Intelligent tension control datamation method and system

Abstract

The invention discloses an intelligent tension control datamation method and system, and relates to the technical field of tension control. An intelligent tension control datamation system comprises a target equipment tension acquisition module, a target equipment tension prediction module and a target equipment tension adjustment module. The real-time tension information and the parameter information of the target equipment are acquired at preset time intervals, so that the state of the equipment is monitored and analyzed in real time; the tension data in the next preset time interval can be predicted by using the target equipment tension prediction model, so that the potential tension deviation problem can be found in advance; by comparing the to-be-analyzed tension data with the predicted tension data, whether the target equipment needs to be adjusted or not can be actively judged, instead of passively responding only after the tension exceeds a preset range; and the target equipment parameter control model can output an optimal parameter control scheme according to the prediction information and the current parameter information, and the equipment performance is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of tension control, and in particular to an intelligent tension control digitization method and system. Background Art

[0002] In the traditional blade winding and unwinding process, if the tension is not digitally controlled, improper tension control often occurs. Improper tension control can cause tight or loose winding, inconsistent pressure between layers of the coil, uneven surface of the coil or uneven roll diameter; during the unwinding process, if the tension is too high, the material may be stretched or torn; and during the winding process, if the tension is too low, the material may wrinkle or curl; in addition, uneven tension will further increase the wear on equipment components, shorten the service life of the equipment, and increase the maintenance and replacement costs of the equipment. Summary of the invention

[0003] The present invention aims to provide an intelligent tension control digitization method and system to perform digitized and intelligent control of tension.

[0004] An intelligent tension control digitization method, comprising: S1. Acquire the real-time tension information of the target device at every preset time interval to obtain the tension data to be analyzed, acquire the real-time parameter information of the target device to obtain the parameter information of the target device to be analyzed; the parameter information of the target device to be analyzed includes various parameter information of the target device that affects the tension data to be analyzed; there is a functional relationship between the tension data to be analyzed and the parameter information of the target device to be analyzed, and the functional relationship is the tension fitting function of the target device; input the parameter information of the target device to be analyzed into the tension prediction model of the target device for analysis to obtain the predicted parameter information of the target device and the predicted tension data of the target device; the tension prediction model of the target device is used to predict the parameter information of the target device within the next preset time interval to obtain the predicted tension data of the target device; S2. Compare the tension data to be analyzed with the predicted target device tension data to determine whether it exceeds the constant difference range of the target device tension; if it exceeds the constant difference range of the target device tension, proceed to step S3; otherwise, continue to control according to the existing target device parameter information; S3. Input the predicted target device tension data and the target device parameter information to be analyzed into the target device parameter control model for analysis to obtain the target device parameter control plan; the target device parameter control model is used to output the target device parameter control plan based on the predicted target device parameter information and the target device parameter information to be analyzed so as to maintain the target device tension data within the constant target device tension difference range; use the target device parameter control plan to adjust the target device.

[0005] As a preferred technical solution of the present invention, the specific steps of determining the tension fitting function of the target device in step S1 include: Collecting several groups of target device tension control training samples; each group of target device tension control training samples contains historical target device parameter information and corresponding historical real tension data; combining several groups of target device tension control training samples to obtain a target device tension control training set; Construct the initial target device tension fitting function. The initial target device tension fitting function is expressed by the formula F=f(X), where F is the simulated tension data, and X=[x1, x2, …, x N ], X represents each parameter in a set of target device parameter information, N is the number of each parameter in the target device parameter information, and f(X) is the functional relationship of the initial target device tension fitting function; Construct K target equipment tension fitting function individuals H k , each target device tension fitting function individual H k contains an initial target device tension fitting function, different target device tension fitting functions H k The functional relationship of the initial target device tension fitting function is different; the K target device tension fitting function individuals H k Combine to obtain the target equipment tension fitting function iterative population; set the maximum number of iterations; The target device tension fitting function individual H in the target device tension fitting function iteration population k , simulate the target equipment tension control training set and obtain the simulated tension data set J k , calculate the simulated tension data set J k The mean square error of the historical real tension data corresponding to the target equipment tension control training set is used to obtain the simulation tension calculation accuracy P k ; The simulation tension calculation accuracy P k The inverse of the target device tension fitting function is used as the individual H k The fitness S k ; For the target device tension fitting function iterative population, a population update operation is performed on the target device tension fitting function iterative population in each iteration to obtain a new target device tension fitting function iterative population; When the maximum number of iterations is reached, the target device tension fitting function individual H corresponding to the maximum fitness is output. k , which is the optimal target device tension fitting function individual; the initial target device tension fitting function in the optimal target device tension fitting function individual is used as the target device tension fitting function for analysis.

[0006] As a preferred technical solution of the present invention, the target equipment tension prediction model in step S1 includes a data preprocessing layer, a prediction analysis layer and a result output layer; The data preprocessing layer is used to preprocess the parameter information of the target device to be analyzed, and obtain the preprocessed parameter information of the target device to be analyzed; The prediction analysis layer is used to perform prediction analysis on the pre-processed target device parameter information to be analyzed, and obtain the predicted target device parameter information and predicted target device tension data within the next preset time interval; The result output layer is used to output the predicted target device parameter information and predicted target device tension data.

[0007] As a preferred technical solution of the present invention, the specific steps of training the prediction analysis layer in the tension prediction model of the target device include: Collecting several groups of continuous target device parameter information prediction samples; each group of target device parameter information prediction samples contains target device parameter information within the same preset time interval; combining several groups of continuous target device parameter information prediction samples to obtain a target device parameter information prediction training set; using the next group of target device parameter information prediction samples as the target value of the previous group of target device parameter information prediction samples; The target device parameter information prediction training set is input into the target device tension prediction model to train the prediction analysis layer with the target value as the target to obtain the initial prediction analysis layer; the initial prediction analysis layer is model evaluated to obtain the initial prediction analysis layer model evaluation result; if the initial prediction analysis layer model evaluation result is passed, the trained initial prediction analysis layer is output as the prediction analysis layer in the target device tension prediction model; otherwise, the target device parameter information prediction training set is used to continue model training.

[0008] As a preferred technical solution of the present invention, the specific steps of determining the constant difference range of the tension of the target device include: Collect a group of abnormal target equipment tension constant training samples B i , i=1,2,…,I; each group of target equipment tension constant training samples B i is a set of continuous target device tension data points, and the target device tension constant training sample B i The corresponding mean tension is the mean tension Z i ; The target equipment tension constant training sample B is group I i Using the Isolation Tree Algorithm to construct I isolated trees T i ; According to the isolated tree T i Calculate the anomaly scores of the child nodes in the anomaly score set C i ; Filter abnormal score set C iThe target device tension data point corresponding to the abnormal score exceeding the abnormal score threshold is obtained, and the abnormal target device tension data point Y is obtained. i ; Calculate the abnormal target equipment tension data point Y i and the mean tension Z i The difference is the constant tension difference D of the target equipment. i ; Traverse all abnormal target equipment tension constant training samples B i , get the constant tension difference D of the target device i , set the tension of the target device to a constant difference value D i The minimum value is recorded as the target equipment tension constant difference range.

[0009] As a preferred technical solution of the present invention, the target device parameter control model in step S3 includes a data analysis layer, a parameter adjustment layer and a solution output layer; The data analysis layer is used to perform difference analysis on the predicted target device tension data and the constant difference range of the target device tension to obtain the target device tension data to be adjusted; The parameter adjustment layer is used to adjust the parameters according to the tension data of the target device to be adjusted and the parameter information of the target device to be analyzed by using the fuzzy control rules of the target device parameter control to obtain the target device parameter control scheme; the fuzzy control rules of the target device parameter control are used to output the target device parameter control scheme; The solution output layer is used to output the target device parameter control solution.

[0010] As a preferred technical solution of the present invention, the specific steps of training the parameter adjustment layer in the target device parameter control model include: Collecting several groups of tension target device parameter control training samples, each group of tension target device parameter control training samples includes tension input and parameter output, the tension input is the target tension adjustment value and the current parameter value, and the parameter output is the target parameter adjustment scheme; combining several groups of tension target device parameter control training samples to obtain a tension target device parameter control training set; Construct fuzzy rules according to the tension target device parameter control training set to obtain the fuzzy control rules for the target device parameter control to be optimized; input the tension input in the tension target device parameter control training sample into the fuzzy control rules for the target device parameter control to be optimized for simulation output to obtain the simulated parameter output; adjust the parameters according to the simulated parameter output and the current parameter value to obtain the simulated target device parameter information; input the simulated target device parameter information into the target device tension fitting function for calculation to obtain the simulated tension adjustment value; compare the simulated tension adjustment value with the target tension adjustment value to obtain the tension control difference; traverse the tension target device parameter control training set to obtain several groups of tension control difference values; Assuming that several groups of tension control differences are all within the error threshold range, the fuzzy control rule for parameter control of the target device to be optimized is used as the fuzzy control rule for parameter control of the target device; otherwise, the fuzzy control rule for parameter control of the target device to be optimized is feedback-adjusted to obtain a new fuzzy control rule for parameter control of the target device to be optimized, and the new fuzzy control rule for parameter control of the target device to be optimized is re-evaluated.

[0011] An intelligent tension control digitization system, comprising: The target device tension acquisition module is used to acquire the real-time tension information of the target device at every preset time interval to obtain the tension data to be analyzed, acquire the real-time parameter information of the target device, and obtain the parameter information of the target device to be analyzed; the parameter information of the target device to be analyzed includes various parameter information of the target device that affects the tension data to be analyzed; there is a functional relationship between the tension data to be analyzed and the parameter information of the target device to be analyzed, and the functional relationship is a tension fitting function of the target device; The target device tension prediction module is used to deploy the target device tension prediction model, input the target device parameter information to be analyzed into the target device tension prediction model for analysis, and obtain the predicted target device parameter information and the predicted target device tension data; the target device tension prediction model is used to predict the target device parameter information within the next preset time interval to obtain the predicted target device tension data; compare the tension data to be analyzed with the predicted target device tension data to determine whether it exceeds the target device tension constant difference range; if it exceeds the target device tension constant difference range, enter step S3; otherwise, continue to control according to the existing target device parameter information; The target device tension adjustment module is used to deploy the target device parameter control model, input the predicted target device tension data and the target device parameter information to be analyzed into the target device parameter control model for analysis, and obtain the target device parameter control plan; the target device parameter control model is used to output the target device parameter control plan based on the predicted target device parameter information and the target device parameter information to be analyzed so as to maintain the target device tension data within the constant difference range of the target device tension; and adjust the target device using the target device parameter control plan.

[0012] The present invention has the following advantages: 1. The present invention realizes real-time monitoring and analysis of the equipment status by acquiring the real-time tension information and parameter information of the target equipment at preset time intervals; using the target equipment tension prediction model, the tension data in the next time interval can be predicted, so as to discover potential tension deviation problems in advance; by comparing the tension data to be analyzed and the predicted tension data, it can actively determine whether the target equipment needs to be adjusted, rather than passively responding only after the tension exceeds the predetermined range; the target equipment parameter control model can output the optimal parameter control solution according to the predicted information and the current parameter information, and optimize the equipment performance; by dynamically adjusting the target equipment parameters, the tension data can be maintained within a constant difference range to ensure the stability of subsequent winding and unwinding operations.

[0013] 2. The present invention establishes a data-driven tension fitting function by collecting historical parameter information and real tension data. This process utilizes advanced data analysis technology to improve the accuracy and predictive power of the model. In order to increase the diversity of the model and improve the probability of finding a better fitting function, multiple tension fitting function individuals with different functional relationships are constructed. In the model evaluation stage, the mean square error is used as the main evaluation indicator, and the fitness is defined as the inverse of the mean square error between the simulated tension data and the historical real tension data, which reasonably reflects the prediction accuracy of the model. Through the iterative optimization process, the tension fitting function individual with the highest fitness is finally determined as the optimal solution. The optimal target device tension fitting function can accurately predict the tension of the target device, providing a reliable basis for subsequent precise control. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 The figure is a schematic diagram of the structure of an intelligent tension control digitization system used in an embodiment of the present invention. DETAILED DESCRIPTION

[0015] In order to enable persons skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0016] Embodiment 1, a method for digitizing intelligent tension control, comprising: S1. Acquire the real-time tension information of the target device at every preset time interval to obtain the tension data to be analyzed, acquire the real-time parameter information of the target device to obtain the parameter information of the target device to be analyzed; the parameter information of the target device to be analyzed includes various parameter information of the target device that affects the tension data to be analyzed; there is a functional relationship between the tension data to be analyzed and the parameter information of the target device to be analyzed, and the functional relationship is the tension fitting function of the target device; input the parameter information of the target device to be analyzed into the tension prediction model of the target device for analysis to obtain the predicted parameter information of the target device and the predicted tension data of the target device; the tension prediction model of the target device is used to predict the parameter information of the target device within the next preset time interval to obtain the predicted tension data of the target device; The preset time interval is set by professional technicians according to actual conditions. The real-time tension information of the target device is captured by the tension sensor, and the real-time parameter information of the target device is captured by various sensors installed on the target device. The parameter information of the target device to be analyzed contains various parameter information of the target device that affects the tension data to be analyzed, such as speed, torque, etc.; by collecting data at preset time intervals, the tension state of the target device can be monitored in real time and the tension change can be responded to in time; the tension data and related parameter information are obtained at the same time, which ensures the comprehensiveness of the analysis and can more accurately understand the various factors affecting the tension; identifying the functional relationship between tension data and parameters is helpful to build an accurate tension prediction model; The specific steps of determining the tension fitting function of the target device in step S1 include: Collecting several groups of target device tension control training samples; each group of target device tension control training samples contains historical target device parameter information and corresponding historical real tension data; combining several groups of target device tension control training samples to obtain a target device tension control training set; the target device tension control training samples are obtained from an existing database of the target device; Construct the initial target device tension fitting function. The initial target device tension fitting function is expressed by the formula F=f(X), where F is the simulated tension data, and X=[x1, x2, …, x N ], X represents each parameter in a set of target device parameter information, N is the number of each parameter in the target device parameter information, and f(X) is the functional relationship of the initial target device tension fitting function; Construct K target equipment tension fitting function individuals H k , each target device tension fitting function individual H k contains an initial target device tension fitting function, different target device tension fitting functions H k The functional relationship of the initial target device tension fitting function is different; the K target device tension fitting function individuals H kcombination, and obtain the target equipment tension fitting function iteration population; set the maximum number of iterations; the maximum number of iterations is set by professional technicians according to actual conditions; the initial target equipment tension fitting function can be fitted by polynomial regression and other methods, and different target equipment tension fitting function individual H k The coefficients in the initial target device tension fitting function are different, and the constructed functional relationship is different; The target device tension fitting function individual H in the target device tension fitting function iteration population k , simulate the target equipment tension control training set and obtain the simulated tension data set J k , calculate the simulated tension data set J k The mean square error of the historical real tension data corresponding to the target equipment tension control training set is used to obtain the simulation tension calculation accuracy P k ; The simulation tension calculation accuracy P k The inverse of the target device tension fitting function is used as the individual H k The fitness S k ; For the target device tension fitting function iteration population, each iteration performs a population update operation on the target device tension fitting function iteration population to obtain a new target device tension fitting function iteration population; the population update operation can refer to the genetic algorithm to perform selection, recombination or mutation operations to update the target device tension fitting function iteration population; the selection operation is to select individuals from the current population according to fitness to generate the next generation. The higher the fitness of the individual, the greater the probability of being selected. Commonly used selection methods include roulette selection, tournament selection, etc.; the recombination operation is to randomly pair selected individuals and generate new individuals through crossover operations. The crossover operation can be single-point crossover, double-point crossover or uniform crossover, etc. The purpose is to combine the excellent genes of the two parents into the offspring and retain the function coefficients with high fitness; the mutation operation is to perform mutation operations on the newly generated individuals to introduce new genetic diversity, such as randomly changing certain parameters or structures in the function. When the maximum number of iterations is reached, the target device tension fitting function individual H corresponding to the maximum fitness is output. k , which is the optimal target device tension fitting function individual; the initial target device tension fitting function in the optimal target device tension fitting function individual is used as the target device tension fitting function for analysis; By collecting historical parameter information and real tension data, a data-driven tension fitting function is established. This process uses advanced data analysis technology to improve the accuracy and predictive power of the model. In order to increase the diversity of the model and improve the probability of finding a better fitting function, multiple tension fitting function individuals with different functional relationships are constructed. In the model evaluation stage, the mean square error is used as the main evaluation indicator, and the fitness is defined as the inverse of the mean square error between the simulated tension data and the historical real tension data, which reasonably reflects the prediction accuracy of the model. Through the iterative optimization process, the tension fitting function individual with the highest fitness is finally determined as the optimal solution. The optimal target equipment tension fitting function can accurately predict the tension of the target equipment, providing a reliable basis for subsequent precise control. The target equipment tension prediction model in step S1 includes a data preprocessing layer, a prediction analysis layer, and a result output layer; The data preprocessing layer is used to preprocess the parameter information of the target device to be analyzed, and obtain the preprocessed parameter information of the target device to be analyzed; The prediction analysis layer is used to perform prediction analysis on the pre-processed target device parameter information to be analyzed, and obtain the predicted target device parameter information and predicted target device tension data within the next preset time interval; The result output layer is used to output the predicted target equipment parameter information and predicted target equipment tension data; The specific steps of training the prediction analysis layer in the target equipment tension prediction model include: Collecting several groups of continuous target device parameter information prediction samples; each group of target device parameter information prediction samples contains target device parameter information within the same preset time interval; combining several groups of continuous target device parameter information prediction samples to obtain a target device parameter information prediction training set; using the next group of target device parameter information prediction samples as the target value of the previous group of target device parameter information prediction samples; Input the target device parameter information prediction training set into the target device tension prediction model to train the prediction analysis layer with the target value as the target to obtain the initial prediction analysis layer; perform model evaluation on the initial prediction analysis layer to obtain the initial prediction analysis layer model evaluation result; if the initial prediction analysis layer model evaluation result is passed, output the trained initial prediction analysis layer as the prediction analysis layer in the target device tension prediction model; otherwise, continue model training using the target device parameter information prediction training set; The BP neural network model can be used as the basic model of the target equipment tension prediction model. By using continuous target equipment parameter information to predict samples, the model can capture dynamic changes in time series data. By taking the next set of parameter information as the target value of the previous set, the model has the characteristics of autoregression, which helps to understand the time dependency between parameters. By predicting potential tension changes, equipment maintenance and adjustments can be carried out in advance to reduce unexpected downtime. S2. Compare the tension data to be analyzed with the predicted target device tension data to determine whether it exceeds the constant difference range of the target device tension; if it exceeds the constant difference range of the target device tension, proceed to step S3; otherwise, continue to control according to the existing target device parameter information; The specific steps for determining the constant differential range of tension of the target equipment include: Collect a group of abnormal target equipment tension constant training samples B i , i=1,2,…,I; each group of target equipment tension constant training samples B i is a set of continuous target device tension data points, and the target device tension constant training sample B i The corresponding mean tension is the mean tension Z i ; Abnormal target equipment constant tension training sample B i To determine the samples containing abnormal data; The target equipment tension constant training sample B is group I i Using the Isolation Tree Algorithm to construct I isolated trees T i ; According to the isolated tree T i Calculate the anomaly scores of the child nodes in the anomaly score set C i ; Filter abnormal score set C i The target device tension data point corresponding to the abnormal score exceeding the abnormal score threshold is obtained, and the abnormal target device tension data point Y is obtained. i ; Calculate the abnormal target equipment tension data point Y i and the mean tension Z i The difference is the constant tension difference D of the target equipment. i ; The anomaly score is calculated by the depth of the tree; Traverse all abnormal target equipment tension constant training samples B i , get the constant tension difference D of the target device i , set the tension of the target device to a constant difference value D i The minimum value of is recorded as the constant difference range of tension of the target equipment; By using the isolation forest algorithm, abnormal tension data points that deviate from the normal pattern can be effectively identified; the difference between the abnormal data point and the tension mean is calculated, providing a method to quantify the tension deviation; by analyzing the tension constant difference D of multiple samplesi , can determine a reasonable tension control threshold range; by maintaining the tension within a controllable range, reduce the impact of tension fluctuations on subsequent operation processes; S3. Input the predicted target device tension data and the target device parameter information to be analyzed into the target device parameter control model for analysis to obtain the target device parameter control scheme; the target device parameter control model is used to output the target device parameter control scheme according to the predicted target device parameter information and the target device parameter information to be analyzed so as to maintain the target device tension data within the target device tension constant difference range; adjust the target device using the target device parameter control scheme; the purpose of adjusting the target device is to predict the impact of tension fluctuations; The target device parameter control model in step S3 includes a data analysis layer, a parameter adjustment layer and a solution output layer; The data analysis layer is used to perform difference analysis on the predicted target device tension data and the constant difference range of the target device tension to obtain the target device tension data to be adjusted; The parameter adjustment layer is used to adjust the parameters according to the tension data of the target device to be adjusted and the parameter information of the target device to be analyzed by using the fuzzy control rules of the target device parameter control to obtain the target device parameter control scheme; the fuzzy control rules of the target device parameter control are used to output the target device parameter control scheme; The solution output layer is used to output the target device parameter control solution; The specific steps of training the parameter adjustment layer in the target device parameter control model include: Collect several groups of tension target device parameter control training samples, each group of tension target device parameter control training samples contains tension input and parameter output, the tension input is the target tension adjustment value and the current parameter value, and the parameter output is the target parameter adjustment plan; combine several groups of tension target device parameter control training samples to obtain a tension target device parameter control training set; the data in the tension target device parameter control training samples can be manually adjusted to set the parameter output; Construct fuzzy rules according to the tension target device parameter control training set to obtain the fuzzy control rules for the target device parameter control to be optimized; input the tension input in the tension target device parameter control training sample into the fuzzy control rules for the target device parameter control to be optimized for simulation output to obtain the simulated parameter output; adjust the parameters according to the simulated parameter output and the current parameter value to obtain the simulated target device parameter information; input the simulated target device parameter information into the target device tension fitting function for calculation to obtain the simulated tension adjustment value; compare the simulated tension adjustment value with the target tension adjustment value to obtain the tension control difference; traverse the tension target device parameter control training set to obtain several groups of tension control difference values; Assuming that the difference values ​​of several groups of tension control are all within the error threshold range, the fuzzy control rule of the target device parameter control to be optimized is used as the fuzzy control rule of the target device parameter control; otherwise, the fuzzy control rule of the target device parameter control to be optimized is feedback-adjusted to obtain a new fuzzy control rule of the target device parameter control to be optimized, and the new fuzzy control rule of the target device parameter control to be optimized is re-evaluated; the error threshold range is set by professional technicians according to actual conditions; By collecting actual tension input and parameter output, data-based fuzzy control rules are constructed to improve the practicality and accuracy of the control rules; using fuzzy logic to deal with tension control problems can handle ambiguity and uncertainty and improve the robustness of the system; by comparing the simulated tension adjustment value and the target tension adjustment value, a closed-loop control system is formed, which helps to reduce the control error; if the control difference exceeds the error threshold, the fuzzy control rules are adjusted through the feedback adjustment mechanism, which enhances the adaptive ability of the model.

[0017] Example 2, an intelligent tension control data system, see Figure 1 As shown, including: The target device tension acquisition module is used to acquire the real-time tension information of the target device at every preset time interval to obtain the tension data to be analyzed, acquire the real-time parameter information of the target device, and obtain the parameter information of the target device to be analyzed; the parameter information of the target device to be analyzed includes various parameter information of the target device that affects the tension data to be analyzed; there is a functional relationship between the tension data to be analyzed and the parameter information of the target device to be analyzed, and the functional relationship is a tension fitting function of the target device; The target device tension prediction module is used to deploy the target device tension prediction model, input the target device parameter information to be analyzed into the target device tension prediction model for analysis, and obtain the predicted target device parameter information and the predicted target device tension data; the target device tension prediction model is used to predict the target device parameter information within the next preset time interval to obtain the predicted target device tension data; compare the tension data to be analyzed with the predicted target device tension data to determine whether it exceeds the target device tension constant difference range; if it exceeds the target device tension constant difference range, enter step S3; otherwise, continue to control according to the existing target device parameter information; The target device tension adjustment module is used to deploy the target device parameter control model, input the predicted target device tension data and the target device parameter information to be analyzed into the target device parameter control model for analysis, and obtain the target device parameter control plan; the target device parameter control model is used to output the target device parameter control plan based on the predicted target device parameter information and the target device parameter information to be analyzed so as to maintain the target device tension data within the constant difference range of the target device tension; and adjust the target device using the target device parameter control plan.

[0018] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all such improvements and changes should fall within the scope of protection of the appended claims of the present invention. Parts not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims

1. An intelligent tension control digitization method, characterized in that: include: S1. Obtain real-time tension information of the target device at preset time intervals, obtain tension data to be analyzed, obtain real-time parameter information of the target device, and obtain parameter information of the target device to be analyzed; The target device parameter information to be analyzed includes various parameter information of the target device that affects the tension data to be analyzed; there is a functional relationship between the tension data to be analyzed and the target device parameter information to be analyzed, and the functional relationship is a target device tension fitting function; the target device parameter information to be analyzed is input into the target device tension prediction model for analysis to obtain predicted target device parameter information and predicted target device tension data; the target device tension prediction model is used to predict the target device parameter information within the next preset time interval to obtain predicted target device tension data; S2. Compare the tension data to be analyzed with the predicted target device tension data to determine whether it exceeds the constant difference range of the target device tension; If it exceeds the constant tension difference range of the target device, then go to step S3; otherwise, continue to control according to the parameter information of the existing target device; S3. The predicted target device tension data and the target device parameter information to be analyzed are input into the target device parameter control model for analysis to obtain the target device parameter control scheme; The target device parameter control model is used to output a target device parameter control scheme according to the predicted target device parameter information and the target device parameter information to be analyzed so as to maintain the target device tension data within a constant difference range of the target device tension; Use the target device parameter control scheme to adjust the target device.

2. The intelligent tension control digitization method according to claim 1 is characterized in that: The specific steps of determining the tension fitting function of the target device in step S1 include: Collecting several groups of target device tension control training samples; each group of target device tension control training samples contains historical target device parameter information and corresponding historical real tension data; combining several groups of target device tension control training samples to obtain a target device tension control training set; Construct the initial target device tension fitting function. The initial target device tension fitting function is expressed by the formula F=f(X), where F is the simulated tension data, and X=[x1, x2, …, x N ], X represents each parameter in a set of target device parameter information, N is the number of each parameter in the target device parameter information, and f(X) is the functional relationship of the initial target device tension fitting function; Construct K target equipment tension fitting function individuals H k , each target device tension fitting function individual H k contains an initial target device tension fitting function, different target device tension fitting functions H k The functional relationship of the initial target device tension fitting function is different; the K target device tension fitting function individuals H k Combine to obtain the target equipment tension fitting function iterative population; set the maximum number of iterations; The target device tension fitting function individual H in the target device tension fitting function iteration population k , simulate the target equipment tension control training set and obtain the simulated tension data set J k , calculate the simulated tension data set J k The mean square error of the historical real tension data corresponding to the target equipment tension control training set is used to obtain the simulation tension calculation accuracy P k ; The simulation tension calculation accuracy P k The inverse of the target device tension fitting function is used as the individual H k The fitness S k ; For the target device tension fitting function iterative population, a population update operation is performed on the target device tension fitting function iterative population in each iteration to obtain a new target device tension fitting function iterative population; When the maximum number of iterations is reached, the target device tension fitting function individual H corresponding to the maximum fitness is output. k , which is the optimal target device tension fitting function individual; the initial target device tension fitting function in the optimal target device tension fitting function individual is used as the target device tension fitting function for analysis.

3. The intelligent tension control digitization method according to claim 2 is characterized in that: The target equipment tension prediction model in step S1 includes a data preprocessing layer, a prediction analysis layer, and a result output layer; The data preprocessing layer is used to preprocess the parameter information of the target device to be analyzed, and obtain the preprocessed parameter information of the target device to be analyzed; The prediction analysis layer is used to perform prediction analysis on the pre-processed target device parameter information to be analyzed, and obtain the predicted target device parameter information and predicted target device tension data within the next preset time interval; The result output layer is used to output the predicted target device parameter information and predicted target device tension data.

4. The method for digitizing intelligent tension control according to claim 3, characterized in that: The specific steps of training the prediction analysis layer in the target equipment tension prediction model include: Collecting several groups of continuous target device parameter information prediction samples; each group of target device parameter information prediction samples contains target device parameter information within the same preset time interval; combining several groups of continuous target device parameter information prediction samples to obtain a target device parameter information prediction training set; using the next group of target device parameter information prediction samples as the target value of the previous group of target device parameter information prediction samples; The target device parameter information prediction training set is input into the target device tension prediction model to train the prediction analysis layer with the target value as the target to obtain the initial prediction analysis layer; the initial prediction analysis layer is model evaluated to obtain the initial prediction analysis layer model evaluation result; if the initial prediction analysis layer model evaluation result is passed, the trained initial prediction analysis layer is output as the prediction analysis layer in the target device tension prediction model; otherwise, the target device parameter information prediction training set is used to continue model training.

5. The intelligent tension control digitization method according to claim 4 is characterized in that: The specific steps for determining the constant differential range of tension of the target equipment include: Collect a group of abnormal target equipment tension constant training samples B i , i=1,2,…,I; each group of target equipment tension constant training samples B i is a set of continuous target device tension data points, and the target device tension constant training sample B i The corresponding mean tension is the mean tension Z i ; The target equipment tension constant training sample B is group I i Using the Isolation Tree Algorithm to construct I isolated trees T i ; According to the isolated tree T i Calculate the anomaly scores of the child nodes in the anomaly score set C i ; Filter abnormal score set C i The target device tension data point corresponding to the abnormal score exceeding the abnormal score threshold is obtained, and the abnormal target device tension data point Y is obtained. i ; Calculate the abnormal target equipment tension data point Y i and the mean tension Z i The difference is the constant tension difference D of the target equipment. i ; Traverse all abnormal target equipment tension constant training samples B i , get the constant tension difference D of the target device i , set the tension of the target device to a constant difference value D i The minimum value is recorded as the target equipment tension constant difference range.

6. The intelligent tension control digitization method according to claim 5, characterized in that: The target device parameter control model in step S3 includes a data analysis layer, a parameter adjustment layer and a solution output layer; The data analysis layer is used to perform difference analysis on the predicted target device tension data and the constant difference range of the target device tension to obtain the target device tension data to be adjusted; The parameter adjustment layer is used to adjust the parameters according to the tension data of the target device to be adjusted and the parameter information of the target device to be analyzed by using the fuzzy control rules of the target device parameter control to obtain the target device parameter control scheme; the fuzzy control rules of the target device parameter control are used to output the target device parameter control scheme; The solution output layer is used to output the target device parameter control solution.

7. The method for digitizing intelligent tension control according to claim 6, characterized in that: The specific steps of training the parameter adjustment layer in the target device parameter control model include: Collecting several groups of tension target device parameter control training samples, each group of tension target device parameter control training samples includes tension input and parameter output, the tension input is the target tension adjustment value and the current parameter value, and the parameter output is the target parameter adjustment scheme; combining several groups of tension target device parameter control training samples to obtain a tension target device parameter control training set; Construct fuzzy rules according to the tension target device parameter control training set to obtain the fuzzy control rules for the target device parameter control to be optimized; input the tension input in the tension target device parameter control training sample into the fuzzy control rules for the target device parameter control to be optimized for simulation output to obtain the simulated parameter output; adjust the parameters according to the simulated parameter output and the current parameter value to obtain the simulated target device parameter information; input the simulated target device parameter information into the target device tension fitting function for calculation to obtain the simulated tension adjustment value; compare the simulated tension adjustment value with the target tension adjustment value to obtain the tension control difference; traverse the tension target device parameter control training set to obtain several groups of tension control difference values; Assuming that several groups of tension control differences are all within the error threshold range, the fuzzy control rule for parameter control of the target device to be optimized is used as the fuzzy control rule for parameter control of the target device; otherwise, the fuzzy control rule for parameter control of the target device to be optimized is feedback-adjusted to obtain a new fuzzy control rule for parameter control of the target device to be optimized, and the new fuzzy control rule for parameter control of the target device to be optimized is re-evaluated.

8. An intelligent tension control digitization system, characterized in that: The system applies an intelligent tension control digitization method according to any one of claims 1 to 7, including: The target device tension acquisition module is used to acquire the real-time tension information of the target device at every preset time interval to obtain the tension data to be analyzed, acquire the real-time parameter information of the target device, and obtain the parameter information of the target device to be analyzed; the parameter information of the target device to be analyzed includes various parameter information of the target device that affects the tension data to be analyzed; there is a functional relationship between the tension data to be analyzed and the parameter information of the target device to be analyzed, and the functional relationship is a tension fitting function of the target device; The target device tension prediction module is used to deploy the target device tension prediction model, input the target device parameter information to be analyzed into the target device tension prediction model for analysis, and obtain the predicted target device parameter information and the predicted target device tension data; the target device tension prediction model is used to predict the target device parameter information within the next preset time interval to obtain the predicted target device tension data; compare the tension data to be analyzed with the predicted target device tension data to determine whether it exceeds the target device tension constant difference range; if it exceeds the target device tension constant difference range, enter step S3; otherwise, continue to control according to the existing target device parameter information; The target device tension adjustment module is used to deploy the target device parameter control model, input the predicted target device tension data and the target device parameter information to be analyzed into the target device parameter control model for analysis, and obtain the target device parameter control plan; the target device parameter control model is used to output the target device parameter control plan based on the predicted target device parameter information and the target device parameter information to be analyzed so as to maintain the target device tension data within the constant difference range of the target device tension; and adjust the target device using the target device parameter control plan.

Images