Intelligent hosiery machine control method and system based on Internet of Things
By using multi-step Nystrom algorithm and support vector description algorithm on the cloud edge collaborative platform, combined with an adaptive Smith estimator, the frequent and delay problems of manual monitoring in smart sock control are solved, and efficient and accurate automated control is achieved, improving Production efficiency and product quality of sock machine.
Patent Information
- Application Number
- CN202411507641.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-28
- Publication Date
- 2024-11-26
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing smart sock machine control scheme requires frequent manual monitoring and adjustment of parameters during remote control, lack of fault identification and automatic correction capabilities, resulting in insufficient control accuracy, increasing manual maintenance costs, and a problem of equipment command delay, resulting in poor system adjustment. precise.
Using the Internet of Things-based smart sock machine control method, using the cloud-edge collaboration platform, an approximate kernel matrix is constructed through a multi-step Nystrom algorithm, combining the support vector description algorithm and an adaptive Smith estimator, identify abnormal parameters and perform parameter corrections, considering execution Mechanism and communication delays to achieve automated control.
It improves the sensitivity and accuracy of abnormal data identification, reduces the cost of manual intervention, ensures the timeliness and accuracy of control instructions, reduces the hysteresis effect, improves the rapid and automatic repair capability of the sock machine, and improves production efficiency.
Smart Images

Figure CN119020910A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of equipment control, and in particular to an intelligent sock machine control method and system based on the Internet of Things. Background Art
[0002] With the continuous upgrading of mechanical equipment, sock machines are also moving towards intelligent improvement in the production process. The intelligent sock machine control method based on the Internet of Things is to use the Internet of Things technology to connect the sock machine equipment to the network, and through real-time monitoring, remote control and data analysis, to achieve precise and automatic adjustment of parameters such as yarn tension and needle speed, to ensure efficient and stable sock production.
[0003] Through automated intelligent sock machine control, remote management and real-time data feedback, the production process is optimized, manual intervention is reduced, production efficiency is improved, and the accuracy of parameter adjustment during the production process is ensured, thereby improving the quality and consistency of sock production.
[0004] However, the existing intelligent sock machine control solution still requires manual frequent monitoring and adjustment of sock machine parameters during the remote control process. It does not have fault identification and automatic correction capabilities, and the degree of automation is insufficient, which increases the manual maintenance cost. Moreover, when remotely controlling the sock machine, the delay in the equipment receiving and executing instructions is not considered, resulting in insufficient control accuracy. The execution instructions of the sock machine cannot take effect immediately, resulting in inaccurate adjustment of the system to the real-time parameters, a lag effect, and an inability to correct errors in a timely manner and accurately execute remote instructions. Summary of the invention
[0005] In order to solve the technical problems that the intelligent sock machine control scheme in the prior art still needs manual frequent monitoring and adjustment of the sock machine parameters during the remote control process, has no fault identification and automatic correction capabilities, has insufficient automation, and increases the manual maintenance cost, and does not consider the delay in the device receiving and executing instructions when remotely controlling the sock machine, resulting in insufficient control accuracy, and the execution instructions of the sock machine cannot take effect immediately, resulting in inaccurate adjustment of the real-time parameters by the system, a lag effect, and failure to correct errors in time and accurately execute remote instructions, the present invention provides an intelligent sock machine control method and system based on the Internet of Things.
[0006] The technical solution provided by the embodiment of the present invention is as follows: First aspect An embodiment of the present invention provides an intelligent sock machine control method based on the Internet of Things, which is applied to a cloud-edge collaborative platform. The cloud-edge collaborative platform includes a sock machine terminal, an edge terminal, a cloud end, and a user terminal connected in sequence, wherein the sock machine terminal is connected to the cloud end; including: S1: Collecting multiple sets of normal operation data of sock machines producing target socks, wherein each set of normal operation data of sock machines includes multiple normal operation parameters of sock machines; S2: Upload the normal operation data of the sock machine to the cloud, and use the multi-step Nystrom algorithm to determine the approximate kernel matrix describing the normal operation parameters of each sock machine in the high-dimensional space; S3: Based on the approximate kernel matrix, an abnormal parameter recognition model based on the support vector description algorithm is constructed to determine the minimum hypersphere describing the normal state of the hosiery machine under the production of the target hosiery; S4: construct a parameter correction output model of a sock machine based on an adaptive Smith predictor, wherein the parameter correction output model of the sock machine includes an actuator delay term and a communication delay term between the actuator and the edge end; S5: Send the support vector description algorithm and the sock machine parameter correction output model to the edge end; S6: collecting real-time operation data of the sock machine, and inputting the real-time operation data of the sock machine into the abnormal parameter recognition model; S7: Determine whether there are abnormal parameters in the real-time operation data of the sock machine. If yes, input the real-time operation data and the abnormal parameters into the sock machine parameter correction output model and output the sock machine parameter correction. Otherwise, return to step S6. S8: According to the sock machine parameter correction amount, the sock machine actuator corresponding to the sock machine operating parameter is controlled to make adjustments.
[0007] Second aspect An embodiment of the present invention provides an intelligent sock machine control system based on the Internet of Things, comprising: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the intelligent sock machine control method based on the Internet of Things as described in the first aspect is implemented.
[0008] The third aspect An embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the method for controlling an intelligent sock machine based on the Internet of Things as described in the first aspect is implemented.
[0009] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least: In the present invention, based on the normal operation data of the sock machine, the multi-step Nystrom algorithm is first used in the cloud to determine the approximate kernel matrix describing the normal operation parameters of each sock machine in the high-dimensional space, and subsets are extracted from the large data set multiple times, and the kernel matrix of the entire data set is approximated using the better subset, which can significantly reduce the computing resources required for calculating the global kernel matrix, effectively avoid the dimension explosion problem, and improve the feasibility of the abnormal data recognition model established based on the approximate kernel matrix. The model covering most of the normal operation data is quickly calculated with the lowest possible computing overhead, and then the recognition sensitivity and accuracy of the abnormal data are reversely improved, and the accuracy of the sock machine control intervention timing is further improved. After identifying the abnormal data, the output model of the sock machine parameter correction amount based on the adaptive Smith predictor is constructed in combination with the actuator delay term and the communication delay term. The delay effect is considered when the sock machine is controlled, and the delay is compensated. In the case of delay, the actuator can still be accurately and timely controlled to weaken the hysteresis effect. Both models are set at the edge end close to the sock machine to ensure that the generated accurate control instructions correct errors in time and accurately execute remote instructions, improve the rapid automatic repair capability of the sock machine, greatly reduce the cost of frequent manual intervention, and increase the production efficiency of the sock machine. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0011] Figure 1 A schematic diagram of a flow chart of an intelligent sock machine control method based on the Internet of Things provided by an embodiment of the present invention; Figure 2 A schematic diagram of the structure of a cloud-edge collaboration platform provided by an embodiment of the present invention; Figure 3 A schematic structural diagram of an Internet of Things-based intelligent sock machine control system provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0012] The technical solution of the present invention is described below in conjunction with the accompanying drawings.
[0013] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.
[0014] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.
[0015] Reference Manual Attached Figure 1 , showing a flow chart of an intelligent sock machine control method based on the Internet of Things provided by an embodiment of the present invention.
[0016] Reference Manual Attached Figure 2 , showing a structural diagram of a cloud-edge collaboration platform provided by an embodiment of the present invention.
[0017] An embodiment of the present invention provides an intelligent sock machine control method based on the Internet of Things, which is applied to a cloud-edge collaborative platform. The cloud-edge collaborative platform includes a sock machine terminal, an edge terminal, a cloud end, and a user terminal connected in sequence, wherein the sock machine terminal is connected to the cloud end.
[0018] It should be noted that through the connection of the sock machine terminal, edge terminal, cloud terminal and user terminal, efficient data collection, real-time monitoring and remote control are achieved. At the same time, complex calculations and analysis can be performed in the cloud, the edge terminal provides fast response, and the user terminal can view and adjust in real time, which improves production efficiency, flexibility and system response speed.
[0019] The processing flow of the intelligent sock machine control method based on the Internet of Things may include the following steps: S1: Collect the normal operation data of multiple groups of sock machines producing target socks.
[0020] Each group of normal operating data of a sock machine includes a plurality of normal operating parameters of the sock machine.
[0021] It should be noted that the staff can determine the normal operation data of the sock machine based on the parameters of the sock machine that can meet the requirements when generating different types of socks and the parameters specified by the sock machine when leaving the factory. The more groups of normal operation data of the sock machine, the greater the tolerance for subsequent abnormal detection. Data that does not meet the normal data status will be identified and corrected to return to the normal operation data of the sock machine. The normal operation data of the sock machine defines the standard of the normal operation status of the sock machine and provides reference data for subsequent abnormal detection and parameter correction.
[0022] In a possible implementation manner, the normal operating parameters of the hosiery machine include normal yarn tension, normal needle cylinder speed, normal number of needles and normal needle density.
[0023] S2: Upload the normal operation data of the sock machine to the cloud, and combine the multi-step Nystrom algorithm to determine the approximate kernel matrix describing the normal operation parameters of each sock machine in the high-dimensional space.
[0024] Among them, the Nystrom algorithm, also known as the Nyström algorithm, is a technique for kernel matrix approximation, which is commonly used when processing large-scale data sets. The calculation of the kernel matrix usually involves high computational complexity, while the Nyström algorithm samples a subset of the data, calculates the kernel matrix of the subset, and then approximates the entire data through the kernel matrix, thereby significantly reducing the amount of calculation.
[0025] The approximate kernel matrix estimates the kernel matrix of the complete data set by partially sampling or simplifying the original data. The kernel matrix is an important tool for describing the similarity between data points, and usually has a high computational complexity. Although the kernel matrix generated by the Nyström algorithm is not completely accurate, it can effectively retain the main features of the original data.
[0026] By using the multi-step Nyström algorithm to generate an approximate kernel matrix, the system can significantly reduce the complexity of kernel matrix calculation while ensuring accuracy, effectively process large-scale sock machine operation data, and improve the efficiency and scalability of the model.
[0027] In a possible implementation, S2 specifically includes: S201: Initialize the number of iterations and select a data subset having all normal operation parameters of the sock machine from the normal operation data of the sock machine.
[0028] S202: Calculate the forward kernel matrix of the complete data set and the data subset, the self-kernel matrix of the data subset, and the inverse kernel matrix of the data subset and the complete data set respectively through a local interactive adaptive kernel function having a local kernel width term, a feature interaction weighting term, and an inter-feature weighting term, wherein the forward kernel matrix and the inverse kernel matrix are transposed matrices of each other, and the complete data set is the normal operation data of the sock machine.
[0029] Among them, the local kernel width term dynamically adjusts the kernel width through local density calculation to adapt to the distribution differences of different samples in local areas, so that the kernel function can adapt to the complexity of the data. Feature interaction weighting introduces interaction weighting between different features, emphasizing the relationship between features, rather than simply treating them as independent features, and capturing the complex nonlinear interactions between features through interaction weighting. Inter-feature weighting terms are used to capture the interactions between features. Gradient embedding enhances the kernel function's ability to capture complex data patterns, making the model more sensitive to mutations or edge features in the data, and can enhance the ability to capture complex patterns, making the model more sensitive to areas with large changes.
[0030] In a possible implementation, the local interactive adaptive kernel function is specifically:
[0031] in, Indicates the complete data set i Data points The first j Data points The similarity value between them, exp represents the natural exponential function, and Respectively k The dimension i Data points and j data points, d Indicates the total dimension of normal operation parameters of the sock machine. express k Dimensionally and The weighting factor between express k Dimensions and l The interaction weighting factors of the dimensions, express l The dimension j data points, and Respectively and The adaptive kernel width of and Respectively and The gradient in the feature space is the rate of change of the feature. represents the gradient embedding weight, represents the square of the two-norm, express The neighborhood set of represents the norm operation, express and The neighborhood set of express k Dimensional data exist The variance of express The variance of express l Dimensional data The variance of express and The covariance of .
[0032] It should be noted that the established local interactive adaptive kernel function can dynamically adjust the calculation method of the kernel function through local density, feature interaction weighting and gradient embedding, so that the kernel function can more flexibly adapt to different data distributions and capture complex nonlinear relationships, thereby improving the sensitivity of the model in detecting data pattern mutations or edge features. It can enhance the accuracy and robustness of the anomaly detection model established based on the approximate kernel matrix calculated by the kernel function, especially when processing complex and high-dimensional data.
[0033] S203: Based on the forward kernel matrix, the self-kernel matrix and the inverse kernel matrix, the approximate kernel matrix of the normal operation data of the sock machine is solved by the Nystrom algorithm:
[0034] in, Indicates i The approximate kernel matrix obtained after +1 iterations is Represents the complete dataset n With i +1 data subset collected m i+1 The kernel matrix between is the forward kernel matrix, for The transposed matrix of is the inverse kernel matrix, Representing a subset of data m i+1 The kernel matrix of itself.
[0035] S204: Increase the number of data points collected in the data subset, increase the number of iterations by one, re-execute steps S202 to S203, and output the approximate kernel matrix:
[0036] in, and Respectively represent i Second and i +The number of data points in the data subset at 1 iteration, Indicates the increase in the number of data points collected.
[0037] S205: Calculate the improvement of the approximate kernel matrix obtained under the current number of iterations:
[0038] in, Indicates i +1 approximate kernel matrix obtained by iteration Relative to the i The approximate kernel matrix obtained by iteration The improvement of the approximate kernel matrix is represents the Frobenius norm.
[0039] S206: When the improvement amount is less than the preset improvement amount, output the approximate kernel matrix obtained under the maximum number of iterations; otherwise, return to step S204.
[0040] It should be noted that those skilled in the art can set the preset improvement amount according to actual needs, and the present invention is not limited thereto.
[0041] Specifically, first, the iteration is initialized by sampling subsets, and the kernel matrix is calculated using the local interactive adaptive kernel function. Then, the approximate kernel matrix is solved by the Nyström algorithm. The amount of data in the subset is gradually increased and iteratively updated until the improvement of the approximate kernel matrix is less than the preset value, thereby outputting the optimal approximate kernel matrix. This process effectively reduces the computational complexity and ensures that data processing remains efficient and accurate on large-scale sock machine datasets.
[0042] It should be noted that due to the huge amount of data of the sock machine, if the data is calculated point by point, it will be difficult to solve the problem. The multi-step Nystrom algorithm is introduced. First, Nystrom approximation is performed on a small sampling subset, and then a second round of Nystrom approximation is performed on the initial approximation result on a larger range of samples. This layer-by-layer approximation method can further reduce the computational complexity and ensure that the retained approximate kernel matrix can summarize the data of the normal state of the sock machine to the greatest extent. It greatly reduces the computational complexity and avoids the problem of dimensionality explosion.
[0043] S3: Based on the approximate kernel matrix, an abnormal parameter identification model based on the support vector description algorithm is constructed to determine the minimum hypersphere that describes the normal state of the sock machine under the production of target socks.
[0044] Among them, the support vector description algorithm is an anomaly detection algorithm based on support vector machines. Its core idea is to construct a minimum hypersphere to surround most of the normal data points in the data set, and regard the points outside the sphere as abnormal data. It is suitable for scenarios with only normal samples and is used to detect abnormal samples that deviate from the normal state. By constructing an abnormal parameter identification model of the support vector description algorithm based on the approximate kernel matrix, it not only reduces the consumption of computing resources, but also can accurately detect abnormal data, ensuring that the system still maintains high efficiency and accuracy when processing large-scale data, can accurately determine the normal state range of the sock machine, and effectively detect abnormal situations in the production process, ensure timely correction, and improve product quality and production efficiency.
[0045] In a possible implementation, the abnormal parameter identification model is specifically:
[0046] in, represents the approximate kernel matrix i Data points With j Data points The similarity of Represents the approximate kernel matrix The self-core value of is the self-similarity, Representation Lagrange multipliers that contribute to the abnormal parameter identification model, Representation Lagrange multipliers that contribute to the abnormal parameter identification model, R and c denote the radius and center point of the smallest hypersphere, respectively. Representation The slack variables for the degree of deviation from the hypersphere, C represents the control slack variable The regularization parameter of It means taking the minimum value of the function R and c and , Indicates the data point to be identified. x Whether it is a normal data point decision function, if ,but x is a normal data point, if ,but x is an abnormal data point, express x and The similarity of Represents the Lagrange multiplier at the minimum value of the function .
[0047] It should be noted that the abnormal parameter identification model based on the approximate kernel matrix can maintain an accurate description of the similarity of data points while reducing the computational complexity, improving the scalability and rapid iteration capability of the abnormal parameter identification model. By minimizing the radius of the hypersphere, the model can effectively surround normal data points and has a high sensitivity to abnormal data points. The introduction of slack variables allows a certain degree of fault tolerance, avoids overfitting, and improves the robustness of the model, which is especially suitable for processing large-scale sock machine data.
[0048] S4: Construct a sock machine parameter correction output model based on the adaptive Smith predictor.
[0049] Among them, the sock machine parameter correction output model includes the actuator delay term and the communication delay term between the actuator and the edge end.
[0050] Among them, the Smith predictor is a control method used to compensate for the influence of time lag in the control system. The sock machine parameter correction output model is a model for calculating the required parameter correction when the parameters of the sock machine deviate from the normal state due to various reasons (such as equipment aging, yarn tension fluctuation, etc.) during the production process.
[0051] By constructing a parameter correction output model for the sock machine based on the adaptive Smith predictor, the influence of actuator delay and communication delay can be considered at the same time to achieve accurate parameter adjustment. This not only improves the real-time performance of the system, but also reduces the error accumulation caused by delay, thereby ensuring that the sock machine can quickly and accurately return to normal operation during the production process, effectively improving production efficiency and product quality.
[0052] In a possible implementation, S4 specifically includes: S401: Establishing a vector similarity metric function for screening the target sock machine normal operating data group by combining the local geometric embedding and curvature transformation value of the abnormal data to be processed.
[0053] Among them, local geometric embedding maps the vector to a low-dimensional space that captures the local geometric relationship. This embedding process can maintain the complex geometric characteristics of the vector in the high-dimensional space and reflect the local similarity through the Euclidean distance of the embedding space. The difference in the geometric curvature of two vectors is described by the change in curvature in the high-dimensional space. The change in curvature between vectors can reflect their similarity in the nonlinear space. Vectors with smaller curvature differences are more similar.
[0054] In a possible implementation, the vector similarity measure function is specifically:
[0055] in, Indicates the abnormal operation data group of the sock machine to which the abnormal data to be processed belongs The target sock machine normal operation data group The vector similarity of and Respectively and The embedding representation of represents geometric embedding, Description and Curvature integral of geometric nonlinear differences, Represents the curvature function used to describe the bending change state of the geometric structure on the path from one vector to another.
[0056] It should be noted that through geometric embedding, data in high-dimensional space is mapped to low-dimensional space. This embedding process maintains the local geometric structure of the data, so that the relationship between the originally complex high-dimensional features can be expressed in a more intuitive way in low-dimensional space. The embedding process captures the local similarity and geometric relationship between data points, so the similarity of vectors can more accurately reflect the local feature similarity while maintaining global features. The curvature integral further quantifies the nonlinear difference between them by calculating the curvature change of the geometric path between two vectors in the embedded space. The Riemann curvature tensor can capture the degree of curvature of the data in the high-dimensional space and reflect the complex geometric changes of the data in the feature space. By calculating the curvature change, the system can identify abnormal data with slight differences, especially in the data distribution changes in nonlinear space. Through the combination of geometric embedding and curvature integral, the vector similarity metric function can capture local and global geometric characteristics at the same time, and effectively measure the similarity between complex and high-dimensional data. Compared with traditional similarity calculation, this method is more refined, can more accurately determine the correction target of abnormal data, and improve the control accuracy of the sock machine.
[0057] Among them, the abnormal operation data group of the sock machine to which the abnormal data to be processed belongs is the data group of all actuators that maintain the operation of the sock machine collected after the abnormal data is detected, specifically the data group composed of yarn tension, needle cylinder speed, number of needles and needle density collected in this abnormal state. Specific geometric embedding methods include Isomap, LaplacianEigenmaps, t-SNE and UMAP. The curvature function is specifically the Riemann curvature tensor.
[0058] S402: Determine the target hosiery machine normal operation data group of the abnormal data to be processed according to the vector similarity measurement function, and output the parameter deviation value:
[0059] in, express t Time parameter deviation value, and They respectively represent the abnormal data to be processed and the expected operating parameters representing the correction targets of the abnormal data to be processed.
[0060] S403: Establishing a parameter correction output model of the sock machine including the actuator delay term and the communication delay term to compensate for the parameter deviation value:
[0061] in, s represents a complex frequency variable, Indicates that there is an actuator delay item and communication delay term The correction amount of the execution parameters of the sock machine under represents the intermediate parameter, represents the system gain describing the sensitivity of the sock machine, Indicates the delay time including the actuator and communication delay The total delay time of the sock machine execution, e represents a natural constant, T It represents the time constant of the sock machine that describes the dynamic response speed of the sock machine. The larger the time constant of the sock machine, the slower the system response. It represents the transfer function that describes the transformation relationship between the output data of the sock machine and the input data.
[0062] It should be noted that the established sock machine parameter correction output model can accurately compensate for the error caused by system response delay by considering the actuator and communication delay terms. By combining the similarity metric function of nonlinear geometric embedding and curvature transformation, it is ensured that the parameter correction can accurately determine the correction direction of abnormal parameters under different operating conditions. This model not only improves the accuracy of parameter adjustment, but also reduces the impact of delay on system control, ensuring the stability and response speed of the sock machine operation, thereby improving production efficiency.
[0063] S5: Send the support vector description algorithm and the sock machine parameter correction output model to the edge end.
[0064] It should be noted that by deploying the support vector description algorithm and parameter correction model on the edge, the delay in cloud data processing is reduced, the real-time and response speed of the system are improved, ensuring that parameter correction can be executed quickly and avoiding abnormal expansion in the production process due to delays.
[0065] S6: Collect the real-time operation data of the sock machine, and input the real-time operation data of the sock machine into the abnormal parameter identification model.
[0066] It should be noted that by collecting the operating data of the sock machine in real time and inputting it into the anomaly detection model, the status of the sock machine can be continuously monitored, dynamic adjustment can be achieved, the development of potential problems can be effectively prevented, and the stability of the production process and product quality can be ensured.
[0067] S7: Determine whether there are abnormal parameters in the real-time operation data of the sock machine. If yes, input the real-time operation data and the abnormal parameters into the sock machine parameter correction output model, and output the sock machine parameter correction. Otherwise, return to step S6.
[0068] It should be noted that this step can quickly determine whether the hosiery machine is abnormal when an abnormal state occurs, and combine the detected abnormal parameters with the actual operation data to immediately generate accurate correction values. This real-time identification and dynamic adjustment mechanism effectively avoids the abnormal state from lasting too long, reduces production failures and product failure rates, and ensures that the hosiery machine can operate efficiently and stably.
[0069] S8: According to the sock machine parameter correction amount, the sock machine actuator corresponding to the sock machine operating parameter is controlled to make adjustments.
[0070] It should be noted that by directly applying the calculated correction amount to the actuator, it is ensured that the adjustment of the sock machine is quick and effective, the abnormal state is corrected in time, the production process is restored to normal, and downtime and the generation of defective products are minimized.
[0071] In a possible implementation, the hosiery machine actuator specifically includes a tension controller relative to the yarn tension, a needle cylinder motor relative to the needle cylinder rotation speed, a needle position selector relative to the needle number, and a yarn feed servo motor relative to the needle density.
[0072] In the actual application process, starting from the collection of normal operation data, the multi-step Nyström algorithm is used to construct an approximate kernel matrix, the support vector description algorithm is used to establish an anomaly detection model, and the adaptive Smith predictor is combined to perform parameter correction. The operating status of the sock machine is monitored in real time. If an abnormality is detected, it is adjusted through the correction model to ensure that the sock machine always remains in a normal state. The advantage of this process is that it can monitor and adjust the operating parameters of the sock machine in real time and accurately, effectively improve production efficiency and product quality, and reduce failures and errors in the production process.
[0073] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least: In the present invention, based on the normal operation data of the sock machine, the multi-step Nystrom algorithm is first used in the cloud to determine the approximate kernel matrix describing the normal operation parameters of each sock machine in the high-dimensional space, and subsets are extracted from the large data set multiple times, and the kernel matrix of the entire data set is approximated using the better subset, which can significantly reduce the computing resources required for calculating the global kernel matrix, effectively avoid the dimension explosion problem, and improve the feasibility of the abnormal data recognition model established based on the approximate kernel matrix. The model covering most of the normal operation data is quickly calculated with the lowest possible computing overhead, and then the recognition sensitivity and accuracy of the abnormal data are reversely improved, and the accuracy of the sock machine control intervention timing is further improved. After identifying the abnormal data, the output model of the sock machine parameter correction amount based on the adaptive Smith predictor is constructed in combination with the actuator delay term and the communication delay term. The delay effect is considered when the sock machine is controlled, and the delay is compensated. In the case of delay, the actuator can still be accurately and timely controlled to weaken the hysteresis effect. Both models are set at the edge end close to the sock machine to ensure that the generated accurate control instructions correct errors in time and accurately execute remote instructions, improve the rapid automatic repair capability of the sock machine, greatly reduce the cost of frequent manual intervention, and increase the production efficiency of the sock machine.
[0074] Reference Manual Attached Figure 3 , showing a structural schematic diagram of an intelligent sock machine control system based on the Internet of Things provided by the present invention.
[0075] The present invention further provides an intelligent sock machine control system 20 based on the Internet of Things, which is applied to the above-mentioned intelligent sock machine control method based on the Internet of Things, and includes: Processor 201.
[0076] The memory 202 stores computer-readable instructions. When the computer-readable instructions are executed by the processor 201, the intelligent sock machine control method based on the Internet of Things as in the method embodiment is implemented.
[0077] The IoT-based smart sock machine control system 20 provided in the present invention can execute the above-mentioned IoT-based smart sock machine control method and achieve the same or similar technical effects. To avoid repetition, the present invention will not go into details.
[0078] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least: In the present invention, based on the normal operation data of the sock machine, the multi-step Nystrom algorithm is first used in the cloud to determine the approximate kernel matrix describing the normal operation parameters of each sock machine in the high-dimensional space, and subsets are extracted from the large data set multiple times, and the kernel matrix of the entire data set is approximated using the better subset, which can significantly reduce the computing resources required for calculating the global kernel matrix, effectively avoid the dimension explosion problem, and improve the feasibility of the abnormal data recognition model established based on the approximate kernel matrix. The model covering most of the normal operation data is quickly calculated with the lowest possible computing overhead, and then the recognition sensitivity and accuracy of the abnormal data are reversely improved, and the accuracy of the sock machine control intervention timing is further improved. After identifying the abnormal data, the output model of the sock machine parameter correction amount based on the adaptive Smith predictor is constructed in combination with the actuator delay term and the communication delay term. The delay effect is considered when the sock machine is controlled, and the delay is compensated. In the case of delay, the actuator can still be accurately and timely controlled to weaken the hysteresis effect. Both models are set at the edge end close to the sock machine to ensure that the generated accurate control instructions correct errors in time and accurately execute remote instructions, improve the rapid automatic repair capability of the sock machine, greatly reduce the cost of frequent manual intervention, and increase the production efficiency of the sock machine.
[0079] It should be understood that the processor in the embodiment of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.
[0080] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).
[0081] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.
[0082] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.
[0083] In the present invention, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.
[0084] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.
[0085] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.
[0086] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0087] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.
[0088] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0089] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
[0090] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.
[0091] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the method for controlling an intelligent sock machine based on the Internet of Things as described in the method embodiment is implemented.
[0092] A computer-readable storage medium provided by the present invention can implement the steps and effects of the intelligent sock machine control method based on the Internet of Things in the above method embodiment. To avoid repetition, the present invention will not go into details.
[0093] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least: In the present invention, based on the normal operation data of the sock machine, the multi-step Nystrom algorithm is first used in the cloud to determine the approximate kernel matrix describing the normal operation parameters of each sock machine in the high-dimensional space, and subsets are extracted from the large data set multiple times, and the kernel matrix of the entire data set is approximated using the better subset, which can significantly reduce the computing resources required for calculating the global kernel matrix, effectively avoid the dimension explosion problem, and improve the feasibility of the abnormal data recognition model established based on the approximate kernel matrix. The model covering most of the normal operation data is quickly calculated with the lowest possible computing overhead, and then the recognition sensitivity and accuracy of the abnormal data are reversely improved, and the accuracy of the sock machine control intervention timing is further improved. After identifying the abnormal data, the output model of the sock machine parameter correction amount based on the adaptive Smith predictor is constructed in combination with the actuator delay term and the communication delay term. The delay effect is considered when the sock machine is controlled, and the delay is compensated. In the case of delay, the actuator can still be accurately and timely controlled to weaken the hysteresis effect. Both models are set at the edge end close to the sock machine to ensure that the generated accurate control instructions correct errors in time and accurately execute remote instructions, improve the rapid automatic repair capability of the sock machine, greatly reduce the cost of frequent manual intervention, and increase the production efficiency of the sock machine.
[0094] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.
[0095] There are a few points to note: (1) The drawings of the embodiments of the present invention only involve structures related to the embodiments of the present invention. Other structures may refer to conventional designs.
[0096] (2) For the sake of clarity, in the drawings used to describe the embodiments of the present invention, the thickness of layers or regions is exaggerated or reduced, that is, these drawings are not drawn according to the actual scale. It is understood that when an element such as a layer, film, region or substrate is referred to as being "on" or "under" another element, the element may be "directly" "on" or "under" the other element or there may be intermediate elements.
[0097] (3) In the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to obtain new embodiments.
[0098] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be based on the protection scope of the claims.
Claims
1. A smart sock machine control method based on the Internet of Things, characterized in that: Applied to a cloud-edge collaborative platform, the cloud-edge collaborative platform includes a sock machine terminal, an edge terminal, a cloud end, and a user terminal connected in sequence, wherein the sock machine terminal is connected to the cloud end; the method includes: S1: Collecting multiple sets of normal operation data of sock machines producing target socks, wherein each set of normal operation data of sock machines includes multiple normal operation parameters of sock machines; S2: uploading the normal operation data of the sock machine to the cloud, and combining the multi-step Nystrom algorithm to determine the approximate kernel matrix describing the normal operation parameters of each sock machine in the high-dimensional space; S3: constructing an abnormal parameter recognition model based on a support vector description algorithm based on the approximate kernel matrix, and determining a minimum hypersphere describing a normal state of a sock machine when producing the target socks; S4: constructing a sock machine parameter correction output model based on an adaptive Smith predictor, wherein the sock machine parameter correction output model includes an actuator delay term and a communication delay term between the actuator and the edge end; S5: Sending the support vector description algorithm and the sock machine parameter correction output model to the edge end; S6: collecting real-time operation data of the sock machine, and inputting the real-time operation data of the sock machine into the abnormal parameter recognition model; S7: Determine whether there are abnormal parameters in the real-time operation data of the sock machine. If yes, input the real-time operation data of the sock machine and the abnormal parameters into the sock machine parameter correction output model and output the sock machine parameter correction. Otherwise, return to step S6. S8: controlling the sock machine actuator corresponding to the sock machine operating parameter to adjust according to the sock machine parameter correction amount.
2. The method for controlling an intelligent sock machine based on the Internet of Things according to claim 1, characterized in that: The normal operating parameters of the hosiery machine include normal yarn tension, normal needle cylinder speed, normal number of needles and normal needle density.
3. The method for controlling an intelligent sock machine based on the Internet of Things according to claim 1, characterized in that: The S2 specifically includes: S201: Initializing the number of iterations, selecting a data subset having all normal operating parameters of the sock machine from the normal operating data of the sock machine; S202: Calculate the forward kernel matrix of the complete data set and the data subset, the self-kernel matrix of the data subset, and the inverse kernel matrix of the data subset and the complete data set respectively through a local interactive adaptive kernel function having a local kernel width term, a feature interaction weighting term, and an inter-feature weighting term, wherein the forward kernel matrix and the inverse kernel matrix are transposed matrices of each other, and the complete data set is the normal operation data of the sock machine; S203: Based on the forward kernel matrix, the self-kernel matrix and the inverse kernel matrix, the approximate kernel matrix of the normal operation data of the sock machine is solved by the Nystrom algorithm: in, Indicates i The approximate kernel matrix obtained after +1 iterations is Represents the complete dataset n With i +1 data subset collected m i+1 The kernel matrix between is the forward kernel matrix, for The transposed matrix of is the inverse kernel matrix, Representing a subset of data m i+1 The own kernel matrix of S204: Increase the number of data points collected in the data subset, increase the number of iterations by one, re-execute steps S202 to S203, and output the approximate kernel matrix: in, and Respectively represent i Second and i +The number of data points in the data subset at 1 iteration, Indicates the increase in the number of data points collected; S205: Calculate the improvement of the approximate kernel matrix obtained under the current number of iterations: in, Indicates i +1 approximate kernel matrix obtained by iteration Relative to the i The approximate kernel matrix obtained by iteration The improvement of the approximate kernel matrix is represents the Frobenius norm; S206: When the improvement amount is less than a preset improvement amount, output the approximate kernel matrix obtained under the maximum number of iterations; otherwise, return to step S204.
4. The method for controlling an intelligent sock machine based on the Internet of Things according to claim 3 is characterized in that: The local interactive adaptive kernel function is specifically: in, Indicates the complete data set i Data points The first j Data points The similarity value between them, exp represents the natural exponential function, and Respectively k The dimension i Data points and j data points, d Indicates the total dimension of normal operation parameters of the sock machine. express k Dimensionally and The weighting factor between express k Dimensions and l The interaction weighting factors of the dimensions, express l The dimension j data points, and Respectively and The adaptive kernel width of and Respectively and The gradient in the feature space is the rate of change of the feature. represents the gradient embedding weight, represents the square of the two-norm, express The neighborhood set of represents the norm operation, express and The neighborhood set of express k Dimensional data exist The variance of express The variance of express l Dimensional data The variance of express and The covariance of .
5. According to the method for controlling an intelligent sock machine based on the Internet of Things according to claim 1, it is characterized in that: The abnormal parameter identification model is specifically: in, represents the approximate kernel matrix i Data points With j Data points The similarity of Represents the approximate kernel matrix The self-core value of is the self-similarity, Representation Lagrange multipliers that contribute to the abnormal parameter identification model, Representation Lagrange multipliers that contribute to the abnormal parameter identification model, R and c denote the radius and center point of the smallest hypersphere, respectively. Representation The slack variables for the degree of deviation from the hypersphere, C represents the control slack variable The regularization parameter of It means taking the minimum value of the function R and c and , Indicates the data point to be identified. x Whether it is a normal data point decision function, if ,but x is a normal data point, if ,but x is an abnormal data point, express x and The similarity of Represents the Lagrange multiplier at the minimum value of the function .
6. The method for controlling an intelligent sock machine based on the Internet of Things according to claim 1, characterized in that: The S4 specifically includes: S401: establishing a vector similarity metric function for screening a target hosiery machine normal operation data group by combining the local geometric embedding and curvature transformation value of the abnormal data to be processed; S402: Determine the target hosiery machine normal operation data group of the abnormal data to be processed according to the vector similarity measurement function, and output the parameter deviation value: in, express t Time parameter deviation value, and represent the abnormal data to be processed and the expected operating parameters representing the correction target of the abnormal data to be processed respectively; S403: Establishing a parameter correction output model of a sock machine including the actuator delay term and the communication delay term to compensate for the parameter deviation value: in, s represents a complex frequency variable, Indicates that there is an actuator delay item and communication delay term The correction amount of the execution parameters of the sock machine under represents the intermediate parameter, represents the system gain describing the sensitivity of the sock machine, Indicates the delay time including the actuator and communication delay The total delay time of the sock machine execution, e represents a natural constant, T It represents the time constant of the sock machine that describes the dynamic response speed of the sock machine. The larger the time constant of the sock machine, the slower the system response. It represents the transfer function that describes the transformation relationship between the output data of the sock machine and the input data.
7. The method for controlling an intelligent sock machine based on the Internet of Things according to claim 6, characterized in that: The vector similarity metric function is specifically: in, Indicates the abnormal operation data group of the sock machine to which the abnormal data to be processed belongs The target sock machine normal operation data group The vector similarity of and Respectively and The embedding representation of represents geometric embedding, Description and Curvature integral of geometric nonlinear differences, Represents the curvature function used to describe the bending change state of the geometric structure on the path from one vector to another.
8. The method for controlling an intelligent sock machine based on the Internet of Things according to claim 2, characterized in that: The sock machine actuator specifically includes a tension controller relative to the yarn tension, a needle cylinder motor relative to the needle cylinder rotation speed, a needle position selector relative to the needle number, and a yarn feeding servo motor relative to the needle density.
9. An intelligent sock machine control system based on the Internet of Things, characterized in that: include: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the intelligent sock machine control method based on the Internet of Things as claimed in any one of claims 1 to 8 is implemented.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, an intelligent sock machine control method based on the Internet of Things as described in any one of claims 1 to 8 is implemented.
Citation Information
Patent Citations
Support vector machine (SVM)-based kernel matrix approximation method
CN102426562A
Nuclear matrix approximation-based carbon dioxide absorption tower fault diagnosis method
CN110209145A
Intelligent hosiery machine control system based on Internet of Things technology and control method
CN113885451A
Power grid measurement data checking method and system based on artificial intelligence algorithm
CN118260541A
Reliability data anomaly detection and correction method based on big data and artificial intelligence
CN118260693A
Cited By
Intelligent control method and system for metal product production line
CN119644967A
