A strength testing system for lining cloth production

Abstract

The present invention discloses a strength testing system for lining cloth production, which is used in the detection field. The strength testing system comprises: a lining cloth sample cutting module, a tensile performance testing module, a tensile strength analysis module, an adhesive strength testing module, an environmental adaptability testing module and a data analysis and evaluation module. The present invention uses a hierarchical clustering algorithm to identify different performance groups on the basis of low-dimensional mapping, which can more accurately reflect the similarities and differences between samples. By statistically analyzing the data in each sample group, the tensile strength and adhesive strength performance of different samples can be compared and evaluated in detail. By analyzing the performance differences of different groups, the key factors affecting the tensile strength and adhesive strength can be identified, providing a scientific basis for material selection and design optimization.

Description

Technical Field

[0001] The invention relates to the field of detection, and in particular to a strength testing system for lining cloth production. Background Art

[0002] Interlining is a kind of fabric or nonwoven material usually used to reinforce other materials or structures. It is widely used in many fields such as clothing, shoes and hats, furniture, and automotive interiors. In the clothing industry, interlining is used in collars, cuffs, pockets and other parts to help clothes maintain their shape and structure; in automotive interiors, it is used in door panels, seats and ceilings to enhance beauty and durability; in furniture manufacturing, interlining is used to improve the wear resistance and comfort of fabric sofas and chairs.

[0003] Strength testing can verify whether the interlining meets specific quality standards and performance requirements. For application scenarios that require a certain hardness and durability (such as automotive interiors or heavy furniture), the strength of the interlining is a key factor. In some applications, the strength of the interlining is directly related to the safety of the final product, especially in the automotive and furniture industries. Interlinings with insufficient strength may cause the product to deform or damage during use, increasing the risk of accidents. Different industries have strict specifications for the strength of interlinings. Through strength testing, manufacturers can ensure that their products meet the safety and performance standards specified by the industry and avoid legal and quality issues caused by non-compliance. Through strength testing, manufacturers can evaluate the effects of different materials and production technologies, optimize production processes and material selection to improve the performance and cost-effectiveness of finished products.

[0004] Traditional tests cannot monitor and respond to changes in force and displacement in real time, resulting in untimely data updates and inability to accurately reflect the real-time behavior of the material. In systems without adaptive adjustment, the test speed is fixed and cannot be adjusted based on the material's response, which may result in long test times and low efficiency. Traditional methods may have difficulty accurately identifying and analyzing complex data structures, and it is difficult to accurately identify anomalies and trends from large amounts of data. In a test system that does not have adaptive adjustment capabilities, if the material suddenly deforms, it may not be able to adjust in time, which may cause sample overload or damage. Traditional analysis methods may not be able to accurately distinguish and identify groups of samples with different performance characteristics.

[0005] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0006] In order to overcome the above problems, the present invention aims to propose a strength testing system for lining cloth production, with the aim of solving the problems that traditional methods may be difficult to accurately identify and analyze complex data structures, and it is difficult to accurately distinguish abnormal points and trends from a large amount of data. In a test system that does not have adaptive adjustment capabilities, if the material suddenly deforms, it may not be adjusted in time, which may cause sample overload or damage, and traditional analysis methods may not be able to accurately distinguish and identify sample groups with different performance characteristics.

[0007] To this end, the specific technical solution adopted by the present invention is as follows:

[0008] A strength testing system for lining cloth production, the strength testing system comprising: a lining cloth sample cutting module, a tensile performance testing module, a tensile strength analysis module, an adhesive strength testing module, an environmental adaptability testing module and a data analysis and evaluation module;

[0009] The lining cloth sample cutting module, the tensile performance testing module, the tensile strength analysis module, the bonding strength testing module, the environmental adaptability testing module and the data analysis and evaluation module are connected in sequence;

[0010] Interlining sample cutting module, used to cut interlining samples according to preset standard sizes and shapes;

[0011] The tensile performance test module is used to dynamically adjust the speed of tensile force application according to the real-time monitored force and displacement data feedback by using an adaptive algorithm until the lining cloth sample breaks, and simultaneously record the force change and displacement data during the stretching process, and convert the force change and displacement data into stress-strain data;

[0012] The tensile strength analysis module is used to analyze stress-strain data using a density-sensitive anomaly detection algorithm, automatically identify and exclude abnormal data points, and calculate the maximum tensile strength of the lining cloth sample;

[0013] Adhesion strength test module, used to measure the stress distribution of the lining sample during the tearing process through the peeling test method, and evaluate the adhesion strength between the lining sample and the fabric;

[0014] Environmental adaptability test module, used to simulate and test the bonding strength of lining samples under various environmental conditions, and evaluate the performance of linings in various environments;

[0015] Data analysis and evaluation module, used to collect and integrate the data of the maximum tensile strength of the lining fabric sample and the bonding strength between the lining fabric sample and the fabric, and evaluate the strength performance of the lining fabric sample based on the integrated data;

[0016] The data analysis and evaluation module collects and integrates the data of the maximum tensile strength of the lining fabric sample and the bonding strength between the lining fabric sample and the fabric, and evaluates the strength performance of the lining fabric sample based on the integrated data, including:

[0017] Obtain the tensile strength data and bonding strength data of the lining cloth samples;

[0018] The collected tensile strength data and bonding strength data were mapped into a low-dimensional space using the t-distributed random neighbor embedding algorithm;

[0019] In low-dimensional space, a weighted average linkage hierarchical clustering algorithm was used to identify groups of samples with different tensile strength and bonding strength.

[0020] Conduct statistical analysis of tensile strength and bonding strength of samples within each sample group, identify the strength performance of each sample group, compare the performance differences between different sample groups, and analyze the commonalities of each sample group;

[0021] Based on the analysis results, identify improvement points in interlining design and production processes and make improvement recommendations.

[0022] Optionally, the tensile performance test module uses an adaptive algorithm to dynamically adjust the speed of tension application according to the real-time monitored force and displacement data feedback until the lining cloth sample breaks, and records the force change and displacement data during the stretching process, and converts the force change and displacement data into stress-strain data, including:

[0023] Apply tension to the lining cloth sample at a preset initial speed, and start recording the force and displacement data during the stretching process;

[0024] Continuously monitor the change data of force and displacement in real time, and calculate the real-time stress and strain;

[0025] According to the stress and strain data calculated in real time, set the stress change rate and strain change rate;

[0026] The speed of applying the tension is adjusted according to the stress change rate and the strain change rate using an adaptive algorithm;

[0027] According to the adjusted application speed, the force change and displacement data during the stretching process are recorded in real time, and the force change and displacement data are converted into stress-strain data.

[0028] Optionally, adjusting the tension application speed according to the stress change rate and the strain change rate using an adaptive algorithm includes:

[0029] Real-time analysis of stress change rate and strain change rate;

[0030] If the stress change rate decreases or the strain change rate increases, the tension application speed is slowed down;

[0031] If the rate of change of stress increases or the rate of change of strain decreases, the speed of applying tension is increased;

[0032] Dynamically adjust the speed at which the tension is applied based on the analysis results.

[0033] Optionally, the tensile strength analysis module analyzes the stress-strain data using a density-sensitive anomaly detection algorithm, automatically identifies and excludes abnormal data points, and calculates the maximum tensile strength of the lining cloth sample, including:

[0034] Obtain stress-strain data of lining cloth samples;

[0035] The Euclidean distance between each pair of data points in the stress-strain data is calculated using the Euclidean distance formula, and a distance matrix is ​​constructed using all the calculated Euclidean distances;

[0036] Sort the distance values ​​of each row in the distance matrix in ascending order, and draw the distance ascending curve of each row. By observing the changing trend of the distance ascending curve, find the point where the slope of the distance ascending curve changes, and select the neighborhood distance values ​​that distinguish different density areas based on the point where the slope changes.

[0037] Based on the expected cluster density, a minimum point count threshold is preset, and a density-sensitive anomaly detection algorithm is run using the selected neighborhood distance value and the preset minimum point count threshold to cluster data points and identify core points, edge points, and abnormal data points;

[0038] Remove data points that are considered as outliers, and use core points and edge points to analyze the stress-strain behavior of the lining cloth samples and evaluate the tensile strength characteristics of the lining cloth samples;

[0039] Based on the evaluation results, the point at which the stress value reaches the maximum is located, and the point at which the stress value reaches the maximum is taken as the maximum tensile strength of the lining cloth sample.

[0040] Optionally, the distance values ​​of each row in the distance matrix are sorted in ascending order, and a distance ascending curve of each row is drawn. By observing the changing trend of the distance ascending curve, the points where the slope of the distance ascending curve changes are found. The neighborhood distance values ​​for different density areas are selected based on the points where the slope changes, including:

[0041] For each row in the distance matrix obtained from the stress-strain test, sort the distance values ​​from smallest to largest;

[0042] For each row of the distance matrix, draw a distance ascending curve, the horizontal axis represents the number of neighboring points, and the vertical axis represents the distance value of the neighboring points;

[0043] The point with the largest slope change in the distance ascending curve is located by calculating the rate of change of the slope between adjacent points in the distance matrix;

[0044] The neighborhood distance value is selected based on the change point with the largest slope;

[0045] Among them, the expression for calculating the rate of change of the slope between adjacent points in the distance matrix is:

[0046] ;

[0047] In the formula, Indicates i Point and i +1 rate of change of slope between points;

[0048] Indicates the distance matrix after ascending sorting. i +1 point distance value;

[0049] Indicates the distance matrix after ascending sorting. i The distance value of each point;

[0050] Represents the index difference between neighboring points.

[0051] Optionally, a minimum point count threshold is preset based on the expected cluster density, and a density-sensitive anomaly detection algorithm is run using the selected neighborhood distance value and the preset minimum point count threshold to cluster data points and identify core points, edge points, and abnormal data points, including:

[0052] Create a list containing all data points and obtain the neighborhood distance value and the preset minimum point number threshold;

[0053] Calculate the distance between the current point and all other points in the list, and obtain all data points whose distance is less than or equal to the neighborhood distance value as neighbors;

[0054] For each data point, if the number of neighbors is greater than or equal to the preset minimum point number threshold, it is marked as a core point;

[0055] Add the unlabeled neighbors of the core point to the same cluster and recursively expand the neighbors of the neighbors;

[0056] For data points that fail to form clusters, check whether the data points have core point neighbors; if they have core point neighbors, they are marked as edge points; if they do not have core point neighbors, they are marked as abnormal data points.

[0057] Optionally, mapping the collected tensile strength data and bonding strength data into a low-dimensional space using a t-distributed random neighbor embedding algorithm includes:

[0058] The tensile strength data and the bonding strength data are integrated into a data matrix, wherein the rows in the data matrix represent the samples and the columns represent the tensile strength and the bonding strength;

[0059] Based on the data matrix, set the parameters of the t-distributed random neighbor embedding algorithm and calculate the similarity probability in the high-dimensional space;

[0060] Define a low-dimensional space, and use the t-distributed random neighbor embedding algorithm to define the similarity probability of all sample point pairs in the low-dimensional space;

[0061] Iteratively update the position of sample points in low-dimensional space by gradient descent method;

[0062] The positions of sample points in the low-dimensional space are iteratively updated to minimize the similarity probability difference between the high-dimensional space and the low-dimensional space, and form a final low-dimensional mapping representing the data structure.

[0063] Optionally, the expression for calculating the similarity probability in high-dimensional space is:

[0064] ;

[0065] In the formula, Indicates fixed sample x i Under the conditions, select the sample x j Selected as fixed sample x i Neighborhood probability;

[0066] Indicates fixed sample x i and select samples x j square of the Euclidean distance in tensile strength and adhesive strength;

[0067] x i represents a fixed sample;

[0068] x j Indicates the selection of samples;

[0069] Represents control points i standard deviation of neighborhood size;

[0070] Indicates that in the data set, except x i All sample points except

[0071] k Represents an index variable.

[0072] Optionally, in a low-dimensional space, a weighted average linkage hierarchical clustering algorithm is used to identify sample groups with different tensile strengths and bonding strengths, including:

[0073] Take low-dimensional data as input and perform hierarchical clustering analysis;

[0074] Initialize each sample point as a separate cluster group;

[0075] In weighted average linkage hierarchical clustering, distances between all cluster groups are calculated;

[0076] Merge the two closest cluster groups to form a new cluster group;

[0077] Recalculate the distance between the merged new cluster and all other clusters, and update the distance matrix;

[0078] The step of merging the two closest cluster groups is repeated, merging the two closest clusters in each iteration until all sample points are merged into a single sample group, and sample groups with different tensile strength and bonding strength characteristics are identified.

[0079] Optionally, statistical analysis of tensile strength and adhesive strength is performed on the samples in each sample group to identify the strength performance of each sample group and compare the performance differences between different sample groups. At the same time, the commonalities of each sample group are analyzed, including:

[0080] Collect sample data including tensile strength and adhesive strength within each sample group;

[0081] Performing statistical analysis on sample data within each sample group;

[0082] Statistical test methods were used to analyze performance differences between groups;

[0083] Based on the statistical analysis results, identify the best and worst performing sample groups;

[0084] The statistical results of different groups were compared to identify the common factors affecting tensile strength and adhesive strength.

[0085] Compared with the prior art, this application has the following beneficial effects:

[0086] 1. The present invention ensures the real-time and accuracy of data by real-time monitoring of force and displacement data. The adaptive algorithm dynamically adjusts the speed of applying tension according to the real-time monitored stress and strain data; when it is detected that the rate of change of stress and strain decreases, the algorithm will slow down the rate of increase of tension to adapt to the deformation behavior of the material; conversely, when the rate of change of stress and strain increases, the algorithm will speed up the rate of increase of tension to keep up with the rapid deformation of the material, and convert force and displacement data into stress and strain data in real time, providing direct mechanical performance indicators for analysis. The adaptive algorithm can avoid material overload or excessive deformation, thereby protecting the sample from damage, which is particularly important for expensive or difficult-to-manufacture samples; the adaptive algorithm can adjust the test process according to the actual behavior of the material, thereby reducing the time required for testing. Since the algorithm can automatically adjust the test parameters, the operator can focus on other tasks, thereby improving the efficiency of the entire test process.

[0087] 2. The present invention ensures the data quality during tensile strength analysis by automatically identifying and excluding abnormal data points. The method of Euclidean distance calculation and distance matrix can analyze the relationship between data points in detail. By identifying the mutation points of data density, the boundary between normal and abnormal data can be more accurately defined, thereby improving the accuracy of analysis. By drawing an ascending distance curve for each data point, the method provides an intuitive way to observe and analyze the density changes between data points. Visualization helps technicians more easily identify anomalies and trends in data sets. The neighborhood distances of different density areas are determined based on the change points with the largest slopes, so that the algorithm can flexibly respond to data sets of different types and sizes. By accurately identifying and excluding abnormal data and through precise tensile strength analysis, it can help the production department identify possible production problems.

[0088] 3. The present invention effectively maps multidimensional data (such as tensile strength and bonding strength) to low-dimensional space through the t-SNE algorithm, which simplifies the data analysis process while maintaining the main relationships and structures between the data; the data representation in low-dimensional space makes complex high-dimensional data relationships easier to visualize and understand, and uses a hierarchical clustering algorithm to identify different performance groups based on low-dimensional mapping, which can more accurately reflect the similarities and differences between samples. By statistically analyzing the data within each sample group, the tensile strength and bonding strength performance of different samples can be compared and evaluated in detail. By analyzing the performance differences of different groups, the key factors affecting the tensile strength and bonding strength can be identified, providing a scientific basis for material selection and design optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] The above characteristics, features and advantages of the present invention and their implementation methods and methods will become more clearly understood in conjunction with the following description of the embodiments, which are described in detail in conjunction with the accompanying drawings. Herein, a schematic diagram is shown:

[0090] Figure 1 The present invention is a block diagram of a strength testing system for lining cloth production according to an embodiment of the present invention.

[0091] In the figure:

[0092] 1. Interlining sample cutting module; 2. Tensile performance testing module; 3. Tensile strength analysis module; 4. Adhesion strength testing module; 5. Environmental adaptability testing module; 6. Data analysis and evaluation module. DETAILED DESCRIPTION

[0093] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0094] According to an embodiment of the present invention, a strength testing system for lining cloth production is provided.

[0095] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to the strength testing system for lining cloth production according to the embodiment of the present invention, the strength testing system comprises: a lining cloth sample cutting module 1, a tensile performance testing module 2, a tensile strength analysis module 3, an adhesive strength testing module 4, an environmental adaptability testing module 5 and a data analysis and evaluation module 6;

[0096] The lining cloth sample cutting module 1, the tensile performance testing module 2, the tensile strength analysis module 3, the bonding strength testing module 4, the environmental adaptability testing module 5 and the data analysis and evaluation module 6 are connected in sequence;

[0097] The lining cloth sample cutting module 1 is used for cutting lining cloth samples according to preset standard sizes and shapes.

[0098] It should be explained that computer-aided design (e.g., CAD software) software is used to design the cutting template for the lining fabric sample; during the design process, the preset standard size and shape are input according to the product design requirements. After the design is completed, the cutting template data is imported into the control terminal of the cutting machine. The computer-aided design allows precise control of the cutting lines and angles to ensure that the size and shape of each sample meet the design specifications; the uncut lining fabric is laid flat on the workbench of the cutting machine. Ensure that the fabric is flat and wrinkle-free; use a fixing device (such as a clamp or vacuum suction cup) to fix the fabric to prevent it from moving during the cutting process. The flatness of the fabric will directly affect the cutting accuracy and the quality of the sample; according to the loaded cutting template, the cutting machine (which may be a laser cutter, blade cutter or other type) starts to automatically cut the fabric, and the cutter head or laser head of the cutting machine moves along the predetermined path to accurately cut the fabric; after cutting, the staff or the automated system will check whether the size and shape of each sample meet the preset standards, and perform necessary cleaning and grinding on the cutting edges to ensure that there are no burrs or overburning; collect the samples that have passed the inspection and post-processing, classify or number them as needed, and use appropriate packaging materials and methods to package the samples to protect the samples from damage during storage and transportation.

[0099] The tensile performance test module 2 is used to use an adaptive algorithm to dynamically adjust the speed of tension application according to the real-time monitored force and displacement data feedback until the lining cloth sample breaks, and at the same time record the force change and displacement data during the stretching process, and convert the force change and displacement data into stress-strain data.

[0100] Preferably, the tensile performance testing module 2 uses an adaptive algorithm to dynamically adjust the speed of applying the tensile force according to the force and displacement data feedback monitored in real time until the lining cloth sample breaks, and records the force change and displacement data during the stretching process, and converts the force change and displacement data into stress-strain data, including:

[0101] Apply tension to the lining cloth sample at a preset initial speed, and start recording the force and displacement data during the stretching process;

[0102] Continuously monitor the change data of force and displacement in real time, and calculate the real-time stress and strain;

[0103] According to the stress and strain data calculated in real time, set the stress change rate and strain change rate;

[0104] The speed of applying the tension is adjusted according to the stress change rate and the strain change rate using an adaptive algorithm;

[0105] According to the adjusted application speed, the force change and displacement data during the stretching process are recorded in real time, and the force change and displacement data are converted into stress-strain data.

[0106] Preferably, using an adaptive algorithm to adjust the tension application speed according to the stress change rate and the strain change rate includes:

[0107] Real-time analysis of stress change rate and strain change rate;

[0108] If the stress change rate decreases or the strain change rate increases, the tension application speed is slowed down;

[0109] If the rate of change of stress increases or the rate of change of strain decreases, the speed of applying tension is increased;

[0110] Dynamically adjust the speed at which the tension is applied based on the analysis results.

[0111] It should be explained that an electric or hydraulic stretching machine is used to start applying tension to the lining cloth sample at a preset initial speed; at the same time, the force and displacement data are recorded through the sensor connected to the stretching machine. The initial speed should be set according to the material type and expected test conditions to ensure the initial stability of the data. The force and displacement sensors must be highly precise and able to provide real-time data to facilitate accurate calculation and adjustment of the algorithm.

[0112] Continuously monitor changes in tension and displacement, and calculate stress and strain in real time through the software. Stress is calculated as the applied force divided by the initial cross-sectional area of ​​the sample, and strain is calculated as the displacement divided by the initial length of the sample. According to the stress and strain data calculated in real time, the software sets the thresholds for the stress change rate and the strain change rate. The stress change rate and the strain change rate are important indicators of material behavior, reflecting the response speed and deformation capacity of the material during the stress process. Setting a reasonable change rate threshold is crucial to optimizing the test process and protecting the sample from damage.

[0113] Real-time analysis of stress change rate and strain change rate data; if the stress change rate decreases or the strain change rate increases, the system adapts to the deformation behavior of the material by slowing down the speed of tension application; if the stress change rate increases or the strain change rate decreases, the system responds to the rapid deformation needs of the material by speeding up the speed of tension application. The adaptive adjustment mechanism is achieved through advanced control algorithms, which require real-time feedback and high-speed processing capabilities. The control algorithm can ensure that the sample is subjected to uniform and appropriate force throughout the test process, thereby obtaining reliable test results, and continuing to record the adjusted tension and displacement data in real time, and converting the recorded force and displacement data into stress and strain data for subsequent analysis.

[0114] The tensile strength analysis module 3 is used to analyze the stress-strain data using a density-sensitive anomaly detection algorithm, automatically identify and exclude abnormal data points, and calculate the maximum tensile strength of the lining cloth sample.

[0115] Preferably, the tensile strength analysis module 3 analyzes the stress-strain data using a density-sensitive anomaly detection algorithm, automatically identifies and excludes abnormal data points, and calculates the maximum tensile strength of the lining cloth sample, including:

[0116] Obtain stress-strain data of lining cloth samples.

[0117] The Euclidean distance formula is used to calculate the Euclidean distance between each pair of data points in the stress-strain data, and all calculated Euclidean distances are used to form a distance matrix.

[0118] Preferably, the distance values ​​of each row in the distance matrix are sorted in ascending order, and a distance ascending curve of each row is drawn. By observing the changing trend of the distance ascending curve, the point where the slope of the distance ascending curve changes is found. The neighborhood distance values ​​for different density areas are selected based on the point where the slope changes, including:

[0119] For each row in the distance matrix obtained from the stress-strain test, sort the distance values ​​from smallest to largest;

[0120] For each row of the distance matrix, draw a distance ascending curve, the horizontal axis represents the number of neighboring points, and the vertical axis represents the distance value of the neighboring points;

[0121] The point with the largest slope change in the distance ascending curve is located by calculating the rate of change of the slope between adjacent points in the distance matrix;

[0122] The neighborhood distance value is selected based on the change point with the largest slope;

[0123] Among them, the expression for calculating the rate of change of the slope between adjacent points in the distance matrix is:

[0124] ;

[0125] In the formula, Indicates i Point and i +1 rate of change of slope between points;

[0126] Indicates the distance matrix after ascending sorting. i +1 point distance value;

[0127] Indicates the distance matrix after ascending sorting. i The distance value of each point;

[0128] Represents the index difference between neighboring points.

[0129] It should be explained that the stress-strain data collected from the test equipment (such as the tensile performance test module) are usually stored in digital form, including the stress value and the corresponding strain value generated by each test. The stress-strain data is a direct representation of the mechanical properties of the material. Stress is the applied force divided by the cross-sectional area of ​​the material, and strain is the ratio of the material deformation to the original length. The Euclidean distance formula is used to calculate the distance between each pair of data points in the stress-strain data set. The Euclidean distance provides a method to measure the absolute distance between two points in a multidimensional space. The dimensions here are stress and strain. The distance matrix is ​​a symmetric matrix. Each element in the matrix represents the distance between two data points. Each row in the distance matrix is ​​sorted in ascending order, and for each row of distance , draw a distance ascending curve, the horizontal axis represents the number of adjacent points, and the vertical axis represents the distance value of the adjacent points. The sorted distance matrix can reveal the density changes between data points. Smaller distances represent higher data density. The ascending curve is an intuitive way to understand the relative position and density changes between data points. Calculate the rate of change of the slope between adjacent points in the distance ascending curve, locate the point with the most significant slope change, and determine the threshold of the outlier point based on the point with the largest slope change, that is, select this point as the boundary of different density areas. The rate of change of the slope can show the mutation of data density. These mutation points may indicate the existence of abnormal data. By identifying these change points, normal data and abnormal data can be more accurately divided to ensure the accuracy and representativeness of the data when calculating the tensile strength.

[0130] Sort the distance values ​​of each row in the distance matrix in ascending order, and draw the distance ascending curve of each row. By observing the changing trend of the distance ascending curve, find the points where the slope of the distance ascending curve changes, and select the neighborhood distance values ​​that distinguish different density areas based on the points where the slope changes.

[0131] A minimum point count threshold is preset based on the expected cluster density, and a density-sensitive anomaly detection algorithm is run using the selected neighborhood distance value and the preset minimum point count threshold to cluster data points and identify core points, edge points, and abnormal data points.

[0132] Preferably, a minimum point count threshold is preset based on the expected cluster density, and a density-sensitive anomaly detection algorithm is run using the selected neighborhood distance value and the preset minimum point count threshold to cluster data points and identify core points, edge points, and abnormal data points, including:

[0133] Create a list containing all data points and obtain the neighborhood distance value and the preset minimum point number threshold;

[0134] Calculate the distance between the current point and all other points in the list, and obtain all data points whose distance is less than or equal to the neighborhood distance value as neighbors;

[0135] For each data point, if the number of neighbors is greater than or equal to the preset minimum point number threshold, it is marked as a core point;

[0136] Add the unlabeled neighbors of the core point to the same cluster and recursively expand the neighbors of the neighbors;

[0137] For data points that fail to form clusters, check whether the data points have core point neighbors; if they have core point neighbors, they are marked as edge points; if they do not have core point neighbors, they are marked as abnormal data points.

[0138] To explain, create a list of all the data points to be analyzed. These data points come from tensile tests, including stress and strain measurements; determine the neighborhood distance value and minimum point count threshold for each data point. The neighborhood distance value is based on the slope change points identified in the previous step, while the minimum point count threshold is preset based on the expected cluster density; the neighborhood distance value is used to define the size of the local neighborhood area around a data point, while the minimum point count threshold is used to determine whether a data point is "dense" enough to be considered a core point.

[0139] For each data point in the list, calculate the distance between it and all other points. This can be achieved using the standard Euclidean distance formula. For each data point, determine all neighbors within its neighborhood distance value range. If the number of neighbors of a data point is greater than or equal to the preset minimum point number threshold, it is marked as a core point. Core points are those points that have enough neighbors in their neighborhood, indicating that there is a high density of data points clustered around these points.

[0140] Starting from the marked core point, add all its neighbors (unmarked) to the same cluster, recursively expand the neighbors of each neighbor until there are no new core points to add. For data points that fail to form a cluster, check whether they have neighbors of core points. If so, mark them as edge points; if not, mark them as abnormal data points. Edge points are usually located at the edge of the cluster, while abnormal points may be noise or data points that do not conform to any clustering pattern; the main difference between edge points and core points is that the number of neighbors around them is not enough to become core points themselves; abnormal points may deviate significantly from the main data clustering area due to various reasons (such as measurement errors or data entry errors).

[0141] Data points considered as outliers were removed, and the stress-strain behavior of the lining cloth samples was analyzed using core points and edge points to evaluate the tensile strength characteristics of the lining cloth samples.

[0142] Based on the evaluation results, the point at which the stress value reaches the maximum is located, and the point at which the stress value reaches the maximum is taken as the maximum tensile strength of the lining cloth sample.

[0143] It should be explained that all data points marked as abnormal are removed from the data set; these points may be caused by measurement errors, data entry errors or equipment failures, and their characteristics may interfere with the actual analysis results. The retained data set includes only core points and edge points, which are considered to be valid data reflecting the real physical behavior. The data set after removing the abnormal points is subjected to stress-strain curve analysis. Use statistical software or dedicated analysis tools to draw stress-strain curves. The behavior of the material during the stress process can be observed through the stress-strain curve, including the elastic stage, yield point and final fracture point. The stress-strain curve is a basic tool used in materials science to describe the mechanical properties of materials. Stress (σ) is usually used as the vertical axis and strain (ε) as the horizontal axis. The analysis of core points and edge points can reveal the uniform behavior of the sample under different loading conditions and any possible variations.

[0144] Find the point where the stress value reaches the maximum in the stress-strain curve. This point represents the maximum force that the sample can withstand before breaking, which usually corresponds to the maximum tensile strength of the material. Use data analysis tools, such as regression analysis or peak detection algorithms, to accurately identify this point. The maximum tensile strength refers to the maximum stress that the material can withstand before breaking. This is a key indicator for evaluating material performance. Record the analysis results and describe the tensile strength characteristics of the lining cloth sample in detail in the report. The report includes the stress-strain curve, the maximum tensile strength value, and other related mechanical performance parameters. The record and report of the results not only provide a detailed description of the sample performance, but can also be used as a basis for future improvements in material design and production processes.

[0145] The bonding strength test module 4 is used to measure the stress distribution of the lining cloth sample during the tearing process through a peeling test method, and to evaluate the bonding strength between the lining cloth sample and the fabric.

[0146] It should be explained that standard-sized samples should be cut from the interlining and fabric to ensure that the sample edges are neat and flawless. The interlining samples should be bonded to the fabric and cured under specified conditions to ensure the quality of the bond. The accuracy of sample preparation has a direct impact on the test results. Standard size helps to achieve comparability of test results. Surface treatment and curing environment (such as temperature, pressure, time) before bonding will directly affect the bond strength. A dedicated peel tester should be used to set the sample on the machine. This usually includes fixing the fabric at one end and connecting the interlining end to the dynamic tension device of the tester. The peel angle and stretching speed should be determined and set. These parameters should be based on the test standard. The peel angle and stretching speed are key factors affecting the bond strength measurement. Factors usually need to be optimized according to specific materials and bonding systems. The peel tester can provide precise force control and data recording to accurately measure the force applied during the peeling process. Start the peel tester and slowly and evenly pull the lining cloth to gradually peel off the bonded part with the fabric. Real-time recording of force and displacement data during the peeling process is completed through the connected sensors. Use dedicated software to analyze the collected force and displacement data, calculate the peel strength, maximum force value and related stress distribution, and evaluate the bonding strength between the lining cloth sample and the fabric based on the data analysis results to determine whether the material meets industrial standards or design requirements. Peel strength is usually expressed as the force required per unit width to peel the bonding interface, which is a key parameter for measuring bonding performance.

[0147] The environmental adaptability test module 5 is used to simulate and test the bonding strength of the lining cloth samples under various environmental conditions and evaluate the performance of the lining cloth under various environments.

[0148] It should be explained that the equipment in the environmental test room, such as the temperature and humidity control box, is used to set up a specific test environment; the test environment may include high temperature, low temperature, high humidity, low humidity, ultraviolet radiation, etc., to ensure that the environmental settings meet the predetermined test standards and parameters to simulate the conditions that the lining may encounter in actual use. Environmental factors such as temperature and humidity have a significant impact on the bonding performance of the material. Controlling these parameters can help evaluate the behavior and durability of the material under different conditions.

[0149] Cut and prepare interlining samples according to standard methods, ensure that all samples have the same size and shape to ensure the consistency of the test, install the samples on a special test device, such as an environmental adaptability test frame, and the consistency of the samples is the key to ensure data comparability. Different cutting or installation methods may lead to deviations in test results; start the test under set environmental conditions, use mechanical testing equipment such as a tension tester to measure the bond strength of the sample, and record the force and displacement data during the test in real time, as well as changes in environmental conditions (such as temperature and humidity). The force and displacement data will be used to calculate the bond strength, while the environmental data provides an exact record of the test conditions, which is crucial for analyzing material properties. Use data analysis software to process the collected test data, analyze the bond strength and performance of interlining samples under different environmental conditions, evaluate the data, and determine whether the material's performance is as expected, and under which conditions it performs best or has potential problems. Changes in bond strength and other mechanical properties can indicate the material's ability to adapt to a specific environment.

[0150] The data analysis and evaluation module 6 is used to collect and integrate the data of the maximum tensile strength of the lining cloth sample and the bonding strength between the lining cloth sample and the fabric, and evaluate the strength performance of the lining cloth sample based on the integrated data.

[0151] The data analysis and evaluation module 6 collects and integrates the data of the maximum tensile strength of the lining cloth sample and the bonding strength between the lining cloth sample and the fabric, and evaluates the strength performance of the lining cloth sample based on the integrated data, including:

[0152] Obtain tensile strength data and bonding strength data of lining cloth samples.

[0153] The collected tensile strength data and bonding strength data were mapped into a low-dimensional space using the t-distributed random neighbor embedding algorithm.

[0154] Preferably, mapping the collected tensile strength data and bonding strength data into a low-dimensional space using a t-distributed random neighbor embedding algorithm comprises:

[0155] The tensile strength data and the bonding strength data are integrated into a data matrix, wherein the rows in the data matrix represent the samples and the columns represent the tensile strength and the bonding strength;

[0156] Based on the data matrix, set the parameters of the t-distributed random neighbor embedding algorithm and calculate the similarity probability in the high-dimensional space;

[0157] Define a low-dimensional space, and use the t-distributed random neighbor embedding algorithm to define the similarity probability of all sample point pairs in the low-dimensional space;

[0158] Iteratively update the position of sample points in low-dimensional space by gradient descent method;

[0159] The positions of sample points in the low-dimensional space are iteratively updated to minimize the similarity probability difference between the high-dimensional space and the low-dimensional space, and form a final low-dimensional mapping representing the data structure.

[0160] Preferably, the expression for calculating the similarity probability in high-dimensional space is:

[0161] ;

[0162] In the formula, Indicates fixed sample x i Under the conditions, select the sample x j Selected as fixed sample x i Neighborhood probability;

[0163] Indicates fixed sample x i and select samples x j square of the Euclidean distance in tensile strength and adhesive strength;

[0164] x i represents a fixed sample;

[0165] x j Indicates the selection of samples;

[0166] Represents control points i standard deviation of neighborhood size;

[0167] Indicates that in the data set, except x i All sample points except

[0168] k Represents an index variable.

[0169] It should be explained that the tensile strength and bonding strength data of the lining samples were collected from the previous test module and integrated into a data matrix, in which the rows represent different samples and the columns represent the tensile strength and bonding strength, respectively. The creation of the data matrix is ​​the first step in efficient data analysis, ensuring the structure and operability of the data. Each column of data should be properly normalized or standardized to eliminate the dimension effect and the difference in the numerical range. Parameters are set for the t-SNE algorithm, including perplexity, learning rate, etc., to calculate the similarity probability of sample point pairs in high-dimensional data, which is usually achieved through the Gaussian distribution function. Perplexity The parameters determine the size of the local neighborhood and affect the interaction between high-dimensional data points. The positions of sample points are randomly initialized in low-dimensional space. The similarity probability of all pairs of sample points is defined in low-dimensional space using t distribution. The t distribution helps to better handle the interaction of distant points in low-dimensional space. The gradient descent method is used to iteratively update the positions of sample points in low-dimensional space. During the iteration, the Kullback-Leibler (KL) divergence of the similarity probability between high-dimensional space and low-dimensional space is gradually reduced, that is, the point position in low-dimensional space is optimized to better reflect the structure of high-dimensional space. Gradient descent is an optimization algorithm used to find the minimum value of the loss function. In t-SNE, the loss function is the divergence between high-dimensional and low-dimensional similarity probabilities. After completing enough iterations, the final low-dimensional mapping is obtained. This mapping represents the structure of the data and the relationship between the sample points. This mapping is used for further analysis, such as cluster analysis or anomaly detection, to evaluate and understand the performance of the lining cloth sample. The results of the low-dimensional mapping can be displayed visually, making the relationship between high-dimensional data more intuitive and easy to understand, identifying the similarities and differences between different samples, and providing a basis for material optimization and improvement.

[0170] In low-dimensional space, a weighted average linkage hierarchical clustering algorithm was used to identify groups of samples with different tensile strength and bonding strength.

[0171] Preferably, in a low-dimensional space, using a weighted average link hierarchical clustering algorithm to identify sample groups with different tensile strengths and bonding strengths includes:

[0172] Take low-dimensional data as input and perform hierarchical clustering analysis;

[0173] Initialize each sample point as a separate cluster group;

[0174] In weighted average linkage hierarchical clustering, distances between all cluster groups are calculated;

[0175] Merge the two closest cluster groups to form a new cluster group;

[0176] Recalculate the distance between the merged new cluster and all other clusters, and update the distance matrix;

[0177] The step of merging the two closest cluster groups is repeated, merging the two closest clusters in each iteration until all sample points are merged into a single sample group, and sample groups with different tensile strength and bonding strength characteristics are identified.

[0178] It should be explained that the data are obtained from the low-dimensional map generated by the t-SNE algorithm. These data now contain low-dimensional representations of the tensile strength and bonding strength of the samples. These low-dimensional data are used as input to start the hierarchical clustering analysis process. Hierarchical clustering is a clustering algorithm that does not require the pre-specified number of clusters. It builds a clustering structure by gradually merging data points or data groups. Initially, each sample point is regarded as an independent cluster group. This means that if there are N sample points, then there are N cluster groups at the beginning of clustering. In weighted average linkage hierarchical clustering, the distance between groups is obtained by calculating the average distance of all pairs of sample points in the two groups. The distance is usually calculated using Euclidean distance, and the weighted average method ensures that the distance calculation takes into account the influence of group size. In each iteration, the two groups with the closest distance are selected to merge to form a new cluster group. This merging process reduces the total number of clusters. After each merge, the distance matrix needs to be updated and the distance between the newly formed cluster and all other clusters needs to be recalculated. The iterative merging process continues, merging the two closest groups each time until all sample points are merged into a single sample group. This process can be visualized by generating a dendrogram (cluster tree or tree structure diagram) to show how the sample points are gradually merged. The final dendrogram provides a clear view of the similarity relationship and clustering hierarchy between samples.

[0179] The tensile strength and bonding strength of the samples in each sample group were statistically analyzed to identify the strength performance of each sample group, compare the performance differences between different sample groups, and analyze the commonalities of each sample group.

[0180] Preferably, statistical analysis of tensile strength and adhesive strength is performed on samples in each sample group to identify the strength performance of each sample group and compare the performance differences between different sample groups. At the same time, the commonalities of each sample group are analyzed, including:

[0181] Collect sample data including tensile strength and adhesive strength within each sample group;

[0182] Performing statistical analysis on sample data within each sample group;

[0183] Statistical test methods were used to analyze performance differences between groups;

[0184] Based on the statistical analysis results, identify the best and worst performing sample groups;

[0185] The statistical results of different groups were compared to identify the common factors affecting tensile strength and adhesive strength.

[0186] Based on the analysis results, identify improvement points in interlining design and production processes and make improvement recommendations.

[0187] It should be explained that, according to the previous hierarchical clustering results, the samples were classified into groups, and the tensile strength and bonding strength data of each group were collected to ensure data integrity and perform appropriate preprocessing, such as missing value processing and outlier inspection; basic statistics of the sample data in each group, such as mean, median, standard deviation, maximum and minimum values, were calculated, and descriptive statistical charts, such as box plots and histograms, were drawn to visualize the data distribution and variability within the group; statistical methods such as ANOVA (analysis of variance), t-test or non-parametric test were applied to test the significant differences in tensile strength and bonding strength between different groups; considering the sample size and data distribution characteristics, appropriate statistical test methods were selected, and the statistical test results were analyzed to determine which groups Showing higher or lower performance, identifying specific performance trends, such as groups with consistent good or poor performance, and identifying groups with extreme performance help understand the extreme conditions of material performance and potential reliability issues; analyzing the commonalities and differences between different groups, trying to find common factors that affect performance, and conducting multivariate analysis, such as regression analysis or principal component analysis, to explore the relationship and influence of multiple variables. Through comparative analysis, key factors that may affect material performance, such as raw material quality, production process, etc., can be found. Based on the above analysis results, specific improvement measures for the design and production process are proposed, and detailed reports are prepared to summarize the analysis findings and recommended measures for reference by the R&D and production teams. The application of the results can directly affect product quality and market competitiveness. Through systematic analysis and feedback, product performance and production efficiency can be continuously optimized.

[0188] The actual application examples are as follows:

[0189] The automotive interior lining fabrics need to meet the requirements of high strength, good bonding performance and environmental adaptability to ensure safety and comfort. Using the test system of the embodiment of the present invention, the performance of the lining fabric samples can be comprehensively evaluated to ensure that they meet the strict standards of the automotive industry.

[0190] Step 1: Use CAD software to design a lining fabric sample that conforms to the shape of the automotive interior parts.

[0191] The interlining is cut to preset standard sizes and shapes on an automatic cutting machine, ensuring precision and consistency for each sample.

[0192] Step 2: Perform a tensile test on the cut lining cloth samples to evaluate their stress-strain response at different tension application speeds.

[0193] An adaptive algorithm adjusts the pulling speed based on real-time data, ensuring the data accurately reflects material properties.

[0194] Step 3: Use density-sensitive anomaly detection algorithm to process stress-strain data and automatically remove abnormal data points.

[0195] Calculate and record the maximum tensile strength of the lining fabric sample.

[0196] Step 4: Perform a peel test to measure the bond strength between the lining and the car interior fabric.

[0197] Analyze stress distribution during bonding and evaluate bonding quality.

[0198] Step 5: Test the performance of the lining fabric sample under simulated high temperature, high humidity and other environmental conditions.

[0199] Evaluate the bond strength and overall performance of the interlining in these environments to ensure it can withstand the various environments of the automotive interior.

[0200] Step 6: Integrate the test data and use the t-SNE algorithm to map it to a low-dimensional space for in-depth analysis.

[0201] The hierarchical clustering algorithm is used to identify sample groups with different tensile strength and bonding strength, providing data support for the improvement of lining cloth.

[0202] Through testing and analysis, we can understand in detail the performance of lining fabric samples under actual use conditions; the test results help to formulate better material formulas and production processes, improve product quality and performance; achieve lining fabric design optimization for specific application requirements, and ensure the safety and durability of automotive interiors.

[0203] In summary, with the help of the above technical solutions of the present invention, the present invention ensures the real-time and accuracy of the data by real-time monitoring of force and displacement data, and the adaptive algorithm dynamically adjusts the speed of applying the tension according to the real-time monitored stress and strain data; when it is detected that the rate of change of stress and strain decreases, the algorithm slows down the rate of increase of the tension to adapt to the deformation behavior of the material; conversely, when the rate of change of stress and strain increases, the algorithm speeds up the rate of increase of the tension to keep up with the rapid deformation of the material, and converts the force and displacement data into stress and strain data in real time, providing direct mechanical performance indicators for analysis, and the adaptive algorithm can avoid material overload or excessive deformation, thereby protecting the sample from damage, which is particularly important for expensive or difficult-to-manufacture samples; the adaptive algorithm can adjust the test process according to the actual behavior of the material, thereby reducing the time required for the test, and since the algorithm can automatically adjust the test parameters, the operator can focus on other tasks, thereby improving the efficiency of the entire test process; the present invention ensures the data quality when performing tensile strength analysis by automatically identifying and excluding abnormal data points, and uses the Euclidean distance calculation and distance matrix method to analyze the relationship between data points in detail, and more accurately define normal and abnormal by identifying the mutation points of data density. The boundary of the data is determined to improve the accuracy of the analysis; by drawing the ascending distance curve of each data point, the method provides an intuitive way to observe and analyze the density changes between data points. Visualization helps technicians to more easily identify anomalies and trends in the data set. The neighborhood distance of different density areas is determined based on the change point with the largest slope, so that the algorithm can flexibly cope with data sets of different types and sizes. By accurately identifying and excluding abnormal data, and through accurate tensile strength analysis, it can help the production department identify possible production problems; the present invention effectively maps multidimensional data (such as tensile strength and bonding strength) to low-dimensional space through the t-SNE algorithm, simplifies the data analysis process, and maintains the main relationship and structure between the data; the data representation in low-dimensional space makes complex high-dimensional data relationships easier to be visualized and understood, and the use of a hierarchical clustering algorithm to identify different performance groups based on low-dimensional mapping can more accurately reflect the similarities and differences between samples. By statistically analyzing the data in each sample group, the tensile strength and bonding strength performance of different samples can be compared and evaluated in detail. By analyzing the performance differences of different groups, the key factors affecting the tensile strength and bonding strength can be identified, providing a scientific basis for material selection and design optimization.

[0204] Although the present invention has been disclosed as above with preferred embodiments, the embodiments are merely examples for the convenience of description and are not intended to limit the present invention. Those skilled in the art may make several changes and modifications without departing from the spirit and scope of the present invention. The scope of protection claimed by the present invention shall be based on the claims.

Claims

1. A strength testing system for lining cloth production, It is characterized in that The strength testing system includes: lining cloth sample cutting module, tensile performance testing module, tensile strength analysis module, bonding strength testing module, environmental adaptability testing module and data analysis and evaluation module; The lining cloth sample cutting module, the tensile performance testing module, the tensile strength analysis module, the bonding strength testing module, the environmental adaptability testing module and the data analysis and evaluation module are connected in sequence; The lining cloth sample cutting module is used to cut the lining cloth sample according to the preset standard size and shape; The tensile performance testing module is used to dynamically adjust the speed of tensile force application according to the force and displacement data feedback monitored in real time by using an adaptive algorithm until the lining cloth sample breaks, and simultaneously record the force change and displacement data during the stretching process, and convert the force change and displacement data into stress-strain data; The tensile strength analysis module is used to analyze stress-strain data using a density-sensitive anomaly detection algorithm, automatically identify and exclude abnormal data points, and calculate the maximum tensile strength of the lining cloth sample; The bonding strength test module is used to measure the stress distribution of the lining cloth sample during the tearing process by a peeling test method, and to evaluate the bonding strength between the lining cloth sample and the fabric; The environmental adaptability test module is used to simulate and test the bonding strength of the lining cloth sample under various environmental conditions, and evaluate the performance of the lining cloth under various environments; The data analysis and evaluation module is used to collect and integrate the data of the maximum tensile strength of the lining cloth sample and the bonding strength between the lining cloth sample and the fabric, and evaluate the strength performance of the lining cloth sample based on the integrated data; The data analysis and evaluation module collects and integrates the data of the maximum tensile strength of the lining cloth sample and the bonding strength between the lining cloth sample and the fabric, and evaluates the strength performance of the lining cloth sample based on the integrated data, including: Obtain the tensile strength data and bonding strength data of the lining cloth samples; The collected tensile strength data and bonding strength data were mapped into a low-dimensional space using the t-distributed random neighbor embedding algorithm; In low-dimensional space, a weighted average linkage hierarchical clustering algorithm was used to identify groups of samples with different tensile strength and bonding strength. Conduct statistical analysis of tensile strength and bonding strength of samples within each sample group, identify the strength performance of each sample group, compare the performance differences between different sample groups, and analyze the commonalities of each sample group; Based on the analysis results, identify improvement points in interlining design and production process and make improvement suggestions; The tensile strength analysis module analyzes stress-strain data using a density-sensitive anomaly detection algorithm, automatically identifies and excludes abnormal data points, and calculates the maximum tensile strength of the lining cloth sample, including: Obtain stress-strain data of lining cloth samples; The Euclidean distance between each pair of data points in the stress-strain data is calculated using the Euclidean distance formula, and a distance matrix is ​​constructed using all the calculated Euclidean distances; Sort the distance values ​​of each row in the distance matrix in ascending order, and draw the distance ascending curve of each row. By observing the changing trend of the distance ascending curve, find the point where the slope of the distance ascending curve changes, and select the neighborhood distance values ​​that distinguish different density areas based on the point where the slope changes. Based on the expected cluster density, a minimum point count threshold is preset, and a density-sensitive anomaly detection algorithm is run using the selected neighborhood distance value and the preset minimum point count threshold to cluster data points and identify core points, edge points, and abnormal data points; Remove data points that are considered as outliers, and use core points and edge points to analyze the stress-strain behavior of the lining cloth samples and evaluate the tensile strength characteristics of the lining cloth samples; Based on the evaluation results, the point at which the stress value reaches the maximum is located, and the point at which the stress value reaches the maximum is taken as the maximum tensile strength of the lining cloth sample; The distance values ​​of each row in the distance matrix are sorted in ascending order, and the distance ascending curve of each row is drawn. By observing the changing trend of the distance ascending curve, the point where the slope of the distance ascending curve changes is found, and the neighborhood distance values ​​for different density areas are selected based on the point where the slope changes, including: For each row in the distance matrix obtained from the stress-strain test, sort the distance values ​​from smallest to largest; For each row of the distance matrix, draw a distance ascending curve, the horizontal axis represents the number of neighboring points, and the vertical axis represents the distance value of the neighboring points; The point with the largest slope change in the distance ascending curve is located by calculating the rate of change of the slope between adjacent points in the distance matrix; The neighborhood distance value is selected based on the change point with the largest slope; The expression for calculating the rate of change of the slope between adjacent points in the distance matrix is: ； In the formula, Indicates i Point and i +1 rate of change of slope between points; Indicates the distance matrix after ascending sorting. i +1 point distance value; Indicates the distance matrix after ascending sorting. i The distance value of each point; Represents the index difference between neighboring points; The method of mapping the collected tensile strength data and bonding strength data to a low-dimensional space using a t-distributed random neighbor embedding algorithm includes: The tensile strength data and the bonding strength data are integrated into a data matrix, wherein the rows in the data matrix represent the samples and the columns represent the tensile strength and the bonding strength; Based on the data matrix, set the parameters of the t-distributed random neighbor embedding algorithm and calculate the similarity probability in the high-dimensional space; Define a low-dimensional space, and use the t-distributed random neighbor embedding algorithm to define the similarity probability of all sample point pairs in the low-dimensional space; Iteratively update the position of sample points in low-dimensional space by gradient descent method; Iteratively update the sample point positions in the low-dimensional space to minimize the similarity probability difference between the high-dimensional space and the low-dimensional space, and form a final low-dimensional mapping representing the data structure; The expression for calculating the similarity probability of the high-dimensional space is: ； In the formula, Indicates fixed sample x i Under the conditions, select the sample x j Selected as fixed sample x i Neighborhood probability; Indicates fixed sample x i and select samples x j square of the Euclidean distance in tensile strength and adhesive strength; x i represents a fixed sample; x j Indicates the selection of samples; Represents control points i standard deviation of neighborhood size; Indicates that in the data set, except x i All sample points except k Represents an index variable; The method of identifying sample groups with different tensile strength and bonding strength using a weighted average link hierarchical clustering algorithm in a low-dimensional space includes: Take low-dimensional data as input and perform hierarchical clustering analysis; Initialize each sample point as a separate cluster group; In weighted average linkage hierarchical clustering, distances between all cluster groups are calculated; Merge the two closest cluster groups to form a new cluster group; Recalculate the distance between the merged new cluster and all other clusters, and update the distance matrix; The step of merging the two closest cluster groups is repeated, merging the two closest clusters in each iteration until all sample points are merged into a single sample group, and sample groups with different tensile strength and bonding strength characteristics are identified.

2. A strength testing system for lining cloth production according to claim 1, It is characterized in that The tensile performance test module uses an adaptive algorithm to dynamically adjust the speed of tensile force application according to the real-time monitored force and displacement data feedback until the lining cloth sample breaks, and records the force change and displacement data during the stretching process. When converting the force change and displacement data into stress-strain data, it includes: Apply tension to the lining cloth sample at a preset initial speed, and start recording the force and displacement data during the stretching process; Continuously monitor the change data of force and displacement in real time, and calculate the real-time stress and strain; According to the stress and strain data calculated in real time, set the stress change rate and strain change rate; The speed of applying the tension is adjusted according to the stress change rate and the strain change rate using an adaptive algorithm; According to the adjusted application speed, the force change and displacement data during the stretching process are recorded in real time, and the force change and displacement data are converted into stress-strain data.

3. A strength testing system for lining cloth production according to claim 2, It is characterized in that The method of adjusting the tension application speed according to the stress change rate and the strain change rate by using an adaptive algorithm includes: Real-time analysis of stress change rate and strain change rate; If the stress change rate decreases or the strain change rate increases, the tension application speed is slowed down; If the rate of change of stress increases or the rate of change of strain decreases, the speed of applying tension is increased; Dynamically adjust the speed at which the tension is applied based on the analysis results.

4. A strength testing system for lining cloth production according to claim 3, It is characterized in that The method of presetting a minimum point count threshold based on the expected cluster density, and running a density-sensitive anomaly detection algorithm using the selected neighborhood distance value and the preset minimum point count threshold to cluster data points and identify core points, edge points, and abnormal data points includes: Create a list containing all data points and obtain the neighborhood distance value and the preset minimum point number threshold; Calculate the distance between the current point and all other points in the list, and obtain all data points whose distance is less than or equal to the neighborhood distance value as neighbors; For each data point, if the number of neighbors is greater than or equal to the preset minimum point number threshold, it is marked as a core point; Add the unlabeled neighbors of the core point to the same cluster and recursively expand the neighbors of the neighbors; For data points that fail to form clusters, check whether the data points have core point neighbors; if they have core point neighbors, they are marked as edge points; if they do not have core point neighbors, they are marked as abnormal data points.

5. A strength testing system for lining cloth production according to claim 1, It is characterized in that The statistical analysis of tensile strength and bonding strength of the samples in each sample group is performed to identify the strength performance of each sample group, and to compare the performance differences between different sample groups. The commonalities of each sample group are analyzed, including: Collect sample data including tensile strength and adhesive strength within each sample group; Performing statistical analysis on sample data within each sample group; Statistical test methods were used to analyze performance differences between groups; Based on the statistical analysis results, identify the best and worst performing sample groups; The statistical results of different groups were compared to identify the common factors affecting tensile strength and adhesive strength.

Images