the predicted tension error, and the rate of change of the tension error;

[0030] Calculate the triggering strength of each fuzzy rule based on the membership degree, and calculate the weighted motor current adjustment amount by the weighted average method according to the triggering strength and the motor current adjustment amount, and obtain the motor current control amount through the weighted motor current adjustment amount.

[0031] Preferably, the method for real-time adjusting the parameters of the adaptive controller by the online parameter optimization method includes: constructing a comprehensive performance index, which consists of a control accuracy index, a control smoothness index, and a tension stability index; based on the comprehensive performance index, optimizing the rule parameters and the membership function parameters by the gradient descent method.

[0032] A spindle motion state control system for covering yarn spinning, comprising:

[0033] A data acquisition unit, which collects the core yarn tension, motor operation data, and environmental data according to the sampling period during the covering yarn spinning process; the motor operation data includes motor torque, motor speed, and motor current; the environmental data includes environmental temperature and environmental humidity;

[0034] A tension prediction unit, which constructs a covering yarn spinning state vector based on the core yarn tension, the motor operation data, and the environmental data, calculates the importance weights of the elements in the covering yarn spinning state vector for the tension change to obtain an importance weight vector, performs weighted processing on the covering yarn spinning state vector according to the importance weight vector to obtain a weighted state vector, and based on the weighted state vector, obtains the predicted core yarn tension at the next sampling moment through a tension prediction model;

[0035] An error analysis unit, which sets a target core yarn tension, calculates the tension error at the current sampling moment according to the target core yarn tension and the core yarn tension at the current sampling moment, calculates the predicted tension error according to the predicted core yarn tension and the target core yarn tension, and calculates the rate of change of the tension error according to the tension error at the current sampling moment and the previous sampling moment;

[0036] The adaptive control unit, based on the tension error, the predicted tension error, and the rate of change of the tension error, obtains the motor current control quantity through an adaptive controller, calculates the motor voltage control quantity according to the motor current control quantity, controls the spindle based on the motor voltage control quantity, and adjusts the parameters of the adaptive controller in real time through an online parameter optimization method.

[0037] Preferably, the process of weighting the wrapped yarn spinning state vector includes:

[0038] Calculating the importance weight of each element in the wrapped yarn spinning state vector for the tension change, and the calculation formula is:

[0039] ;

[0040] Wherein, is the importance weight of the th element, is the natural exponential function, is the basic importance coefficient of the th element, is the number of elements in the wrapped yarn spinning state vector, is the th element and the correlation coefficient of the tension;

[0041] Construct the importance weight vectors with the importance weights, and perform weighted processing on the wrapped yarn spinning state vector according to the importance weight vectors to obtain the weighted state vector, and the formula is:

[0042] ;

[0043] Wherein, is the weighted state vector at the current sampling time , is the importance weight vector, is the wrapped yarn spinning state vector at the current sampling time , is the Hadamard product symbol.

[0044] Preferably, the process of the tension prediction model includes:

[0045] Extract the weighted state vectors at the nearest sampling times and construct a weighted state vector sequence;

[0046] Input the weighted state vector sequence into a gated recurrent unit neural network to predict the predicted core yarn tension at the next sampling moment, including: calculating a forgetting coefficient and a retention coefficient based on the weighted state vector at the current sampling moment and the hidden state at the previous sampling moment; selectively forgetting the hidden state at the previous sampling moment by using the forgetting coefficient, and fusing it with the weighted state vector at the current sampling moment to calculate the candidate hidden state at the current sampling moment; using the retention coefficient to fuse the hidden state at the previous sampling moment with the candidate hidden state at the current sampling moment to obtain the hidden state at the current sampling moment; obtaining the predicted core yarn tension at the next sampling moment through a fully connected layer based on the hidden state at the current sampling moment.

[0047] Preferably, the adaptive controller includes:

[0048] Set fuzzy sets for the tension error, the predicted tension error, and the rate of change of the tension error respectively;

[0049] Construct an adaptive Gaussian membership function based on each fuzzy set, and initialize the membership function parameters of each adaptive Gaussian membership function; the membership function parameters include the function center value and the function standard deviation;

[0050] Calculate the membership degrees of the tension error, the predicted tension error, and the rate of change of the tension error with the corresponding fuzzy sets respectively through the adaptive Gaussian membership function;

[0051] Construct an initial fuzzy rule base based on the fuzzy sets, and initialize the rule parameters of each fuzzy rule; the initial fuzzy rule base includes fuzzy rules; the rule parameters include a first rule parameter, a second rule parameter, and a third rule parameter; the first rule parameter is related to the tension error, the second rule parameter is related to the predicted tension error, and the third rule parameter is related to the rate of change of the tension error;

[0052] Calculate the motor current adjustment amount under each fuzzy rule according to the rule parameters in combination with the tension error, the predicted tension error, and the rate of change of the tension error;

[0053] Calculate the triggering intensity of each fuzzy rule based on the membership degree, and calculate the weighted motor current adjustment amount through the weighted average method according to the triggering intensity and the motor current adjustment amount, and obtain the motor current control amount through the weighted motor current adjustment amount.

[0054] Preferably, the real-time adjustment of the parameters of the adaptive controller by the online parameter optimization method includes: constructing a comprehensive performance index, which is composed of a control accuracy index, a control smoothness index, and a tension stability index; based on the comprehensive performance index, the gradient descent method is used to optimize the rule parameters and the membership function parameters.

[0055] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0056] 1. By constructing a covered yarn spinning state vector including core yarn tension, motor operation data, and environmental data, and calculating the importance weights of each data through the basic importance coefficient and Pearson correlation coefficient set in combination with historical experience, and performing weighted processing, it is possible to adaptively adjust the influence degree of different parameters on tension control, and pay more attention to the key data that has a significant impact on tension changes. By combining the importance weight vector with the covered yarn spinning state vector, the obtained weighted state vector more accurately reflects the actual contribution of each data to system control. It effectively improves the response sensitivity of the control method to changes in key parameters, and thus improves the accuracy of predicting the core yarn tension.

[0057] 2. The gated recurrent unit neural network is used to construct a tension prediction model. By extracting the weighted state vector sequence as the input and using the forgetting gate and update gate mechanisms, selective memory and update of historical state information are realized. The model calculates the forgetting coefficient and retention coefficient, selectively forgets the hidden state of the previous moment, and fuses it with the current state vector to obtain a candidate hidden state; then, the historical information and new information are fused through the retention coefficient, and finally the core yarn tension of the next moment is predicted through the fully connected layer. The prediction mechanism makes full use of the continuity characteristics of production data, can accurately capture the tension change trend, improves the pre-control ability of the control method, and improves the control accuracy in a changing production environment.

[0058] 3. An adaptive controller is designed. By setting fuzzy sets of tension error, predicted tension error, and tension error change rate, constructing an adaptive Gaussian membership function, and combining with a fuzzy rule base to obtain the motor current control amount, the underlying control decision-making is realized, avoiding the interference of the production environment on the control accuracy. The controller uses a comprehensive performance index for online parameter optimization, including a control accuracy index, a control smoothness index, and a tension stability index, and uses the gradient descent method to optimize the rule parameters and membership function parameters in real time, reasonably balancing the multi-dimensional requirements of control accuracy, smoothness, and stability, overcoming the problem of insufficient adaptability of traditional control methods, and significantly improving the control accuracy of the control method in a complex and changing environment. Description of the Drawings

[0059] Figure 1 It is a flowchart of a method for controlling the motion state of a spindle for covered yarn spinning according to the present invention;

[0060] Figure 2 is the flow chart of the adaptive controller of the present invention;

[0061] Figure 3 is the structure diagram of a control system for the motion state of a spindle in covered yarn spinning of the present invention. Specific embodiments

[0062] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

[0063] Embodiment 1

[0064] Factory A uses a control method for the motion state of a spindle in covered yarn spinning in order to control the core yarn tension through precise control of the spindle during the covered yarn spinning process, thereby improving product quality and work efficiency. Referring to Figure 1 , including:

[0065] During the covered yarn spinning process, the core yarn tension, motor operation data, and environmental data are collected according to the sampling period; the motor operation data includes motor torque, motor speed, and motor current; the environmental data includes environmental temperature and environmental humidity;

[0066] Based on the core yarn tension, the motor operation data, and the environmental data, a covered yarn spinning state vector is constructed;

[0067] Calculate the importance weights of each element in the covered yarn spinning state vector for the tension change to obtain an importance weight vector, and perform weighted processing on the covered yarn spinning state vector according to the importance weight vector to obtain a weighted state vector;

[0068] Based on the weighted state vector, the predicted core yarn tension at the next sampling moment is obtained through a tension prediction model;

[0069] Set the target core yarn tension, calculate the tension error at the current sampling moment according to the target core yarn tension and the core yarn tension at the current sampling moment, calculate the predicted tension error according to the predicted core yarn tension and the target core yarn tension, and calculate the tension error change rate according to the tension errors at the current sampling moment and the previous sampling moment;

[0070] Based on the tension error, the predicted tension error, and the rate of change of the tension error, an adaptive controller is used to obtain the motor current control quantity. The motor voltage control quantity is calculated according to the motor current control quantity. The spindle is controlled based on the motor voltage control quantity, and the parameters of the adaptive controller are adjusted in real time through an online parameter optimization method.

[0071] Furthermore, the target core yarn tension is the ideal core yarn tension value set according to the yarn specifications and covering process requirements, which can ensure that the covered yarn is evenly wound around the core yarn to meet the physical properties and appearance quality requirements of the covered yarn. In practical applications, the target core yarn tension can be preset according to the process requirements and adjusted appropriately according to changes in environmental temperature and humidity to ensure the quality stability of the covered yarn.

[0072] Furthermore, the process of weighting the covered yarn textile state vector includes:

[0073] Calculating the importance weight of each element in the covered yarn textile state vector for the tension change. The calculation formula is:

[0074] ;

[0075] where, is the importance weight of the th element, is the natural exponential function, is the basic importance coefficient of the th element, which is set according to historical experience, is the number of elements in the covered yarn textile state vector, is the th element's correlation coefficient with the tension, which is calculated using the Pearson correlation coefficient;

[0076] Constructing the importance weight vector with importance weights. The weighted state vector is obtained by weighting the covered yarn textile state vector according to the importance weight vector. The formula is:

[0077] ;

[0078] where, is the weighted state vector at the current sampling time , is the importance weight vector, is the covered yarn textile state vector at the current sampling time , is the Hadamard product symbol.

[0079] By introducing a weight calculation method based on the natural exponential function and combining the basic importance coefficient and the Pearson correlation coefficient, the importance of each element of the state vector is accurately quantified, enabling the system to more accurately grasp the influence degree of each element on tension control and improving the pertinence and effectiveness of the control system.

[0080] Further, the process of the tension prediction model includes:

[0081] Extract the weighted state vectors at the most recent

[0082] sampling moments and construct a weighted state vector sequence;

[0083] Input the weighted state vector sequence into the gated recurrent unit neural network to predict the predicted core yarn tension at the next sampling moment, including: calculating the forgetting coefficient and the retention coefficient based on the weighted state vector at the current sampling moment and the hidden state at the previous sampling moment; selectively forgetting the hidden state at the previous sampling moment using the forgetting coefficient and fusing it with the weighted state vector at the current sampling moment to calculate the candidate hidden state at the current sampling moment; using the retention coefficient to fuse the hidden state at the previous sampling moment with the candidate hidden state at the current sampling moment to obtain the hidden state at the current sampling moment; obtaining the predicted core yarn tension at the next sampling moment through a fully connected layer based on the hidden state at the current sampling moment.

[0084] Further, referring to Figure 2 , the adaptive controller includes:

[0085] Set fuzzy sets for the tension error, the predicted tension error, and the tension error change rate respectively; specifically, the fuzzy sets are shown in Table 1, and 5 fuzzy sets are set respectively: {negative large, negative small, zero, positive small, positive large};

[0086] Construct an adaptive Gaussian membership function based on each fuzzy set and initialize the membership function parameters of each adaptive Gaussian membership function; the membership function parameters include the function center value and the function standard deviation;

[0087] Calculate the membership degrees of the tension error, the predicted tension error, and the tension error change rate with the corresponding fuzzy sets respectively through the adaptive Gaussian membership function;

[0088] Construct an initial fuzzy rule base based on fuzzy sets and initialize the rule parameters of each fuzzy rule; the initial fuzzy rule base includes fuzzy rules; the rule parameters include a first rule parameter, a second rule parameter, and a third rule parameter; the first rule parameter is related to the tension error, the second rule parameter is related to the predicted tension error, and the third rule parameter is related to the rate of change of the tension error;

[0089] Calculate the motor current adjustment amount under each fuzzy rule according to the rule parameters in combination with the tension error , the predicted tension error and the rate of change of the tension error . The calculation formula is:

[0090] ;

[0091] where, is the motor current adjustment amount under the th fuzzy rule, is the first rule parameter corresponding to the th fuzzy rule, is the second rule parameter corresponding to the th fuzzy rule, is the third rule parameter corresponding to the th fuzzy rule;

[0092] Calculate the triggering intensity of each fuzzy rule based on the membership degree. According to the triggering intensity and the motor current adjustment amount, calculate the weighted motor current adjustment amount by the weighted average method, and obtain the motor current control amount through the weighted motor current adjustment amount. The calculation formula is:

[0093] ;

[0094] where, is the triggering intensity of the th fuzzy rule, is the membership degree of the tension error and the corresponding th fuzzy set, is the membership degree of the predicted tension error and the corresponding th fuzzy set, is the membership degree of the rate of change of the tension error and the corresponding th fuzzy set, is the weighted motor current adjustment amount. Add the weighted motor current adjustment amount to the motor current at the previous sampling moment to obtain the motor current control amount.

[0095] Table 1 Fuzzy Sets and Corresponding Membership Function Parameters

[0096]

[0097] The formula for calculating the motor voltage control quantity based on the motor current control quantity is as follows:

[0098] ;

[0099] Wherein, is the motor voltage control quantity, is the armature resistance, is the back electromotive force coefficient, is the motor angular velocity, is the motor current control quantity. The armature resistance and the back electromotive force coefficient are equipment characteristic parameters of the motor.

[0100] Taking the tension error as an example, the Gaussian membership function of the corresponding th fuzzy set is:

[0101] ;

[0102] Wherein, is the tension error corresponding to the th Gaussian membership function of the fuzzy set, is the natural exponential function, is the center value of the function, is the standard deviation of the function. The initialization of the center value of the function and the standard deviation of the function is shown in Table 1.

[0103] Furthermore, a trigger intensity threshold is set. When the trigger intensity is less than the trigger intensity threshold, it indicates that the corresponding fuzzy rule is not activated. A statistical period is set to statistically analyze the number of times each of the fuzzy rules is not activated within the period, and the fuzzy rules with the number of times not activated greater than the preset threshold are deleted from the initial fuzzy rule base.

[0104] Specifically, some fuzzy rules and the corresponding rule parameters are shown in Table 2. Taking the first fuzzy rule as an example, the rule content is: map the tension error, the predicted tension error, and the rate of change of the tension error to the fuzzy set {negative large}, calculate the membership degrees respectively through the center value of the function and the standard deviation of the function corresponding to the fuzzy set {negative large}, calculate the trigger intensity of the first fuzzy rule according to the membership degrees, and calculate the motor current adjustment amount under the first fuzzy rule through the corresponding first rule parameter, second rule parameter, and third rule parameter.

[0105] Table 2 Fuzzy Rules and Corresponding Parameters

[0106]

[0107] The adaptive controller adopts Gaussian membership functions and a dynamically optimized fuzzy rule base, and combines a trigger intensity threshold mechanism to achieve adaptive adjustment of control rules, improve the degree of intelligence and control accuracy, and enhance the ability of the control method to handle complex working conditions.

[0108] Furthermore, the parameters of the adaptive controller are adjusted in real time through an online parameter optimization method, including: constructing a comprehensive performance index, which consists of a control accuracy index, a control smoothness index, and a tension stability index; based on the comprehensive performance index, the gradient descent method is used to optimize the rule parameters and the membership function parameters.

[0109] The calculation formula for the comprehensive performance index is:

[0110] ;

[0111] where is the comprehensive performance index at the current sampling time , the first term of the formula is the control accuracy index, the second term is the control smoothness index, and the third term is the tension stability index, is the tension error at the current sampling time , is the motor current control amount at the current sampling time , is the core yarn tension at the current sampling time , is the core yarn tension at the previous sampling time , , and are weight coefficients.

[0112] By constructing a comprehensive performance index that includes control accuracy, smoothness, and stability, and using the gradient descent method for parameter optimization, real-time optimization adjustment of control parameters is achieved, ensuring balanced optimization of multiple performance indicators and improving the overall performance of the control method.

[0113] By constructing a covered yarn spinning state vector containing multi-dimensional data and performing weighted processing based on importance weights, it is possible to adaptively focus on key parameters that have a greater impact on tension changes. The tension prediction model constructed using a gated recurrent unit neural network makes full use of the continuity characteristics of time-series data, enhancing the predictive control ability of the system. At the same time, the designed adaptive controller combined with an online parameter optimization mechanism can output a motor current control amount to control the spindle, and at the same time, realize real-time optimization and adjustment of the controller parameters. It not only ensures the precise control of the spindle movement but also enhances the adaptability of the system to environmental changes, overcomes the problems of inaccurate quantification of parameter effects, lack of predictive control, and insufficient adaptability in traditional control methods, significantly improves the control accuracy in the covered yarn spinning process, and thus improves the quality and production efficiency of the covered yarn.

[0114] Example Two

[0115] Factory B uses a control system for the movement state of the spindle in covered yarn spinning to reduce the defect rate of covered yarn products, referring to Figure 3 , including:

[0116] A data acquisition unit that collects core yarn tension, motor operation data, and environmental data according to a sampling period during the covered yarn spinning process; the motor operation data includes motor torque, motor speed, and motor current; the environmental data includes environmental temperature and environmental humidity;

[0117] A tension prediction unit that constructs a covered yarn spinning state vector based on the core yarn tension, the motor operation data, and the environmental data, calculates the importance weights of the elements in the covered yarn spinning state vector for tension changes to obtain an importance weight vector, performs weighted processing on the covered yarn spinning state vector according to the importance weight vector to obtain a weighted state vector, and based on the weighted state vector, obtains the predicted core yarn tension at the next sampling moment through a tension prediction model;

[0118] An error analysis unit that sets a target core yarn tension, calculates the tension error at the current sampling moment according to the target core yarn tension and the core yarn tension at the current sampling moment, calculates the predicted tension error according to the predicted core yarn tension and the target core yarn tension, and calculates the tension error change rate according to the tension errors at the current sampling moment and the previous sampling moment;

[0119] An adaptive control unit that, based on the tension error, the predicted tension error, and the tension error change rate, obtains a motor current control amount through an adaptive controller, calculates a motor voltage control amount according to the motor current control amount, realizes the control of the spindle according to the motor voltage control amount, and adjusts the parameters of the adaptive controller in real time through an online parameter optimization method.

[0120] Further, the process of weighting the wrapped yarn spinning state vector includes:

[0121] Calculating the importance weights of each element in the wrapped yarn spinning state vector with respect to the change in tension, and the calculation formula is:

[0122] ;

[0123] where, is the importance weight of the th element, is the natural exponential function, is the basic importance coefficient of the th element, is the number of elements in the wrapped yarn spinning state vector, is the th element's correlation coefficient with tension;

[0124] Taking such importance weights to form the importance weight vector, and performing weighting processing on the wrapped yarn spinning state vector according to the importance weight vector to obtain the weighted state vector, and the formula is:

[0125] ;

[0126] where, is the weighted state vector at the current sampling time , is the importance weight vector, is the wrapped yarn spinning state vector at the current sampling time , is the Hadamard product symbol.

[0127] Further, the process of the tension prediction model includes:

[0128] Extracting the weighted state vectors at the most recent sampling times and constructing a weighted state vector sequence;

[0129] Inputting the weighted state vector sequence into a gated recurrent unit neural network to predict the predicted core yarn tension at the next sampling moment includes: calculating a forgetting coefficient and a retention coefficient based on the weighted state vector at the current sampling moment and the hidden state at the previous sampling moment; selectively forgetting the hidden state at the previous sampling moment by using the forgetting coefficient, and fusing it with the weighted state vector at the current sampling moment to calculate the candidate hidden state at the current sampling moment; using the retention coefficient to fuse the hidden state at the previous sampling moment with the candidate hidden state at the current sampling moment to obtain the hidden state at the current sampling moment; and obtaining the predicted core yarn tension at the next sampling moment through a fully connected layer based on the hidden state at the current sampling moment.

[0130] Further, the adaptive controller includes:

[0131] Setting fuzzy sets for the tension error, the predicted tension error, and the rate of change of the tension error respectively;

[0132] Constructing an adaptive Gaussian membership function based on each fuzzy set and initializing the membership function parameters of each adaptive Gaussian membership function; the membership function parameters include the function center value and the function standard deviation;

[0133] Calculating the membership degrees of the tension error, the predicted tension error, and the rate of change of the tension error with respect to the corresponding fuzzy sets respectively through the adaptive Gaussian membership function;

[0134] Constructing an initial fuzzy rule base based on the fuzzy sets and initializing the rule parameters of each fuzzy rule; the initial fuzzy rule base includes fuzzy rules; the rule parameters include a first rule parameter, a second rule parameter, and a third rule parameter; the first rule parameter is related to the tension error, the second rule parameter is related to the predicted tension error, and the third rule parameter is related to the rate of change of the tension error;

[0135] Calculating the motor current adjustment amount under each fuzzy rule according to the rule parameters in combination with the tension error, the predicted tension error, and the rate of change of the tension error;

[0136] Calculating the triggering intensity of each fuzzy rule based on the membership degree, and calculating the weighted motor current adjustment amount through the weighted average method according to the triggering intensity and the motor current adjustment amount, and obtaining the motor current control amount through the weighted motor current adjustment amount.

[0137] Furthermore, real-time adjustment of the parameters of the adaptive controller by the online parameter optimization method includes: constructing a comprehensive performance index, which is composed of a control accuracy index, a control smoothness index, and a tension stability index; and optimizing the rule parameters and the membership function parameters based on the comprehensive performance index by using the gradient descent method.

[0138] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A control method for the motion state of a spindle rod used in covered yarn spinning, characterized in that, Including: During the wrapping yarn spinning process, the core yarn tension, motor operation data, and environmental data are collected according to the sampling period; The motor operation data includes motor torque, motor speed, motor angular velocity, and motor current; the environmental data includes environmental temperature and environmental humidity; Based on the core yarn tension, the motor operation data, and the environmental data, a wrapping yarn spinning state vector is constructed; Combining the basic importance coefficient set according to historical experience and the correlation coefficient of tension calculated using the Pearson correlation coefficient, calculate the importance weight of each element in the wrapping yarn spinning state vector for tension change, obtain the importance weight vector, and perform weighted processing on the wrapping yarn spinning state vector according to the importance weight vector to obtain the weighted state vector; Use a gated recurrent unit neural network to construct a tension prediction model. Based on the weighted state vector, obtain the predicted core yarn tension at the next sampling moment through the tension prediction model; Set the target core yarn tension, calculate the tension error at the current sampling moment according to the target core yarn tension and the core yarn tension at the current sampling moment, calculate the predicted tension error according to the predicted core yarn tension and the target core yarn tension, and calculate the tension error change rate according to the tension errors at the current sampling moment and the previous sampling moment; Based on the tension error, the predicted tension error, and the tension error change rate, obtain the motor current control amount through an adaptive controller, calculate the motor voltage control amount according to the motor current control amount, control the spindle according to the motor voltage control amount, and adjust the parameters of the adaptive controller in real time through an online parameter optimization method; obtaining the motor current control amount through the adaptive controller specifically includes: setting fuzzy sets of tension error, predicted tension error, and tension error change rate, constructing an adaptive Gaussian membership function, and obtaining the motor current control amount in combination with a fuzzy rule base; The parameters of the adaptive controller include rule parameters and membership function parameters; the rule parameters include the first rule parameter, the second rule parameter, and the third rule parameter; the first rule parameter is related to the tension error, the second rule parameter is related to the predicted tension error, and the third rule parameter is related to the tension error change rate.

2. A method for controlling the motion state of a spindle rod for covered yarn spinning according to claim 1, characterized in that, The process of performing weighted processing on the wrapping yarn spinning state vector includes: Calculate the importance weight of each element in the wrapping yarn spinning state vector for tension change, and the calculation formula is: ； Among them, is the importance weight of the th element, is the natural exponential function, is the basic importance coefficient of the th element, is the number of elements in the covered yarn spinning state vector, is the correlation coefficient between the th element and the tension, calculated using the Pearson correlation coefficient; The importance weights form the importance weight vector. The weighted state vector is obtained by performing a weighting process on the covered yarn spinning state vector according to the importance weight vector. The formula is as follows: ； Among them, is the current sampling moment of the weighted state vector is the importance weight vector is the current sampling moment of the covered yarn spinning state vector is the Hadamard product symbol 3. A method for controlling the motion state of a spindle rod for covering yarn spinning according to claim 1, characterized in that, The process of the tension prediction model includes: Extract the most recent weighted state vectors at sampling instants and construct a sequence of weighted state vectors; Input the weighted state vector sequence into a gated recurrent unit neural network to predict the predicted core yarn tension at the next sampling moment, including: calculating a forgetting coefficient and a retention coefficient based on the weighted state vector at the current sampling moment and the hidden state at the previous sampling moment; selectively forgetting the hidden state at the previous sampling moment using the forgetting coefficient, and fusing it with the weighted state vector at the current sampling moment to calculate the candidate hidden state at the current sampling moment; using the retention coefficient to fuse the hidden state at the previous sampling moment with the candidate hidden state at the current sampling moment to obtain the hidden state at the current sampling moment; and obtaining the predicted core yarn tension at the next sampling moment through a fully connected layer based on the hidden state at the current sampling moment.

4. A method for controlling the motion state of a spindle rod for covered yarn spinning according to claim 1, characterized in that, The adaptive controller includes: Set fuzzy sets for the tension error, the predicted tension error, and the tension error change rate respectively ; Construct an adaptive Gaussian membership function based on each fuzzy set, and initialize the membership function parameters of each adaptive Gaussian membership function; the membership function parameters include the function center value and the function standard deviation; Calculate the membership degrees of the tension error, the predicted tension error, and the change rate of the tension error with respect to the corresponding fuzzy sets through the adaptive Gaussian membership function respectively; Construct an initial fuzzy rule base based on fuzzy sets and initialize the rule parameters of each fuzzy rule; the initial fuzzy rule base includes fuzzy rules; the rule parameters include a first rule parameter, a second rule parameter, and a third rule parameter; the first rule parameter is related to the tension error, the second rule parameter is related to the predicted tension error, and the third rule parameter is related to the rate of change of the tension error; Calculate the motor current adjustment amount under each fuzzy rule according to the rule parameters in combination with the tension error, the predicted tension error, and the change rate of the tension error; Calculate the triggering intensity of each fuzzy rule based on the membership degree, and calculate the weighted motor current adjustment amount through the weighted average method according to the triggering intensity and the motor current adjustment amount, and obtain the motor current control amount through the weighted motor current adjustment amount.

5. A method for controlling the motion state of a spindle rod for covering yarn spinning according to claim 4, characterized in that, Real-time adjustment of the adaptive controller parameters through an online parameter optimization method includes: constructing a comprehensive performance index, which consists of a control accuracy index, a control smoothness index, and a tension stability index; and optimizing the rule parameters and the membership function parameters based on the comprehensive performance index by using the gradient descent method.

6. A control system for the motion state of a spindle rod used in covered yarn spinning, characterized in that, It includes: A data acquisition unit that collects core yarn tension, motor operation data, and environmental data according to a sampling period during the covered yarn spinning process; The motor operation data includes motor torque, motor speed, and motor current; the environmental data includes environmental temperature and environmental humidity; A tension prediction unit that constructs a covered yarn spinning state vector based on the core yarn tension, the motor operation data, and the environmental data, combines the basic importance coefficient set according to historical experience and the correlation coefficient of the tension calculated using the Pearson correlation coefficient, calculates the importance weights of the elements in the covered yarn spinning state vector for the tension change to obtain an importance weight vector, performs weighted processing on the covered yarn spinning state vector according to the importance weight vector to obtain a weighted state vector, constructs a tension prediction model using a gated recurrent unit neural network, and obtains the predicted core yarn tension at the next sampling moment through the tension prediction model based on the weighted state vector; An error analysis unit sets a target core yarn tension, calculates a tension error at the current sampling moment based on the target core yarn tension and the core yarn tension at the current sampling moment, calculates a predicted tension error based on the predicted core yarn tension and the target core yarn tension, and calculates a rate of change of the tension error based on the tension errors at the current sampling moment and the previous sampling moment; An adaptive control unit, based on the tension error, the predicted tension error, and the rate of change of the tension error, obtains a motor current control amount through an adaptive controller, calculates a motor voltage control amount based on the motor current control amount, controls the spindle based on the motor voltage control amount, and adjusts the parameters of the adaptive controller in real time through an online parameter optimization method; Obtaining the motor current control amount through the adaptive controller specifically includes: setting fuzzy sets of the tension error, the predicted tension error, and the rate of change of the tension error, constructing an adaptive Gaussian membership function, and obtaining the motor current control amount in combination with a fuzzy rule base; The parameters of the adaptive controller include rule parameters and membership function parameters; the rule parameters include a first rule parameter, a second rule parameter, and a third rule parameter; the first rule parameter is related to the tension error, the second rule parameter is related to the predicted tension error, and the third rule parameter is related to the rate of change of the tension error.

7. A control system for the motion state of a spindle rod used in covered yarn spinning, characterized in that, The process of weighting the covered yarn spinning state vector includes: Calculating the importance weights of the elements in the covered yarn spinning state vector for the tension change, and the calculation formula is: ； Among them, is the importance weight of the element, is the natural exponential function, is the basic importance coefficient of the element, is the number of elements in the covered yarn spinning state vector, is the correlation coefficient between the element and the tension, calculated using the Pearson correlation coefficient; The importance weights form the importance weight vector. The weighted state vector is obtained by weighting the covered yarn spinning state vector according to the importance weight vector. The formula is: ； Among them, is the weighted state vector at the current sampling moment ; is the importance weight vector ; is the covered yarn spinning state vector at the current sampling moment is the Hadamard product symbol.

8. A control system for the motion state of a spindle rod used for covered yarn spinning, characterized in that, The process of the tension prediction model includes: Extract the most recent weighted state vectors at sampling instants and construct a sequence of weighted state vectors; Inputting the weighted state vector sequence into a gated recurrent unit neural network to predict the predicted core yarn tension at the next sampling moment, including: calculating a forgetting coefficient and a retention coefficient based on the weighted state vector at the current sampling moment and the hidden state at the previous sampling moment; selectively forgetting the hidden state at the previous sampling moment by using the forgetting coefficient, and fusing it with the weighted state vector at the current sampling moment to calculate the candidate hidden state at the current sampling moment; using the retention coefficient to fuse the hidden state at the previous sampling moment with the candidate hidden state at the current sampling moment to obtain the hidden state at the current sampling moment; obtaining the predicted core yarn tension at the next sampling moment through a fully connected layer based on the hidden state at the current sampling moment.

9. A control system for the motion state of a spindle rod used for covered yarn spinning, according to claim 6, characterized in that The adaptive controller includes: Set fuzzy sets for the tension error, the predicted tension error, and the rate of change of the tension error, respectively; Constructing an adaptive Gaussian membership function based on each fuzzy set and initializing the membership function parameters of each adaptive Gaussian membership function; the membership function parameters include the function center value and the function standard deviation; Calculating the membership degrees of the tension error, the predicted tension error, and the rate of change of the tension error with the corresponding fuzzy sets respectively through the adaptive Gaussian membership function; Construct an initial fuzzy rule base based on fuzzy sets and initialize the rule parameters of each fuzzy rule; the initial fuzzy rule base includes fuzzy rules; the rule parameters include a first rule parameter, a second rule parameter, and a third rule parameter; the first rule parameter is related to the tension error, the second rule parameter is related to the predicted tension error, and the third rule parameter is related to the rate of change of the tension error; Calculating the motor current adjustment amount under each fuzzy rule according to the rule parameters in combination with the tension error, the predicted tension error, and the rate of change of the tension error; Calculating the triggering intensity of each fuzzy rule based on the membership degree, and calculating the weighted motor current adjustment amount through a weighted average method according to the triggering intensity and the motor current adjustment amount, and obtaining the motor current control amount through the weighted motor current adjustment amount.

10. A control system for the motion state of a spindle rod used for covered yarn spinning according to claim 9, characterized in that, Real-time adjustment of the parameters of the adaptive controller through the online parameter optimization method includes: constructing a comprehensive performance index, which is composed of a control accuracy index, a control smoothness index, and a tension stability index; based on the comprehensive performance index, using the gradient descent method to optimize the rule parameters and the membership function parameters.

Images