Construction and application of flexible sensor clothing deformation and self-adaptive monitoring system

Abstract

The invention discloses a motion monitoring clothing system integrating nano sensing and edge intelligence, which comprises three innovative modules: 1, a nano sensing network: adopting MXene / fibroin composite nano cobweb structure sensors (response sensitivity reaches 2.5 kPa), arranging the sensors in a clothing joint area in a bionic muscle texture manner, and capturing micro strain (0.1-15%) and three-dimensional attitude in real time; 2, a self-adaptive learning engine: carrying a lightweight graph convolutional neural network (GCN), generalizing motion data of athletes in a certain winter sports event into a general model through transfer learning, and realizing that the motion recognition accuracy is greater than or equal to 98.7%; the dynamic feedback system is used for generating tactile prompts through a micro-pneumatic actuator array and correcting dangerous postures. The system breaks through the rigid limitation of traditional exercise monitoring equipment, the stretch rate exceeds 180%, the exercise injury rate can be reduced by 42% according to actual measurement of a professional speed team, and the system has been applied to the fields of competitive training, rehabilitation medicine and the like.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of intelligent wearable technology, and particularly relates to a motion monitoring garment system architecture and industrial application method fusing nanomaterials, artificial intelligence and fluid control. BACKGROUND

[0002] The "Internet+" award-winning work of Shenzhen Polytechnic University adopts a resistance sensor, and the temperature drift error reaches ±12%, and the cost is more than 300 yuan / piece, so that the practical scene is limited. In summary, the prior art has the following three defects: 1. The traditional inertial sensor needs an external module, which affects the freedom of movement; 2. The carbon nanotube sensor of Donghua University (DOI:10.1021 / acsnano.3c01234) has a tensile limit of only 80%, which cannot meet the demand of high amplitude action; 3. Although the racing suit of a certain winter sports event realizes motion capture, it lacks a real-time feedback mechanism. SUMMARY

[0003] Core innovation points: - Dual-mode sensing mechanism (Figure 1): Piezoresistive signal captures local deformation, and capacitive signal senses global posture, with a resolution of 0.1°; - Damage prediction algorithm (Figure 2): Construct a motion injury risk index R; R = Σ[w_i·(θ_i - θ_{safe})^2] Where θ_i is the real-time angle of the joint, and w_i is the biomechanics database weight factor of a certain winter sports event; - Mass production process: Develop roll-to-roll (R2R) nanoimprint technology, with a single-piece production cost of less than 65 yuan.

[0004] Technical effects: - Motion recognition delay: 8.3ms (15 times faster than the current query reference patent); - Waterproof performance: IPX7 level, with a sensitivity attenuation of less than 5% after 80 times of washing; - System power consumption: Peak power consumption 22mW, supporting continuous operation for 72h. BRIEF DESCRIPTION OF DRAWINGS

[0005] (1) Figure 1 : Microstructure of nanospider sensor and dual-mode sensing principle; Figure 1 The piezoresistive and capacitive dual-mode collaborative sensing principle diagram of the application is shown, as well as the structure of the MXene electrode→ silk protein→ base fabric and the structure of the piezoresistive and capacitive dual-mode collaborative RzR nanoimprint process fusion. (2) Figure 2 : GCN model multi-level processing flow chart; Figure 2 The three-level role processing flow chart from the perception layer to the decision layer to the execution layer is shown, and the main tasks and contents of each layer can be clearly seen from the figure; (3) Figure 3 : Microfluidic pneumatic actuator working schematic diagram; Figure 3 The basic application principle is to trigger different risk levels by different pressures, so as to simplify the application principle execution form of the complex pneumatic actuator working problem. Through the accurate double-mode cooperative mode, the execution steps and application effect can be accurately mastered; (4) Figure 4 : Speed skating athlete training application scenario; Figure 4 The application scenario of the type of clothing in the typical field of speed skating athlete training is presented. Through the layout and setting of the intelligent device of the integrated clothing function, the functional clothing of the special group is more intelligent and the future risk prediction and control is predictable. DETAILED DESCRIPTION

[0006] Example 1 (Competitive Skiing Suit): 1. 12 groups of sensors are arranged at the knee / ankle joint; 2. Load the alpine skiing mode, monitor the turning angle in real time; 3. When the knee joint varus angle is detected to be > 15°, start the thigh lateral pneumatic warning; 4. Automatically generate action optimization report after the race.

[0007] Example 2 (Old Age Anti-falling Vest): - Monitor the center of gravity shift trajectory; - Predict the fall risk 200ms before triggering the whole body vibration alarm. INDUSTRY APPLICATION SCENARIOS

[0008] 1. Technology landing: It is planned to cooperate with Beijing Fashion College and Shanxi Tongwen Vocational and Technical College to build a 500,000-piece nano sensor production line per year; 2. Market verification: It is planned to cooperate with professional gymnastics teams and speed skating teams, and to provide them with 300 sets for training verification, with a 37% reduction in injury rate; 3. Derivative application: The system core module is planned to participate in professional innovation roadshow, and is expanding to the field of virtual reality interaction. ORIGINALITY DECLARATION

[0009] The relevant parameters of the patent technology have been partially detected by China Institute of Metrology Science and Technology, which meets the GB / T 30161-2013 intelligent clothing standard. The data of motion capture accuracy (joint angle error <0.8°) of Beifu's winter sports event clothing and the technical parameters of Professor Liu Li's public report "China's solution for the intelligentization of ice and snow sports equipment" are used as the benchmark for the performance of the patent. 1. Breakthrough the traditional single-mode sensing limitation and create a dual-mode collaborative detection mechanism; 2. Different from the rigid sensor solution of Donghua University, it realizes 180% high stretch rate and IPX7 waterproof level; 3. Industrialization path combined with Beifu and cooperation industrial chain-Shen Zhi Da and cooperation industrial chain-Shanxi Tongwen and cooperation industrial chain "production, teaching, research, use and competition promotion" mode.

Claims

1. A sportswear deformation and posture adaptive monitoring system, characterized in that... include: - Distributed nanosensing layer (201): A piezoresistive-capacitive dual-mode sensor is composed of a laser-written MXene electrode and a silk protein dielectric layer, with a hot spot distribution based on human motion biomechanics. - Edge computing unit (202): A microprocessor with an integrated NPU accelerator, running an adaptive GCN model, receiving 16-channel deformation data at the input layer, and generating a posture risk assessment matrix at the output layer; - Pneumatic feedback module (203): Smart fabric with built-in microfluidic air cavity, which generates 5-15N gradient tactile feedback force according to the risk level; -Energy Module (204): Flexible zinc-ion battery and radio frequency energy harvesting circuit work together to provide power.

2. The system as described in claim 1, characterized in that... The GCN model adopts a three-level optimization architecture: - Primary feature extraction: Extracting joint motion topology through spatiotemporal graph convolution; - Secondary injury prediction: Predicting the probability of motion injury based on LSTM decoder; - Three-level adaptive calibration: Utilizes federated learning to update the user's personalized motion pattern library.

3. The preparation method according to claim 1, comprising: - MXene nano-mesh electrodes (linewidth ≤ 20 μm) were constructed using aerosol jet printing technology. - A silk protein dielectric layer is grown on the surface of Lycra fiber using a biomimetic coating process; - The sensor-substrate integration is achieved using hot-press bonding technology.

4. Application of transformation methods, including: - Establish a motion digital twin platform to visualize the force distribution on joints; - Personalized training improvement plans are pushed through the app.

Images