Textile dyeing production quality management system based on multi-source data monitoring

Abstract

The application relates to the technical field of quality management systems, and particularly discloses a textile dyeing production quality management system based on multi-source data monitoring, which comprises the following modules: a critical recognition module that constructs a continuous airflow field model and identifies a vortex core area in combination with dynamic modal decomposition; a migration analysis module that maps the vortex core area to a dye concentration gradient field and identifies a vortex dominant linkage; a migration analysis module that couples and constructs an airflow-dye coupling transfer function, quantifies migration intensity and obtains a migration intensity index; an inter-cluster extraction module that extracts color difference surge clusters based on the migration intensity index and quantifies features; and a convergence control module that judges and controls effects through a convergence regression model, so as to realize dynamic convergence regulation. The present application integrates multi-source data and physical models, realizes intelligent monitoring of the whole process from migration risk early warning, driving mechanism identification to active control, and improves the uniformity of textile dyeing and the stability of production.

Description

Technical Field

[0001] The present invention relates to the technical field of quality management systems, and in particular to a textile dyeing production quality management system based on multi-source data monitoring. Background Art

[0002] In textile dyeing production, heat setting is a key process to ensure the quality of textile dyeing. During this process, the airflow state near the cloth surface directly affects the uniformity of dye distribution on the fabric surface. If there are disturbances such as irregular vortices in the airflow, it is easy to cause dye migration, resulting in color difference, color spots and other quality defects on the cloth surface.

[0003] Existing technologies have difficulty analyzing the correlation between airflow fields and dye migration. Failure to establish a dynamic linkage analysis mechanism between airflow vortex intensity and dye concentration gradients makes it impossible to distinguish whether migration is driven by airflow vortexes or caused by changes in the concentration gradient itself. This results in a lack of clear mechanistic basis for regulatory measures, making it easy to make unintentional interventions.

[0004] Existing technologies lack a quantitative model for the coupling effect between airflow and dye, making it difficult to accurately extract the migration trigger boundary and quantify the migration intensity, and unable to provide numerical support for risk assessment; the identification of possible chromatic aberration risk areas is vague, and existing technologies find it difficult to focus on key risk clusters, resulting in a wide control range and insufficient targeting. It is difficult to adjust the control strategy in real time according to changes in migration risk, affecting the stability of the control effect.

[0005] To this end, the present invention provides a textile dyeing production quality management system based on multi-source data monitoring. Summary of the Invention

[0006] The purpose of the present invention is to provide a textile dyeing production quality management system based on multi-source data monitoring to solve the above background problems.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] A textile dyeing production quality management system based on multi-source data monitoring, including the following modules:

[0009] Migration analysis module: used to collect critical warning signals to construct a dye concentration gradient field, identify the overlapping field between the vortex core area and the dye concentration gradient field, collect the airflow vortex intensity and local concentration characteristics of the grid within the overlapping field, and perform linkage analysis to determine whether the overlapping field is vortex-dominated linkage;

[0010] Migration analysis module: If the linkage is vortex-dominated, the airflow-dye coupling transfer function is constructed to extract the migration trigger boundary. If the grid is on the migration trigger boundary, the out-of-bounds transfer function value is extracted and integrated in the time domain to obtain the migration intensity index;

[0011] Inter-cluster extraction module: Builds a color difference emergence probability model based on the migration intensity index. The model outputs a three-dimensional probability distribution map, extracts color difference surge clusters from the three-dimensional probability distribution map, collects inter-cluster labeling coefficients of the color difference surge clusters, and constructs a surge cluster sequence.

[0012] Convergence control module: Based on the anchor point characteristics of the surging cluster sequence, an anti-phase strategy is constructed and the anti-phase airflow field is stimulated. The convergence distance of the convergence index after the anti-phase airflow field is stimulated is extracted, and a convergence regression model of the convergence distance is constructed to judge the dynamic convergence trend.

[0013] As a further solution of the present invention: a textile dyeing production quality management system based on multi-source data monitoring also includes the following modules:

[0014] Critical identification module: Constructs a spatiotemporal continuous airflow field and extracts the disturbance characteristic components of the modal energy density. The disturbance characteristic components are subjected to a migration critical analysis to obtain the vortex core area. If the vortex core area reaches the migration critical boundary, a critical warning signal is triggered.

[0015] As a further solution of the present invention: the method of performing the migration critical analysis is:

[0016] Obtain the velocity vector and monitoring time step of each grid point in the three-dimensional grid of the spatiotemporal continuous airflow field output, and construct a time series matrix containing the velocity vectors and monitoring time steps of all grid points;

[0017] Perform dynamic modal decomposition on the time series matrix to extract the modal energy density of different frequencies;

[0018] The modal energy density of the characteristic frequency is superimposed to construct the disturbance characteristic component;

[0019] The grid point velocity vector corresponding to the characteristic component of the disturbance is obtained, and the vortex area is located using the Q criterion in the vortex identification algorithm.

[0020] As a further solution of the present invention: the method of performing the linkage analysis is:

[0021] Obtain the frequency of occurrence of vortex-driven concentration gradient grids in the overlapping field under N monitoring time steps to obtain the vortex drive rate;

[0022] Obtain the frequency of occurrence of the grid of the concentration gradient-driven vortex in the overlapping field under N monitoring time steps to obtain the concentration drive rate;

[0023] For N monitoring time steps, the vortex drive entropy of the vortex drive rate and the concentrated drive entropy of the concentrated drive rate are calculated using the entropy algorithm;

[0024] Where N is the total number of monitoring steps;

[0025] The deviation ratio of the vortex drive rate to the concentration drive rate and the deviation ratio of the vortex drive entropy to the concentration drive entropy are calculated to determine whether the linkage mode is vortex-dominated linkage.

[0026] As a further solution of the present invention: the method of obtaining the vortex-driven concentration gradient and the concentration gradient-driven vortex is:

[0027] Mapping the vortex core area with the dye concentration gradient field to obtain the overlapping field of the mapped vortex core area and the concentration field;

[0028] Obtain the direction vector of the concentration gradient at each grid point in the overlapping field, calculate the modulus of the direction vector, and the Q value of the grid point;

[0029] Obtain the Q value at different monitoring time steps t and construct the vortex intensity series;

[0030] Obtain the modulus of the direction vector at different monitoring time steps t and construct the concentration feature sequence;

[0031] The lag time is calculated from the vortex intensity series and the concentration characteristic series through the cross-correlation function;

[0032] The vortex-driving concentration gradient is determined based on the lag time, and the concentration gradient drives the vortex.

[0033] As a further solution of the present invention: the migration intensity index is obtained by:

[0034] Obtain out-of-bounds grid points, establish the time decay contribution equation of a single grid point through time domain integration, and obtain the time decay contribution of a single grid point;

[0035] All out-of-bounds grid points in the overlapping field are traversed, and the time decay contributions of all out-of-bounds grid points are accumulated to obtain the migration intensity index.

[0036] As a further solution of the present invention: the method of obtaining the out-of-bounds grid point is:

[0037] Combining the Navier-Stokes equations and Fick's diffusion law, the airflow-dye coupling transfer function is constructed;

[0038] Obtain the eddy current modulus and transfer function value of each grid point in the overlapping field. If the grid point meets the grid point verification conditions, perform a spatial continuity test on the grid point and extract the connected area as the migration trigger boundary.

[0039] The grid points at the migration trigger boundary are regarded as out-of-bounds grid points.

[0040] As a further solution of the present invention, the method of collecting the inter-cluster marker coefficients of the chromatic aberration surge cluster is:

[0041] Obtain the growth rate of probability space clusters under multiple monitoring time steps, screen the probability space clusters based on the growth rate, and obtain the color difference surge clusters;

[0042] Calculate the ratio of the area of ​​the color difference surge cluster to the total area of ​​all color difference surge clusters to obtain the inter-cluster area ratio;

[0043] Calculate the mean probability of color difference emergence within a single color difference surge cluster and the mean probability of all color difference surge clusters;

[0044] The probability ratio between clusters is obtained by performing ratio processing on the mean probability of color difference emergence within a single color difference surge cluster and the mean probability of all color difference surge clusters.

[0045] The inter-cluster area ratio and the inter-cluster probability ratio are summed to obtain the inter-cluster labeling coefficient.

[0046] As a further solution of the present invention: the probability space cluster is obtained as follows:

[0047] A color difference emergence probability model is constructed. Based on the color difference emergence probability model, the color difference emergence probability of each out-of-bounds grid point in the overlapping field is obtained, and a three-dimensional probability distribution map containing the coordinates of the out-of-bounds grid points is constructed.

[0048] The multi-target tracking algorithm is used to obtain the spatiotemporal motion trajectory of the vortex core in the overlapping field of each monitoring time step t, and the dynamic influence radius of the spatiotemporal motion trajectory is extracted;

[0049] By setting the spatiotemporal coordinate filtering criteria, candidate coordinate points of the three-dimensional probability distribution map are extracted;

[0050] The candidate coordinate points obtained by screening are spatially clustered according to the color difference emergence probability to obtain multiple probability space clusters.

[0051] As a further solution of the present invention: the method for determining the dynamic convergence trend is:

[0052] Obtain the convergence index within K monitoring time steps after exciting the anti-phase airflow field and construct a convergence index vector;

[0053] Get the convergence index vector before K monitoring time steps as the anchor point index vector;

[0054] Extract the convergence index vector after K monitoring time steps, calculate the Euclidean distance between the convergence index vector and the anchor point index vector, and obtain the convergence distance;

[0055] Collect M convergence index vectors and their convergence distances with the anchor point index vector, build a convergence regression model of the convergence distance using a linear regression model algorithm, and calculate the slope of the convergence regression equation;

[0056] The trend of dynamic convergence is determined based on the slope of the convergence regression equation.

[0057] Beneficial effects of the present invention:

[0058] (1) The dynamic data of airflow on the fabric surface is collected through an array of micro hot wire anemometers. The radial basis function interpolation is combined to construct a three-dimensional spatiotemporal continuous airflow field. The high-frequency disturbance characteristic components are extracted using dynamic mode decomposition. The Q criterion is then used to accurately identify the vortex core area and determine whether it has reached the critical migration boundary. Based on multi-source data fusion and fluid dynamics analysis, it is beneficial to early identify the potential risk of dye migration caused by airflow vortexes during dyeing heat setting, achieve early warning of risks, and reduce the potential development of risks that may lead to subsequent dyeing quality problems.

[0059] (2) The vortex core area is mapped to the dye concentration gradient field. The linkage relationship between the airflow vortex intensity and the local concentration characteristics is extracted through cross-correlation analysis to clarify the vortex-dominated linkage. By establishing the spatial correlation and dynamic response analysis of the airflow field and concentration field, the driving mechanism of dye migration can be distinguished, providing a clear mechanism basis for subsequent targeted regulation, avoiding resource waste or regulation failure caused by blind intervention.

[0060] (3) The airflow-dye coupling transfer function is constructed. The migration trigger boundary is extracted by setting the vortex mode and the transfer function threshold, and the migration intensity index is obtained by time-domain integration of the out-of-bounds transfer function value. Based on the physical field coupling model and quantitative calculation, the risk level of dye migration can be quantitatively characterized, which is conducive to reflecting the intensity and range of migration and providing quantifiable data analysis support for locating key risk areas. The color difference emergence probability model is constructed through Gaussian process regression to generate a three-dimensional probability distribution map. The color difference surge clusters are then screened by combining the vortex motion trajectory and the dynamic influence radius, and the inter-cluster labeling coefficient is calculated to construct a sequence. Through the probability model and spatial cluster analysis, the high-risk color difference surge areas in the dyeing process can be focused on, the characteristics and priority of each risk cluster can be clarified, and indiscriminate regulation of the entire area can be avoided, thereby improving the targetedness and efficiency of risk treatment.

[0061] (4) The anti-phase airflow field is constructed by the anchor point characteristics of the chromatic aberration surge cluster, and the convergence distance changes after the anti-phase airflow field is stimulated are dynamically monitored through the convergence regression model. The closed-loop logic of risk characteristics, control strategy, and effect feedback can implement active control based on specific risk characteristics, and at the same time, judge the dynamic trend of the control effect in real time, which is conducive to effectively suppressing the risk of dye migration and ensuring the stability of the dyeing process. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The present invention will be further described below with reference to the accompanying drawings.

[0063] Figure 1This is a module diagram of a textile dyeing production quality management system based on multi-source data monitoring according to the present invention;

[0064] Figure 2 This is a flow chart for determining the driving mode in the present invention;

[0065] Figure 3 The present invention is a flowchart of a textile dyeing production quality management method based on multi-source data monitoring. DETAILED DESCRIPTION

[0066] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0067] Example 1

[0068] See also Figure 1 As shown, the present invention is a textile dyeing production quality management system based on multi-source data monitoring, which includes the following modules:

[0069] Critical Identification Module: This module constructs a spatiotemporal continuous airflow field during textile dyeing and heat setting, extracts the disturbance characteristic components of the modal energy density, performs migration criticality analysis on the disturbance characteristic components, and obtains the vortex core area. If the vortex core area reaches the migration critical boundary, a critical warning signal is triggered.

[0070] Among them, the method of constructing a spatiotemporal continuous airflow field in textile dyeing heat setting and extracting the disturbance characteristic component of the modal energy density is as follows:

[0071] Preferably, during the textile dyeing heat setting process, multi-source dynamic data of the airflow near the fabric surface is extracted by an array of micro hot wire anemometers deployed in the heat setting equipment;

[0072] Among them, the multi-source dynamic data includes the velocity value and direction of the airflow;

[0073] The collected multi-source dynamic data are pre-processed and interpolated with radial basis functions in combination with the structural parameters of the heat setting equipment to generate a spatiotemporal continuous airflow field of a three-dimensional grid in the fabric surface area;

[0074] Those skilled in the art will appreciate that the collected multi-source dynamic data is first pre-processed to remove noise, calibrate sensor errors, and unify the data format to ensure the accuracy and consistency of the original data. Subsequently, the spatial coordinate boundaries of the fabric area and the size and node distribution density of the three-dimensional grid of the meshing are determined in combination with the structural parameters of the heat setting equipment cavity size, the fabric transmission path, and the anemometer array installation position.

[0075] Based on the above pre-processed data and spatial boundaries, the radial basis function interpolation method is adopted. Based on the dynamic data of discrete anemometer collection points, a spatial interpolation model is constructed through radial basis functions. The model uses the mapping relationship between sampling point data and spatial distance to calculate the velocity vector (including x, y, and z direction components) of each node in the three-dimensional grid, and combines time series data to supplement the dynamic changes in velocity at different timestamps. Finally, a three-dimensional airflow field model covering the surface area and containing continuous spatiotemporal velocity information of all grid points is generated, transforming discrete monitoring data into a continuous physical field.

[0076] The spatiotemporal continuous airflow field outputs the velocity vector and monitoring time step of each grid point in the three-dimensional grid, and constructs a time series matrix containing the velocity vectors and monitoring time steps of all grid points;

[0077] Perform dynamic modal decomposition on the time series matrix to extract the modal energy density of different frequencies;

[0078] Filter the frequencies to extract frequencies above a preset critical frequency range to obtain characteristic frequencies;

[0079] The modal energy density of the characteristic frequency is superimposed to construct the disturbance characteristic component;

[0080] Among them, the method of performing migration critical analysis on the disturbance characteristic component to obtain the vortex core area is as follows:

[0081] Obtain the grid point velocity vector corresponding to the disturbance characteristic component and locate the vortex area using the Q criterion in the vortex identification algorithm;

[0082] It will be understood by those skilled in the art that the velocity vector (including the velocity components in the x, y, and z directions) of the corresponding space-time grid point is extracted from the high-frequency disturbance characteristic components obtained by modal decomposition, and the velocity gradient tensor is approximately calculated for each grid point by the difference method. The tensor is then decomposed into an antisymmetric rotation tensor Ω representing the rotational motion and a symmetric strain tensor S representing the tensile and compressive strains.

[0083] By formula: Get the Q value. When the Q value is greater than 0, it reflects the relative strength of the rotation effect and strain effect at the grid point.

[0084] In the entire space-time continuous airflow field composed of grid points, all grid points with Q values ​​greater than 0 are classified into one category, because the rotation effect of these points exceeds the strain effect. The area they form together is the vortex area identified by the Q criterion. The size of the Q value can represent the vortex intensity.

[0085] In the vortex area, the formula: Get eddy current mode ;

[0086] in, They are the vorticity components in the x-, y-, and z-directions respectively. The velocity vector of the vorticity passing through the grid point is V, and the three components of the velocity curl are calculated by the velocity difference between adjacent grids. , get the vorticity of the grid point;

[0087] Obtaining the local maximum of the vortex modulus in the vortex zone to locate the vortex core point, and connecting the vortex core points to form a continuous vortex core zone;

[0088] Calculate the maximum vortex mode in the vortex core area ,like , it is determined that the vortex core area has reached the critical migration boundary, and a critical warning signal is triggered;

[0089] in, is the eddy mode of the critical boundary of migration;

[0090] It is understood that the purpose of identifying the vortex core area is to:

[0091] Objective 1: To achieve early warning of dye migration risks. By identifying the vortex core area and determining whether it has reached the critical migration boundary, such as whether the maximum vortex modulus in the vortex core area exceeds the critical value, a critical warning signal can be triggered in time to detect the potential risk of dye migration caused by airflow vortexes during dyeing heat setting in advance, reducing the potential development of risks that may lead to subsequent quality problems such as color difference and color fringing on the fabric.

[0092] Objective 2: To provide a spatial correlation basis for clarifying the driving mechanism of dye migration. The vortex core area needs to be mapped to the dye concentration gradient field. By extracting the airflow vortex intensity and local concentration characteristics of the concentration gradient field, a linkage analysis can be performed to determine whether the migration is vortex-dominated. This provides a key spatial reference for distinguishing whether the migration is driven by vortices or the concentration gradient itself.

[0093] Objective 3: Provide a targeted region for subsequent risk quantification and precise control. The identified vortex core serves as the foundation for the migration analysis module to construct the airflow-dye coupling transfer function and extract the migration trigger boundary. It also serves as the spatial reference for the inter-cluster extraction module to locate chromatic aberration surge clusters and the convergence control module to construct the anti-phase airflow field. This ensures that subsequent risk intensity quantification, key risk area location, and targeted control can precisely focus on the core region affected by the vortex.

[0094] Migration analysis module: used to collect critical warning signals to construct a dye concentration gradient field, identify the overlapping field between the vortex core area and the dye concentration gradient field, collect the airflow vortex intensity and local concentration characteristics of the grid within the overlapping field, and perform linkage analysis to determine whether the overlapping field is vortex-dominated linkage;

[0095] The critical warning signal is collected to construct the dye concentration gradient field, and the overlapping field between the vortex core area and the dye concentration gradient field is identified as follows:

[0096] Preferably, the colorimeter is used to collect the colorimetric information of discrete sampling points set on the fabric surface, and the colorimetric value is converted into the dye concentration by establishing a colorimetric-concentration calibration model;

[0097] Obtain the concentration of discrete sampling points and establish the dye concentration gradient field through spatial interpolation algorithm;

[0098] Mapping the vortex core area with the dye concentration gradient field to obtain the overlapping field of the mapped vortex core area and the concentration field;

[0099] The method for collecting the airflow vortex intensity and local concentration characteristics of the grids in the overlapping field and performing linkage analysis is as follows:

[0100] Obtain the direction vector of the concentration gradient at each grid point in the overlapping field, calculate the modulus C of the direction vector, and the Q value of the grid point;

[0101] Obtain the Q value at different monitoring time steps t and construct the vortex intensity sequence ;

[0102] Where n is the total number of monitoring times;

[0103] Obtain the modulus of the direction vector at different monitoring time steps t and construct the concentration feature sequence ;

[0104] It can be understood that the magnitude of the direction vector of the concentration gradient serves as a local concentration characteristic;

[0105] Extract each pair of time series from the vortex intensity series and concentration feature series Calculate lag time by cross-correlation function Under the correlation, get different lag time Correlation coefficient , and perform significance test to exclude randomly associated grids;

[0106] in, Indicates lag time The correlation coefficient under

[0107] It is understandable that by analyzing the dynamic response relationship between the two in the time dimension, such as whether the change in vortex intensity precedes the change in concentration gradient, or vice versa, the driving direction between the airflow vortex and the dye concentration gradient can be clarified, that is, whether the vortex drives the concentration gradient change, or the concentration gradient itself drives the vortex. This provides a quantitative basis for judging whether the linkage mode of dye migration is vortex-dominated linkage, distinguishes the key steps of the migration driving mechanism, and focuses on the real driving factors for subsequent migration analysis and the formulation of control strategies, reducing control deviations caused by misjudgment of the driving mechanism.

[0108] Get the lag time of significance of each grid after excluding random correlation grids , and determine the driving mode;

[0109] like Figure 2 As shown, if the lag time >0, it means Before changes, i.e., vortex-driven concentration gradients;

[0110] If the lag time <0, it means Before changes, i.e., the concentration gradient drives the vortex;

[0111] Obtain the frequency of occurrence of vortex-driven concentration gradient grids in the overlapping field under N monitoring time steps to obtain the vortex drive rate;

[0112] Obtain the frequency of occurrence of the grid of the concentration gradient-driven vortex in the overlapping field under N monitoring time steps to obtain the concentration drive rate;

[0113] For N monitoring time steps, the vortex drive entropy of the vortex drive rate and the concentrated drive entropy of the concentrated drive rate are calculated using the entropy algorithm;

[0114] The entropy algorithm includes: information entropy algorithm. The entropy algorithm calculates the discreteness and distribution uniformity of the two sets of frequency sequences to obtain vortex drive entropy and concentrated drive entropy. The entropy value reflects the stability of the corresponding driving mode in the time dimension: the higher the entropy value, the more dispersed the frequency distribution of the driving mode at different time steps, the greater the fluctuation, and the more unstable the driving mode; the lower the entropy value, the more concentrated the frequency distribution, the smaller the fluctuation, and the more stable the driving mode. It provides a quantitative basis for reflecting the time dynamic characteristics of the driving mode by comparing the deviation ratio of vortex drive entropy and concentrated drive entropy to determine whether the linkage mode is vortex-dominated.

[0115] Where N is the total number of monitoring steps;

[0116] Calculate the deviation ratio of the vortex drive rate relative to the concentration drive rate, and the deviation ratio of the vortex drive entropy relative to the concentration drive entropy to determine whether the linkage mode is vortex-dominated linkage;

[0117] For example, assume that in the analysis of N = 10 monitoring time steps, the average total frequency (eddy drive rate) of the grids with eddy-driven concentration gradients appearing in the overlapping field is 55%, that is, within 10 time steps, on average 55% of the overlapping grids show the characteristics of eddy-driven concentration gradients;

[0118] The average total frequency of the grids of concentration gradient driven vortices appearing in the overlapping field is 25%, that is, on average, 25% of the overlapping grids show the characteristics of concentration gradient driven vortices;

[0119] Calculated by entropy algorithm: the vortex drive entropy corresponding to the distribution discreteness of vortex drive rate in 10 time steps is 0.7; the concentration drive entropy corresponding to the distribution discreteness of concentration drive rate in 10 time steps is 1.1;

[0120] The deviation of the vortex drive rate relative to the concentration drive rate is (55%-25%) / 25%×100%=120%, indicating that the vortex drive rate is significantly higher than the concentration drive rate. The deviation of the vortex drive entropy relative to the concentration drive entropy is (0.7-1.1) / 1.1×100%≈-36.4%, indicating that the vortex drive entropy is lower than the concentration drive entropy, that is, the temporal stability of the vortex drive mode is better than that of the concentration gradient drive mode.

[0121] In summary, the vortex drive rate accounts for a higher proportion and has a more stable time distribution, which indicates that the linkage mode is vortex-dominated linkage;

[0122] It can be understood that the role of determining whether the linkage mode is vortex-dominated linkage is:

[0123] Function 1: Provides a basis for determining when to activate the migration analysis module, ensuring that subsequent airflow-dye coupling transfer function construction, migration trigger boundary extraction, and migration intensity index calculation are performed only for dye migration driven by vortexes. This reduces ineffective quantitative analysis of non-vortex-driven migration (such as that driven by the concentration gradient itself), thereby improving the pertinence of risk assessment.

[0124] Function 2: Clarify the core driving mechanism of dye migration and provide a key basis for the convergence control module to formulate precise control strategies. If it is determined to be a vortex-dominated linkage, the anti-phase strategy can focus on the anchor point characteristics of the vortex (such as motion trajectory, rotation direction, etc.) to construct an anti-phase airflow field, reduce blind intervention in non-dominant factors, ensure the matching of control measures with risk-driven mechanisms, and improve control efficiency.

[0125] Example 2

[0126] like Figure 1 As shown, the present invention is a textile dyeing production quality management system based on multi-source data monitoring, which also includes the following modules:

[0127] Migration analysis module: If the linkage is vortex-dominated, the airflow-dye coupling transfer function is constructed to extract the migration trigger boundary. If the grid is on the migration trigger boundary, the out-of-bounds transfer function value is extracted and integrated in the time domain to obtain the migration intensity index;

[0128] The method for constructing the airflow-dye coupling transfer function and extracting the migration triggering boundary is as follows:

[0129] Combining the Navier-Stokes equations and Fick's diffusion law, the formula is:

[0130] Construct airflow-dye coupling transfer function;

[0131] in, is the disturbance velocity vector, are the disturbance characteristic components of the grid eddy current mode and modal energy density respectively;

[0132] is the grid average velocity vector, is the dye concentration gradient vector;

[0133] is the comprehensive diffusion coefficient, is the Laplace operator of concentration, that is, the second-order rate of change of concentration in three dimensions of space;

[0134] It can be understood that the purpose of constructing the airflow-dye coupling transfer function is:

[0135] Function 1: Provide a quantitative basis for extracting the migration trigger boundary. By constructing the airflow-dye coupling transfer function and combining it with the preset eddy current mode threshold and transfer function threshold, grid points that meet the grid point inspection conditions can be screened out. After the spatial continuity test, the migration trigger boundary is determined and the key area that triggers dye migration is extracted.

[0136] Function 2: Provides a basis for quantifying dye migration intensity. The function can calculate the transfer function value of the out-of-bounds grid point. By integrating these values ​​in the time domain, the migration intensity index can be obtained, which can quantitatively characterize the risk of dye migration under vortex dominance and reflect the dynamic impact of migration.

[0137] Function 3: Establish a dynamic correlation model between airflow and dye migration. The coupled transfer function integrates the Navier-Stokes equations (which describe airflow motion) and Fick's diffusion law (which describes dye diffusion). This mathematically quantifies the interaction between airflow disturbances (such as vortex intensity and disturbance velocity) and dye concentration gradients, reveals the coupling between the two, and provides theoretical support for analyzing the intrinsic mechanism of vortex-driven dye migration.

[0138] Obtain the eddy current modulus and transfer function value H of each grid point in the overlapping field. If the grid point meets the grid point test conditions: ;

[0139] in, and are the preset eddy current mode threshold and transfer function threshold respectively;

[0140] Then the spatial continuity test of the grid points is performed, and the connected areas are extracted as the migration triggering boundaries;

[0141] If the grid is at the migration trigger boundary, the out-of-bounds transfer function value is extracted and integrated in the time domain to obtain the migration intensity index as follows:

[0142] The grid points at the migration trigger boundary are regarded as out-of-bounds grid points, and the transfer function value of the out-of-bounds grid points, as well as the spatial coordinates and monitoring time step of the out-of-bounds grid points are obtained for each monitoring time step t.

[0143] For each out-of-bounds grid point, the time decay contribution equation of a single grid point is established by time domain integration: Get the time decay contribution of a single grid point ;

[0144] in, is the time decay coefficient, N is the total number of monitoring time steps, and i is the number of the monitoring time step;

[0145] t is the monitoring time step;

[0146] Traverse all the out-of-bounds grid points in the overlapping field and accumulate the time decay contributions of all the out-of-bounds grid points to obtain the migration intensity index I;

[0147] It should be noted that the overall intensity and cumulative impact of dye migration under vortex-dominated linkage are quantitatively characterized. The migration intensity index comprehensively reflects the cumulative degree of dye deviation from uniform distribution and abnormal migration driven by airflow vortex within a specific time range by integrating the airflow-dye coupling transfer function value of the out-of-bounds grid points within the migration trigger boundary in the time domain (superimposing the time decay contribution of each grid point). It includes the impact of key areas within the migration trigger boundary in space, and also reflects the persistence and attenuation characteristics of this migration effect in the time dimension. It provides a quantifiable physical measurement for intuitively evaluating the severity of vortex-driven dye migration and judging its potential threat to dyeing uniformity.

[0148] Inter-cluster extraction module: Builds a color difference emergence probability model based on the migration intensity index. The model outputs a three-dimensional probability distribution map, extracts color difference surge clusters from the three-dimensional probability distribution map, collects inter-cluster labeling coefficients of the color difference surge clusters, and constructs a surge cluster sequence.

[0149] Among them, the method of constructing the color difference emergence probability model based on the migration intensity index is:

[0150] By formula: Construct a color difference emergence probability model;

[0151] in, is the Gaussian process regression function, used to output probability distribution , is the model hyperparameter;

[0152] It is understood by those skilled in the art that when constructing a color difference emergence probability model based on the migration intensity index, the migration intensity index I is used as the input variable, and the probability modeling capability of Gaussian process regression (GP) is used in combination with the model hyperparameters (such as kernel function type, regularization coefficient, used to control model complexity and fitting characteristics), build mapping relationship : Gaussian process regression learns the nonlinear relationship between historical migration intensity and corresponding color difference emergence sample data. The final output L is the probability distribution of color difference emergence, which includes the expected probability of color difference emergence and also reflects the prediction uncertainty through the degree of dispersion of the distribution. It thus converts the continuous change of migration intensity into a probabilistic description of the risk of color difference emergence, providing support for the probabilistic assessment of dyeing quality.

[0153] Based on the color difference emergence probability model, the color difference emergence probability of each out-of-bounds grid point in the overlapping field is obtained, and a three-dimensional probability distribution map containing the coordinates of the out-of-bounds grid points is constructed;

[0154] The method of extracting the color difference surge clusters from the three-dimensional probability distribution map, collecting the inter-cluster labeling coefficients of the color difference surge clusters, and constructing the surge cluster sequence is as follows:

[0155] The spatiotemporal motion trajectory of the vortex core in the overlapping field of each monitoring time step t is obtained through a multi-target tracking algorithm, and the dynamic influence radius R(t) of the spatiotemporal motion trajectory is extracted;

[0156] Preferably, by the equation: Get the dynamic influence radius R(t);

[0157] By setting the spatiotemporal coordinate filtering criteria, candidate coordinate points of the three-dimensional probability distribution map are extracted;

[0158] Filtering criterion 1: retain the out-of-bounds grid points whose centers of the spatiotemporal motion trajectories within the overlapping field are lower than the dynamic influence radius;

[0159] Filtering criterion 2: retain out-of-bounds grid points whose gradient direction and vortex rotation direction are less than a preset deviation angle;

[0160] The candidate coordinate points obtained by screening are spatially clustered according to the probability of color difference emergence to obtain multiple probability space clusters;

[0161] Obtain the growth rate of probability space clusters under multiple monitoring time steps, screen the probability space clusters based on the growth rate, and obtain the color difference surge clusters;

[0162] Calculate the ratio of the area of ​​the color difference surge cluster to the total area of ​​all color difference surge clusters to obtain the inter-cluster area ratio;

[0163] Calculate the mean probability of color difference emergence within a single color difference surge cluster and the mean probability of all color difference surge clusters;

[0164] The probability ratio between clusters is obtained by performing ratio processing on the mean probability of color difference emergence within a single color difference surge cluster and the mean probability of all color difference surge clusters.

[0165] The inter-cluster area ratio and the inter-cluster probability ratio are summed to obtain the inter-cluster labeling coefficient;

[0166] It can be understood that the inter-cluster labeling coefficient is a comprehensive quantitative representation of the relative importance of a single color difference surge cluster among all risk clusters. The inter-cluster labeling coefficient reflects the relative scale of the cluster in spatial distribution through the inter-cluster area ratio, and reflects the relative intensity of the color difference risk within the cluster through the inter-cluster probability ratio. The superposition and summation of the two, combining the spatial proportion with the risk intensity, reflects the potential threat of a single color difference surge cluster to the overall dyeing quality, providing a quantifiable basis for distinguishing the priorities of different risk clusters and focusing on key risk areas, ensuring that subsequent regulatory interventions can be implemented in the risk clusters with the greatest impact.

[0167] Obtain the inter-cluster labeling coefficient of each chromatic aberration surge cluster and construct a surge cluster sequence;

[0168] Convergence control module: Based on the anchor point characteristics of the surge cluster sequence, an anti-phase strategy is constructed and an anti-phase airflow field is stimulated. The convergence distance of the convergence indicator after the anti-phase airflow field is stimulated is extracted, and a convergence regression model of the convergence distance is constructed to determine the dynamic convergence trend.

[0169] Among them, the method of exciting the anti-phase airflow field based on the surge cluster sequence is:

[0170] The motion trajectory of each color difference surge cluster in the surge cluster sequence, the rotation direction and vortex intensity of the cluster vortex, and the inter-cluster marking coefficient are extracted as the anchor point features of the anti-phase airflow field.

[0171] Among them, the vortex intensity includes the Q value and the mean value of the vortex modulus;

[0172] An anti-phase strategy is constructed based on the anchor point characteristics of the anti-phase airflow field, stimulating the anti-phase airflow field until the risk mode converges dynamically;

[0173] It can be understood that the construction of the anti-phase strategy based on the anchor point characteristics of the anti-phase airflow field needs to be based on the motion trajectory of the chromatic aberration surge cluster, the rotation direction of the vortex within the cluster, the vortex intensity (Q value and vortex modulus mean) and the inter-cluster marking coefficient: generate a counter-rotating airflow according to the vortex rotation direction. For example, if the original vortex is clockwise, the anti-phase airflow is counterclockwise to offset the rotation driving force; match the anti-phase airflow intensity (combined with the vortex intensity and the inter-cluster marking coefficient) to ensure that the energy is sufficient to neutralize the disturbance; at the same time, dynamically adjust the airflow action area following the motion trajectory of the surge cluster so that the anti-phase airflow continuously covers the risk area. Through the coordinated regulation of direction offset, intensity adaptation and trajectory following, the risk mode is accurately neutralized until dynamic convergence.

[0174] Obtain the convergence index within K monitoring time steps after exciting the anti-phase airflow field and construct a convergence index vector;

[0175] Get the convergence index vector before K monitoring time steps as the anchor point index vector;

[0176] Extract the convergence index vector after K monitoring time steps, calculate the Euclidean distance between the convergence index vector and the anchor point index vector, and obtain the convergence distance;

[0177] Collect M convergence index vectors and their convergence distances with the anchor point index vector, build a convergence regression model of the convergence distance using a linear regression model algorithm, and calculate the slope of the convergence regression equation;

[0178] Determine the trend of dynamic convergence based on the slope of the convergence regression equation;

[0179] It can be understood that if the slope is negative, it indicates that the convergence distance is gradually decreasing over time, that is, the difference between the current state and the anchor point state is shrinking, which means that the system state is developing in a more stable direction, that is, the effect of the anti-phase airflow field is effective, and the risk mode is being neutralized, which can be judged as convergence;

[0180] If the slope is close to 0, it means that the convergence distance changes very little over time and the difference is basically stable, which means that the system state has become stable and can be judged as close to convergence or has converged.

[0181] If the slope is positive, it means that the convergence distance increases over time, that is, the difference between the current state and the anchor point state is widening, which means that the effect of the anti-phase airflow field has not effectively suppressed the risk mode and may even aggravate it. It can be judged as non-convergence or deterioration.

[0182] Example 3

[0183] like Figure 3 As shown, the present invention is a textile dyeing production quality management method based on multi-source data monitoring, which also includes the following steps:

[0184] S1. Construct a spatiotemporal continuous airflow field in textile dyeing and heat setting, extract the disturbance characteristic component of the modal energy density, perform migration criticality analysis on the disturbance characteristic component, and obtain the vortex core area. If the vortex core area reaches the migration critical boundary, a critical warning signal is triggered;

[0185] S2. Collect critical warning signals to construct a dye concentration gradient field, identify the overlapping field between the vortex core area and the dye concentration gradient field, collect the airflow vortex intensity and local concentration characteristics of the grid within the overlapping field, and perform linkage analysis to determine whether the overlapping field is vortex-dominated linkage;

[0186] S3. If the linkage is vortex-dominated, construct the airflow-dye coupling transfer function and extract the migration trigger boundary. If the grid is on the migration trigger boundary, extract the transfer function value that crosses the boundary and perform time domain integration to obtain the migration intensity index.

[0187] S4. Constructing a color difference emergence probability model based on the migration intensity index, the model outputs a three-dimensional probability distribution map, extracting color difference surge clusters from the three-dimensional probability distribution map, collecting inter-cluster labeling coefficients of the color difference surge clusters, and constructing a surge cluster sequence;

[0188] S5. Based on the anchor point characteristics of the surging cluster sequence, an anti-phase strategy is constructed and the anti-phase airflow field is stimulated. The convergence distance of the convergence index after the anti-phase airflow field is stimulated is extracted, and a convergence regression model of the convergence distance is constructed to determine the dynamic convergence trend.

[0189] The above is a detailed description of an embodiment of the present invention. However, the content described is only a preferred embodiment of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. An intelligent monitoring system for textile dyeing production based on multi-source data monitoring, characterized by: Includes the following modules: Migration analysis module: used to collect critical warning signals to construct a dye concentration gradient field, identify the overlapping field between the vortex core area and the dye concentration gradient field, collect the airflow vortex intensity and local concentration characteristics of the grid within the overlapping field, and perform linkage analysis to determine whether the overlapping field is vortex-dominated linkage; Migration analysis module: If the linkage is vortex-dominated, the airflow-dye coupling transfer function is constructed to extract the migration trigger boundary. If the grid is on the migration trigger boundary, the out-of-bounds transfer function value is extracted and integrated in the time domain to obtain the migration intensity index; The method for constructing the airflow-dye coupling transfer function is: By formula: Construct airflow-dye coupling transfer function; in, is the disturbance velocity vector, are the disturbance characteristic components of the grid eddy current mode and modal energy density respectively; is the grid average velocity vector, is the dye concentration gradient vector; is the comprehensive diffusion coefficient, is the Laplace operator of concentration; The method for extracting the migration trigger boundary is as follows: Obtain the eddy current modulus and transfer function value of each grid point in the overlapping field. If the grid point meets the grid point verification conditions, perform a spatial continuity test on the grid point and extract the connected area as the migration trigger boundary. Inter-cluster extraction module: Builds a color difference emergence probability model based on the migration intensity index. The model outputs a three-dimensional probability distribution map, extracts color difference surge clusters from the three-dimensional probability distribution map, collects inter-cluster labeling coefficients of the color difference surge clusters, and constructs a color difference surge cluster sequence. Convergence control module: Based on the anchor point characteristics of the surging cluster sequence, an anti-phase strategy is constructed and the anti-phase airflow field is stimulated. The convergence distance of the convergence index after the anti-phase airflow field is stimulated is extracted, and a convergence regression model of the convergence distance is constructed to judge the dynamic convergence trend.

2. The intelligent monitoring system for textile dyeing production based on multi-source data monitoring according to claim 1, characterized in that: Also includes the following modules: Critical identification module: Constructs a spatiotemporal continuous airflow field and extracts the disturbance characteristic components of the modal energy density. The disturbance characteristic components are subjected to a migration critical analysis to obtain the vortex core area. If the vortex core area reaches the migration critical boundary, a critical warning signal is triggered.

3. The intelligent monitoring system for textile dyeing production based on multi-source data monitoring according to claim 2, characterized in that: The migration criticality analysis was performed as follows: Obtain the velocity vector and monitoring time step of each grid point in the three-dimensional grid of the spatiotemporal continuous airflow field output, and construct a time series matrix containing the velocity vectors and monitoring time steps of all grid points; Perform dynamic modal decomposition on the time series matrix to extract the modal energy density of different frequencies; The modal energy density of the characteristic frequency is superimposed to construct the disturbance characteristic component; The grid point velocity vector corresponding to the characteristic component of the disturbance is obtained, and the vortex area is located using the Q criterion in the vortex identification algorithm.

4. The intelligent monitoring system for textile dyeing production based on multi-source data monitoring according to claim 1, characterized in that: The linkage analysis is performed as follows: Obtain the frequency of occurrence of vortex-driven concentration gradient grids in the overlapping field under N monitoring time steps to obtain the vortex drive rate; Obtain the frequency of occurrence of the grid of the concentration gradient-driven vortex in the overlapping field under N monitoring time steps to obtain the concentration drive rate; For N monitoring time steps, the vortex drive entropy of the vortex drive rate and the concentrated drive entropy of the concentrated drive rate are calculated using the entropy algorithm; Where N is the total number of monitoring steps; The deviation ratio of the vortex drive rate to the concentration drive rate and the deviation ratio of the vortex drive entropy to the concentration drive entropy are calculated to determine whether the linkage mode is vortex-dominated linkage.

5. The intelligent monitoring system for textile dyeing production based on multi-source data monitoring according to claim 4 is characterized in that: The method of obtaining the vortex-driven concentration gradient and the concentration gradient-driven vortex is as follows: Mapping the vortex core area with the dye concentration gradient field to obtain the overlapping field of the mapped vortex core area and the concentration field; Obtain the direction vector of the concentration gradient at each grid point in the overlapping field, calculate the modulus of the direction vector, and the Q value of the grid point; Obtain the Q value at different monitoring time steps t and construct the vortex intensity series; Obtain the modulus of the direction vector at different monitoring time steps t and construct the concentration feature sequence; The lag time is calculated from the vortex intensity series and the concentration characteristic series through the cross-correlation function; The vortex-driving concentration gradient is determined based on the lag time, and the concentration gradient drives the vortex.

6. The intelligent monitoring system for textile dyeing production based on multi-source data monitoring according to claim 1, characterized in that: The migration intensity index is obtained as follows: Obtain out-of-bounds grid points, establish the time decay contribution equation of a single grid point through time domain integration, and obtain the time decay contribution of a single grid point; All out-of-bounds grid points in the overlapping field are traversed, and the time decay contributions of all out-of-bounds grid points are accumulated to obtain the migration intensity index.

7. The intelligent monitoring system for textile dyeing production based on multi-source data monitoring according to claim 6, characterized in that: The method for obtaining the out-of-bounds grid point is: Combining the Navier-Stokes equations and Fick's diffusion law, the airflow-dye coupling transfer function is constructed; Obtain the eddy current modulus and transfer function value of each grid point in the overlapping field. If the grid point meets the grid point verification conditions, perform a spatial continuity test on the grid point and extract the connected area as the migration trigger boundary. The grid points at the migration trigger boundary are regarded as out-of-bounds grid points.

8. The intelligent monitoring system for textile dyeing production based on multi-source data monitoring according to claim 1, characterized in that: The method of collecting the inter-cluster marker coefficients of the chromatic aberration surge cluster is: Obtain the growth rate of probability space clusters under multiple monitoring time steps, screen the probability space clusters based on the growth rate, and obtain the color difference surge clusters; Calculate the ratio of the area of ​​the color difference surge cluster to the total area of ​​all color difference surge clusters to obtain the inter-cluster area ratio; Calculate the mean probability of color difference emergence within a single color difference surge cluster and the mean probability of all color difference surge clusters; The probability ratio between clusters is obtained by performing ratio processing on the mean probability of color difference emergence within a single color difference surge cluster and the mean probability of all color difference surge clusters. The inter-cluster area ratio and the inter-cluster probability ratio are summed to obtain the inter-cluster labeling coefficient.

9. The intelligent monitoring system for textile dyeing production based on multi-source data monitoring according to claim 8, characterized in that: The probability space cluster is obtained as follows: A color difference emergence probability model is constructed. Based on the color difference emergence probability model, the color difference emergence probability of each out-of-bounds grid point in the overlapping field is obtained, and a three-dimensional probability distribution map containing the coordinates of the out-of-bounds grid points is constructed. The multi-target tracking algorithm is used to obtain the spatiotemporal motion trajectory of the vortex core in the overlapping field of each monitoring time step t, and the dynamic influence radius of the spatiotemporal motion trajectory is extracted; By setting the spatiotemporal coordinate filtering criteria, candidate coordinate points of the three-dimensional probability distribution map are extracted; The candidate coordinate points obtained by screening are spatially clustered according to the color difference emergence probability to obtain multiple probability space clusters.

10. The intelligent monitoring system for textile dyeing production based on multi-source data monitoring according to claim 1, characterized in that: The method for determining the dynamic convergence trend is: Obtain the convergence index within K monitoring time steps after exciting the anti-phase airflow field and construct a convergence index vector; Get the convergence index vector before K monitoring time steps as the anchor point index vector; Extract the convergence index vector after K monitoring time steps, calculate the Euclidean distance between the convergence index vector and the anchor point index vector, and obtain the convergence distance; Collect M convergence index vectors and their convergence distances with the anchor point index vector, build a convergence regression model of the convergence distance using a linear regression model algorithm, and calculate the slope of the convergence regression equation; The trend of dynamic convergence is determined based on the slope of the convergence regression equation.

Images