A system for stretch fabric strain sensor layout design

Abstract

The application discloses a system for stretch fabric strain sensor layout design, comprising: a fabric sensor layout algorithm, including the following steps: processing a human motion dataset, calculating the stretch rate of edges to represent skin deformation in motion; through skin deformation data, and through four characteristic edge selection methods of stretch extremum method, variance method, autocovariance and correlation coefficient method, and composite standard method, screening characteristic edges capable of representing skin deformation caused by motion to constitute a stretch hot zone; through a K-Means clustering algorithm, classifying the stretch hot zone according to an expected sensor number, and generating virtual sensors in each cluster; defining coverage as a main objective function, and redundancy as an auxiliary objective function, optimizing the sensor layout to maximize the coverage of the sensor on the stretch hot zone, and finally outputting a fabric sensor layout scheme. According to the application, the layout precision is significantly improved, the layout design cost is greatly reduced, and the design iteration efficiency is greatly improved.

Description

Technical Field

[0001] This invention relates to the technical field of fabric sensors, and in particular to a system for the layout design of strain sensors for tensile fabrics. Background Technology

[0002] Human motion capture technology, capable of real-time monitoring and recording of human movement data, has been widely applied in various fields. Among them, human motion capture based on fabric strain sensors has become a highly promising research direction. Fabric sensors, through the piezoresistive properties of their conductive fiber structure, convert the stretching or compressive deformation of the skin during movement into changes in resistance. These resistance changes can be measured and analyzed, thereby enabling the capture and monitoring of human movement. Fabric sensors offer advantages in motion monitoring such as comfort, flexibility, and non-invasiveness, making the motion capture process more natural and unrestricted, suitable for continuous monitoring of body movement. Fabric strain sensors have demonstrated significant application value in tasks such as joint angle prediction, posture estimation, and motion classification, showing great potential in rehabilitation medicine, sports, and human-computer interaction.

[0003] Human movement is generated by muscle contraction driving the skeletal system, resulting in the stretching or compression of surrounding skin tissue. When fabric stretch sensors come into contact with the skin, they undergo similar changes due to skin deformation, thus enabling motion monitoring. In sensor placement design, research often uses skin deformation to approximate the deformation of tight-fitting clothing. Due to the complex characteristics of skin, including high elasticity, pre-stress, and anisotropy, deformation patterns are complex and diverse, and skin deformation models differ significantly among individuals with different body characteristics. This variability necessitates targeted sensor placement based on different populations and application scenarios to avoid problems such as sensors failing to produce effective responses or outputs exceeding their response range, ensuring accurate motion capture. Therefore, the placement of fabric stretch sensors is crucial to the results of human motion monitoring.

[0004] Currently, there are various methods for placing fabric sensors. Experience-based methods rely primarily on researchers' past practical experience, depending on individual intuition regarding specific scenarios and simple movements to determine sensor placement and quantity. While these methods do not require complex computing equipment, they lack systematic scientific rigor and struggle to address diverse populations and complex movements. Anatomy-based methods utilize human anatomy and movement physiology, placing sensors based on key muscle groups and skeletal joints; however, this method demands a high level of specialized knowledge. Optical motion capture system-based methods can accurately measure skin deformation during human movement to plan sensor placement, providing accurate data for optimized layout and detailed capture. However, they are costly, complex to operate, require demanding environmental conditions, and rely on specialized personnel, limiting widespread application. These challenges significantly restrict the effectiveness of fabric sensor technology in motion monitoring, thus hindering its broader application and development. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the present invention aims to provide a system for the layout design of strain sensors for tensile fabrics. Based on three-dimensional human point cloud data analysis, it achieves intelligent sensor placement through calculation of skin deformation thermal zones, feature edge filtering, and K-Means clustering optimization. The system adopts a modular architecture, integrating a parametric customization interface, an automated calculation engine, and a visualization output module, providing non-professionals with an efficient and accurate sensor layout design tool. To achieve the above-mentioned objectives and other advantages of the present invention, a system for the layout design of strain sensors for tensile fabrics is provided, comprising:

[0006] A fabric sensor layout algorithm, which generates a sensor layout scheme through data-driven processing and multiple parameters, specifically includes the following steps:

[0007] Process human motion datasets and calculate edge stretching rates to represent skin deformation during motion;

[0008] Using skin deformation data, and through four feature edge selection methods—stretching extreme value method, variance method, autocovariance and correlation coefficient method, and composite standard method—feature edges that can represent skin deformation caused by motion are selected to form a stretching hot zone.

[0009] The tensile hot zone is classified by the desired number of sensors using the K-Means clustering algorithm, and a virtual sensor is generated in each cluster.

[0010] Coverage is defined as the primary objective function, and redundancy is defined as the secondary objective function. By optimizing the sensor layout to maximize the sensor coverage of the stretching hot zone, the final output is the fabric sensor layout scheme.

[0011] Preferably, the specific steps for processing the human motion dataset are as follows:

[0012] Create an adjacency matrix containing adjacency information and coordinate information of the connecting points;

[0013] Construct a vertex-edge matrix using an adjacency matrix to record the coordinates and indexes of the vertices at both ends of an edge;

[0014] Construct the point-coordinate matrix corresponding to each action and body part;

[0015] The changes of each edge endpoint are identified using the point-edge matrix and the point-coordinate matrix, the edge length is calculated, and an edge-edge length matrix corresponding to each movement is constructed.

[0016] The edge-stretch rate matrix is ​​constructed by calculating the stretch rate of each edge in each frame relative to the first frame using an edge-edge length matrix. This matrix is ​​used to calculate the deformation of the human skin surface caused by motion.

[0017] Preferably, the method for filtering feature edges that can represent skin deformation caused by motion includes:

[0018] Stretching extreme value method: When the maximum stretch of an edge during the movement exceeds a user-defined threshold, the edge is included in the feature set;

[0019] Difference method: Calculate the variance of the stretch rate of each edge throughout the entire movement. An edge with a large variance in its stretch rate indicates a significant stretch change, and is therefore selected as a characteristic edge.

[0020] Autocovariance and correlation coefficient method: Calculate the autocovariance and correlation coefficient of each edge in different action sequences to assess its similarity across different actions. Edges with large autocovariance and small correlation coefficients can better distinguish motions, and these edges are selected as feature edges;

[0021] Composite criterion: Ranking edges based on a combination of elongation, variance, autocovariance, and correlation coefficient to establish a comprehensive ranking index.

[0022] Preferably, the step of classifying the tensile hot zone using the K-Means clustering algorithm to the desired number of sensors and generating a virtual sensor in each cluster specifically includes the following steps:

[0023] K-Means clustering is used to divide the selected feature edges into clusters equal to the number of sensors;

[0024] A virtual sensor is generated based on the center of the feature edge as the center of the sensor and the custom sensor length.

[0025] The orientation of each sensor is determined through target optimization calculations, ultimately generating a sensor layout scheme.

[0026] Compared with the prior art, the beneficial effects of this invention are:

[0027] 1. Accurately determine the sensor layout

[0028] This invention achieves scientific and precise sensor positioning through four-dimensional point cloud data analysis and multiple feature extraction algorithms. Based on high-precision three-dimensional human point cloud data and temporal motion information, the system can accurately calculate the skin deformation pattern during movement. Specifically, based on 40,000+ grid points and 129 dynamic sequences from the D-Faust dataset, the system calculates a 6890×3×200 point cloud matrix to accurately capture skin deformation patterns. Four feature selection methods are employed: stretching extremum method, variance method, autocovariance and correlation coefficient method, and composite standard method, enabling adaptive matching for different joint types and motion patterns. For example, the stretching extremum method can be used to improve accuracy for single-degree-of-freedom joints, while the variance method effectively suppresses skin wrinkle interference for multi-degree-of-freedom joints. Simultaneously, the system supports parameterized customization based on gender, BMI, body part, and motion pattern, ensuring that each layout scheme accurately adapts to the target population.

[0029] 2. Reduce technical complexity and cost

[0030] This invention simplifies the traditional sensor layout design workflow, transforming a manual approach reliant on expensive equipment and complex operations into an automated solution based on a software platform. By utilizing pre-calculated and stored human deformation models, the system eliminates the need for on-site testing and equipment calibration, thus removing reliance on optical motion capture devices. The digital design process supports batch processing and template reuse, resulting in extremely low marginal costs. The entire layout design process can be completed using only standard computing equipment, significantly reducing hardware investment. Simultaneously, the system's automation reduces labor costs and time investment, making professional-grade sensor layout design services affordable even for small organizations and resource-constrained environments.

[0031] 3. Reduce reliance on specialized knowledge

[0032] This invention transforms complex professional knowledge into standardized algorithms and an intuitive user interface, significantly lowering the barrier to entry. Through modular design, the system encapsulates expertise from multiple fields, including human anatomy, sensor technology, and data analysis, within its algorithms, allowing users to complete layout designs without requiring specific professional backgrounds. The visual user interface and parametric design process enable even non-professionals to quickly learn and achieve high-quality design results. The standardized operating procedures ensure the consistency and reliability of the design results, unaffected by the operator's individual experience or knowledge level. This design philosophy greatly promotes the popularization and application of fabric sensor technology.

[0033] 4. Improve design efficiency

[0034] This invention significantly improves the efficiency of sensor layout design through algorithm automation and process optimization. The system automates the entire process from parameter setting to layout scheme generation, allowing users to obtain a complete design scheme by selecting only a few parameters. Real-time 3D visualization preview supports rapid iterative optimization, enabling users to immediately view the effects of parameter adjustments and make fine-tuning adjustments. Batch design functionality allows the system to handle multiple design requirements simultaneously, making it suitable for large-scale application scenarios. Compared to traditional manual design methods, this invention reduces the design cycle from weeks to hours, achieving an efficiency improvement of two orders of magnitude. This high efficiency not only saves significant time and labor costs but also enables rapid prototyping and agile design methodologies.

[0035] 5. The system has high stability and strong scalability.

[0036] This invention employs a modular architecture design, with each functional module relatively independent and standardized interfaces, ensuring system stability and maintainability. The data processing module supports multiple input data formats, exhibiting good compatibility and scalability. The layout algorithm utilizes a validated mathematical model, resulting in stable, reliable, and highly reproducible calculations. The system reserves interfaces for algorithm upgrades and functional expansion, supporting future technology integration and upgrades. The standardized output format facilitates integration with other tools and platforms, supporting end-to-end digital management. The modular design philosophy and open system architecture ensure the invention's long-term technological viability and broad development prospects. Attached Figure Description

[0037] Figure 1 This is a diagram of the core interface of the software tool for the system of strain sensor layout design for tensile fabrics according to the present invention.

[0038] Figure 2 The flowchart of the layout optimization algorithm for the system of strain sensor layout design for tensile fabrics according to the present invention is shown.

[0039] Figure 3 The feature edge graph of the system for the layout design of strain sensors for tensile fabrics according to the present invention is selected by the autocovariance and correlation coefficient method.

[0040] Figure 4 The flowchart of the sensor layout method based on K-Means clustering optimization for the system of tensile fabric strain sensor layout design according to the present invention is as follows: (a) Clustering feature edges into six groups; (b) Generating virtual sensors in each cluster; (c) Determining the layout scheme.

[0041] Figure 5 A schematic diagram illustrating the calculation of sensor coverage and redundancy in a system for the layout design of strain sensors for tensile fabrics according to the present invention.

[0042] Figure 6 A UI diagram showing the selection of gender, body type, placement location, and motion parameters for a system based on the present invention for the layout design of strain sensors for tensile fabrics.

[0043] Figure 7 The system UI diagram of the system for the layout design of strain sensors for tensile fabrics according to the present invention is shown in (a) the interface for displaying tensile hot zones selected by four feature edge selection methods and (b) the interface for displaying and adjusting layout schemes. Detailed Implementation

[0044] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] Reference Figure 1 A system for designing a layout of strain sensors for tensile fabrics, comprising:

[0046] A fabric sensor layout algorithm, which generates a sensor layout scheme through data-driven processing and multiple parameters, specifically includes the following steps:

[0047] This study processes human motion datasets, calculating edge stretching rates to represent skin deformation during movement; and employs digital human models to calculate skin deformation. The open-source dataset D-Faust contains two genders, five body types, and 129 dynamic sequences. It captures over 40,000 grid points on the human surface, each corresponding to a specific location on the surface. The dataset achieves near-optical motion reproduction accuracy, with 93% of data points having an error within 1mm. This high precision ensures a one-to-one correspondence between the dataset and actual human surface points, guaranteeing the reliability of the calculation results. The dataset is reshaped into a 6890×3×200 format, containing 3D coordinate data of 6890 vertices on the human surface over 200 frames. The distance between two points on the human surface is defined as an edge. As the human moves, the coordinates of points change, causing variations in the length and direction of edges in different human datasets. To capture this temporal variation, the raw data is processed as follows:

[0048] Create an adjacency matrix containing the adjacency information of the connecting points and the coordinate information of the points.

[0049] Use an adjacency matrix to construct a vertex-edge matrix that records the coordinates and indexes of the vertices at both ends of an edge.

[0050] Construct the point-coordinate matrix corresponding to each action and body part.

[0051] The changes of each edge endpoint are identified using the point-edge matrix and the point-coordinate matrix, the edge length is calculated, and an edge-edge length matrix corresponding to each movement is constructed.

[0052] The edge-stretch ratio matrix is ​​constructed by calculating the stretch ratio of each edge in each frame relative to the first frame, which is set as the reference frame. This matrix is ​​used to calculate the deformation of the human skin surface caused by motion.

[0053] Using skin deformation data, and employing four feature edge selection methods—the stretching extreme value method, the variance method, the autocovariance and correlation coefficient method, and the composite standard method—feature edges representing motion-induced skin deformation are selected to form a stretching hot zone. The four feature edge selection methods specifically include:

[0054] The stretching extreme value method: When the maximum stretch of an edge during the movement exceeds a custom threshold, the edge is included in the feature set.

[0055] Variance method: Calculate the variance of the stretching rate of each edge throughout the entire movement. A larger variance in the stretching rate of an edge indicates a significant stretching change, and the edge is selected as the characteristic edge.

[0056] Autocovariance and correlation coefficient method: Calculate the autocovariance and correlation coefficient of each edge in different action sequences to assess its similarity across different actions. Edges with large autocovariance and small correlation coefficients can better distinguish motions, and these edges are selected as feature edges.

[0057] Composite criterion: Ranking edges based on a combination of elongation, variance, autocovariance, and correlation coefficient to establish a comprehensive ranking index.

[0058]

[0059] Where λ represents the stretching ratio of the edge, s 2 Let represent the variance of the edges, and α, β, and γ be the weights assigned to the stretching extreme value method, variance method, autocovariance method, and correlation coefficient method, respectively. f(n) is the sum of rankings based on these weighted criteria.

[0060] These four feature selection methods can be used to determine a set of feature edges, which constitute the stretching thermal zone on the human body surface, representing the area of ​​skin deformation caused by movement.

[0061] The stretched thermal zone is classified using the K-Means clustering algorithm based on the desired number of sensors, and a virtual sensor is generated in each cluster. K-Means clustering is also used to divide selected feature edges into clusters equal to the number of sensors. Virtual sensors are generated based on the center of each feature edge and the custom sensor length. The orientation of each sensor is determined through target optimization calculations, ultimately generating a sensor layout scheme. The entire process is as follows: Figure 4 As shown.

[0062] To determine the sensor layout scheme corresponding to the stretching heat zone, two metrics are defined: coverage, as the primary objective function, measures the sensor's effectiveness in capturing the stretching heat zone; redundancy, as an auxiliary evaluation function, assesses the degree of overlap in the sensor's effective range. The following assumptions are made regarding the sensor placement on the skin:

[0063] The sensor is placed close to the skin surface, and its thickness is negligible.

[0064] The longitudinal deformation of the sensor is much greater than its lateral deformation; therefore, the sensor's width w and lateral elongation λ are... β It is considered a constant value.

[0065] The effective operating range of the sensor is determined by the product of its length l and width w, and is further determined by the longitudinal tensile ratio λ. α and transverse elongation λ β Adjust the proportions.

[0066] The coverage area of ​​each sensor is determined by its center coordinates (C). x C y C z The coordinates of the four corners (x1, y1, z1) to (x4, y4, z4) are defined. If a point in the human body point cloud falls within the coverage area of ​​a sensor, it is considered to be covered by that sensor. Simultaneously, redundancy is calculated by determining whether a point is captured by multiple sensors. A schematic diagram of this equivalent model is shown below. Figure 5 As shown. The calculation methods for coverage and redundancy are as follows:

[0067]

[0068] Where m is the number of points within the tensile thermal zone covered by each generated sensor, n is the number of sensors, and N is the total number of coordinate points within all tensile thermal zones. min and s max These are the minimum and maximum longitudinal stretch ratios, respectively. In the redundancy calculation, C... j This indicates the number of sensors covering a specific tensile heat zone. When a point is covered by multiple sensors, the function I(C) represents the number of sensors covering that zone. j The value of >1) is 1, otherwise it is 0.

[0069] Coverage is defined as the primary objective function, and redundancy is defined as the secondary objective function. By optimizing the sensor layout to maximize the sensor coverage of the stretching hot zone, the final output is the fabric sensor layout scheme.

[0070] Furthermore, based on the aforementioned fabric sensor layout algorithm, this invention has developed a digital sensor layout design system on the MATLAB platform (MathWorks, R2021b). This system is used to design and optimize the layout of fabric strain sensors in human motion capture. Although implemented using MATLAB, the algorithm logic is universal and can be ported to other development platforms such as Python and C++. The system includes:

[0071] The data initialization module is used to load and parse human motion data. The system uses the open-source D-Faust human motion dataset as its base data. This dataset was obtained through high-precision optical scanning and includes two genders (male and female), five body types (based on BMI classification: underweight, standard, overweight, obese, and severely obese), and 129 dynamic sequences. Each sequence records the deformation information of the human body surface during a specific movement. The system is designed with a flexible data interface, supporting users to import custom datasets. Users only need to prepare data files according to the data format based on the SMPL model (including point cloud format, marker information, time series labels, etc.), and the system can automatically complete data loading and preprocessing. This design enables the system to adapt to different research scenarios and special application needs.

[0072] The parameter setting module provides a visual interface for users to adjust multi-dimensional parameters. The system offers a graphical user interface (GUI) with a tabbed layout, categorizing and organizing complex parameter settings. Through this interactive interface, users can select gender, body type, body part, and type of exercise to monitor. The system provides three gender options (female, male, and unrestricted), five different BMI ranges, and allows users to choose from preset body parts or customize the sensor placement area using coordinates. Users can also select individual actions or combinations of actions to meet specific needs. These selections are subsequently used to process data loaded during the initialization phase, such as... Figure 6 As shown.

[0073] Furthermore, the system's visual interface design fully considers user-friendliness and functional completeness, providing gender selection (male / female / unisex design options), body shape customization based on BMI values ​​(covering 5 standard ranges), flexible selection of body parts (both preset parts and custom coordinates are supported), and comprehensive coverage of movement types (single or combined movements are available). Users can adjust key sensor parameters according to actual needs, including the number, length, and stretch ratio. All setting changes can be previewed in real time in the 3D model, ensuring users can intuitively evaluate and optimize design schemes. This integrated design concept not only improves operational efficiency but also significantly lowers the professional knowledge threshold, enabling non-professionals to quickly master and use the system for high-quality sensor layout design.

[0074] The system includes a layout generation module, which automatically calculates the optimal sensor position based on an algorithm; and a deformation calculation engine, which performs real-time calculations of skin deformation patterns based on preprocessed point cloud data. This engine employs a hierarchical computing architecture, first establishing the topological structure of the human body surface, and then dynamically maintaining an edge-to-edge length matrix to achieve real-time updates of the distance between any two points on the human body surface. Using the initial frame of the motion sequence as a reference, the system calculates the deformation rate of all edges in each subsequent frame, thereby quantifying the degree of skin stretching and compression. Through an efficient mapping algorithm, the system remaps the two-dimensional stretch rate data onto a three-dimensional human body surface model, generating an intuitive deformation heatmap.

[0075] Feature extraction algorithm implementation: The system adopts a parallel computing architecture and executes four feature extraction algorithms simultaneously. Users can choose a single method or combine the results of multiple methods by setting weight parameters according to their actual needs.

[0076] The stretching extreme value algorithm systematically scans all edges by setting a threshold parameter based on the minimum stretching range of the sensor, marks the edges whose stretching rate exceeds the threshold, generates a binary feature map, and automatically selects the top 5% of edges with the largest stretching amplitude as candidates.

[0077] The analysis of variance algorithm calculates the variance of the stretching change of each edge in the complete motion sequence, and uses an adaptive threshold mechanism to identify the edges with the largest stretching fluctuations. Typically, the edges with the top 5% of variance are selected.

[0078] The correlation analysis algorithm constructs a correlation matrix between different action modes. By calculating the autocovariance and cross-correlation coefficient of each edge, it sorts the edges based on composite indices and selects the feature edges that best distinguish different actions.

[0079] The composite evaluation algorithm integrates the calculation results of the first three methods, allows users to adjust the weight parameters of each component, and finally outputs the edge set with the highest comprehensive score.

[0080] Layout scheme generation process: Based on the output of four feature edge selection methods, the system calculates and visualizes the skin deformation area caused by motion, and generates the corresponding layout scheme after the user determines the sensor parameters. The system uses an optimized K-Means clustering algorithm to process the feature edge data. First, it randomly selects k edge center points as initial cluster centers, then calculates the distance from all feature edges to each cluster center, and assigns each edge to the nearest center. After clustering, the system traverses all edges in each cluster and calculates the coverage index based on the geometric definition of coverage, finding the edge position that achieves the maximum coverage through iterative optimization. When the user selects one of the four feature methods, the system displays the layout scheme generated by the corresponding method, and provides detailed quantitative analysis results, such as... Figure 7 As shown.

[0081] The adjustment and optimization module allows users to adjust parameters and recalculate layout schemes in real time. The system presents the layout scheme generated based on default parameters in 3D, and users can adjust it in real time through various methods. The interface provides intuitive slider controls, allowing users to adjust the number of sensors and standard length parameters, and modify the sensor stretch ratio settings through a dedicated input window. Whenever the user adjusts the parameters, the system immediately triggers a recalculation process, generating an updated sensor layout scheme and displaying the adjustment effect in real time on the current interface. This instant feedback mechanism enables users to quickly iterate and optimize design schemes and intuitively evaluate the impact of different parameter combinations on the layout effect.

[0082] The output module generates a final visual layout diagram and a detailed quantitative evaluation report. The system produces rich visual output content, providing users with comprehensive design information. In the 3D layout diagram, the system accurately displays the specific location of each sensor on the human body, allowing users to view detailed information, including key parameters such as sensor orientation and angle, through rotation and zoom. The deformation heatmap function uses color-coding technology to intuitively display the spatial distribution of skin deformation during movement; different colors represent different tensile intensities, allowing users to clearly observe key deformation areas. The coverage analysis diagram provides a quantitative relationship curve between sensor coverage and the number of sensors, helping users find the optimal balance between sensor cost and monitoring accuracy. All visualization results support high-resolution export, facilitating subsequent report creation and technical communication.

[0083] The number of devices and processing scale described herein are for simplification of the invention. Applications, modifications, and variations of this invention will be readily apparent to those skilled in the art. Although embodiments of the invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. It can be applied to various fields suitable for this invention, and further modifications can be readily implemented by those skilled in the art. Therefore, without departing from the general concept defined by the claims and their equivalents, this invention is not limited to the specific details and illustrations shown and described herein.

Claims

1. A system for the layout design of strain sensors for tensile fabrics, characterized in that, include: A fabric sensor layout algorithm, which generates a sensor layout scheme through data-driven processing and multiple parameters, specifically includes the following steps: Process human motion datasets and calculate edge stretching rates to represent skin deformation during motion; Using skin deformation data, and through four feature edge selection methods—stretching extreme value method, variance method, autocovariance and correlation coefficient method, and composite standard method—feature edges that can represent skin deformation caused by motion are selected to form a stretching hot zone. The stretching hot zone is classified by the desired number of sensors using the K-Means clustering algorithm, and a virtual sensor is generated in each cluster. Coverage is defined as the primary objective function, and redundancy is defined as the secondary objective function. By optimizing the sensor layout to maximize the sensor coverage of the stretching hot zone, the final output is the fabric sensor layout scheme.

2. The system for designing a strain sensor layout for tensile fabrics as described in claim 1, characterized in that, The specific steps for processing the human motion dataset are as follows: Create an adjacency matrix containing adjacency information and coordinate information of the connecting points; Construct a vertex-edge matrix using an adjacency matrix to record the coordinates and indexes of the vertices at both ends of an edge; Construct the point-coordinate matrix corresponding to each action and body part; The changes of each edge endpoint are identified using the point-edge matrix and the point-coordinate matrix, the edge length is calculated, and an edge-edge length matrix corresponding to each movement is constructed. The edge-stretch rate matrix is ​​constructed by calculating the stretch rate of each edge in each frame relative to the first frame using an edge-edge length matrix. This matrix is ​​used to calculate the deformation of the human skin surface caused by motion.

3. The system for designing a strain sensor layout for tensile fabrics as described in claim 1, characterized in that, The method for filtering feature edges that can represent skin deformation caused by motion includes: Stretching extreme value method: When the maximum stretch of an edge during the movement exceeds a user-defined threshold, the edge is included in the feature set; Variance method: Calculate the variance of the stretching rate of each edge throughout the entire process of the action. If the variance of the stretching rate of an edge is large, it indicates that its stretching change is significant, and the edge is selected as the characteristic edge. Autocovariance and correlation coefficient method: Calculate the autocovariance and correlation coefficient of each edge in different action sequences to evaluate its similarity in different actions. Edges with large autocovariance and small correlation coefficient can better distinguish motions, and such edges are selected as feature edges. Composite criterion: Ranking edges based on a combination of elongation, variance, autocovariance, and correlation coefficient to establish a comprehensive ranking index.

4. The system for designing a strain sensor layout for tensile fabrics as described in claim 1, characterized in that, The process of classifying the tensile hot zones using the K-Means clustering algorithm to determine the desired number of sensors and generating virtual sensors in each cluster specifically includes the following steps: K-Means clustering is used to divide the selected feature edges into clusters equal to the number of sensors; A virtual sensor is generated based on the center of the feature edge as the center of the sensor and the custom sensor length. The orientation of each sensor is determined through target optimization calculations, ultimately generating a sensor layout scheme.

5. The system for designing a strain sensor layout for tensile fabrics as described in claim 1, characterized in that, It also includes a data initialization module, which is used to load and parse human motion data; The parameter setting module is used to support users in adjusting parameters in multiple dimensions through a visual interface. A layout generation module, which is used to automatically calculate the optimal position of the sensor according to an algorithm; An adjustment and optimization module is provided to allow users to adjust parameters and recalculate layout schemes in real time. The output module is used to generate a final visual layout diagram and a detailed quantitative evaluation report.

Images