Au-VG-based flexible strain sensor, and preparation method and application thereof

Abstract

The invention discloses an Au-VG-based flexible strain sensor and a preparation method and application thereof, the flexible strain sensor is based on Au-VG, and under the conditions that the original length is 3 cm and the strain range is 3.3%-13.3% (delta L is equal to 1-4 mm), the sensor can realize that the average GF exceeds 240 and the linearity is Rgt; the surface resistance is less than or equal to 500 omega / sq, and the response time is lt; the resistance drift is less than 3% after 10,000 times of circulation, and excellent comprehensive performance is shown. The Au-VG sensor is integrated in the intelligent glove, and finally high-precision classification of ASL gestures of 26 English letters is achieved and reaches up to 99.6%. Compared with an existing identification system based on a crack metal or carbon-based composite material, the micro-strain identification system has remarkable advantages in the aspects of micro-strain identification precision, power consumption control, device service life, signal stability and the like.

Description

Technical Field

[0001] The present invention belongs to the field of sensor technology, and in particular relates to an Au-VG based flexible strain sensor, a preparation method thereof and applications. Background Art

[0002] With the continued global growth in demand for accessible information technology, natural interaction systems, and smart wearable devices, the need for flexible wearable devices with high sensitivity, low power consumption, structural integration, and mass production capabilities is becoming increasingly urgent. This is particularly true as governments and international organizations actively promote accessible information technology, leading to the continuous release of relevant regulations and standards. The rapid expansion of the global wearable device and drone markets is driving demand for new, high-performance sensing materials. Furthermore, the integration of new conductive materials (such as vertical graphene) with micro-nanofabrication technologies is enhancing sensor performance and manufacturability. Furthermore, the significant demand for high-precision wearable gesture recognition devices to assist communication among special populations is expanding into broader areas such as consumer electronics, industrial control, and medical rehabilitation.

[0003] Currently, sign language recognition and natural gesture interaction technologies are still in the research and development or initial commercialization stages, facing issues such as low recognition accuracy, high latency, and bulky equipment. While some camera- and visual recognition-based methods have been commercialized, they rely on specific environments, are sensitive to occlusion, and are computationally intensive, making them difficult to popularize in practical applications. In contrast, solutions based on flexible strain sensors can operate in non-visual environments and offer greater practicality and reliability. As a representative of emerging electronic materials, the commercialization of flexible sensors is still limited by unstable material properties and the difficulty of mass production.

[0004] In recent years, flexible, wearable strain sensors have garnered widespread attention as key sensing elements in gesture recognition, human-computer interaction, and health monitoring systems. Current research focuses on material configuration design and conductivity optimization. Typical technologies include metal crack structures, laser-induced graphene (LIG), carbon nanotube / graphene composite conductive networks, and three-dimensional vertical graphene (VG).

[0005] In the metal crack structure category, this approach typically involves depositing a thin metal film, such as silver (Ag) or gold (Au), on the surface of an elastomer. This induces crack propagation during strain loading, resulting in a strain gauge response. While this approach offers high initial sensitivity (GFs of up to a thousand), the irreversible propagation of the crack structure leads to significant device response degradation after more than 1,000 cycles, making it difficult to meet long-term operational requirements and resulting in insufficient stability.

[0006] Laser-induced graphene (LIG) can be obtained by direct laser etching of polyimide (PI), offering convenient preparation and excellent flexibility. Previous studies have reported biomimetic, hydrophobic LIG strain sensors with GFs as high as 565, demonstrating a certain degree of environmental adaptability. However, the slow change in ΔR / R0 within the microstrain range of 0–4.5% leads to insufficient sensitivity, and the high sheet resistance (>10 kΩ / sq) limits their application in low-power precision sensing scenarios.

[0007] Related research has also used a flexible pressure sensor based on an ion-composite nanofiber membrane, combined with a BPNN (back-propagation neural network) model, to achieve intelligent sign language recognition. The overall recognition accuracy for 24 alphabetical hand gestures, collected from 650 data samples, was 96.8%. However, this method performed poorly in distinguishing between the letters U and V, a shortcoming of the system. In practical applications, the accuracy of sign language recognition systems is crucial, and improving recognition accuracy is directly related to their practical value in assisting communication for the deaf and hard of hearing.

[0008] US 12025437 B2 Stretchable Strain Sensor Based on Vertical Graphene and Its Application discloses a VG–PDMS flexible strain sensor. This patent utilizes a PECVD process to grow a vertical graphene (VG) network on a PDMS surface, creating a three-dimensional conductive framework. This sensor achieves a linear response under high strain (up to 50%) and excellent cycling stability (>10,000 cycles). The device exhibits excellent mechanical compliance and overall conductivity, making it suitable for wearable applications.

[0009] However, this solution has the following shortcomings: the initial surface resistance of the material is still high (1~5kΩ / sq), which is not conducive to low-power electronic systems; the GF in the microstrain region (0~10%) is relatively low, which makes it difficult to meet the requirements of high-resolution gesture or physiological signal recognition; no crack structure or sensitization mechanism is introduced, and there is limited room for improving response sensitivity.

[0010] A recent study reported a VG / CNT / PDMS sandwich structure, constructing a multilayer conductive structure composed of wrinkled vertical graphene (VG) and cross-linked carbon nanotubes (CNTs). The device exhibited an extremely high gauge factor (GF>1300) in the 60–100% strain range, along with good linearity (R² ≈ 0.98), demonstrating strong tensile strength and structural continuity. However, the GF was only ≈ 45 in the 0–10% range, limiting microstrain detection capabilities. The CNT composition significantly affected interfacial stability, making localized delamination more likely during high-strain cycles. Furthermore, device fabrication involved multiple steps, including CNT coating, transfer, and compounding, hindering mass production and consistent control.

[0011] In an existing study, a sign language recognition system based on a flexible pressure sensor of an ion composite nanofiber membrane combined with a BPNN learning model achieved an accuracy of 96.8% for the recognition of the 24 English letters of American Sign Language. However, the two letters U\V are very easy to confuse, and it is impossible to recognize all 26 letters with high precision. The fuzzy recognition of the U / V letters remains a key bottleneck in this field. Summary of the Invention

[0012] The object of the present invention is to provide an Au-VG based flexible strain sensor, a preparation method thereof and an application in gesture recognition.

[0013] Based on the above objectives, the present invention adopts the following technical solutions: A preparation method based on Au-VG flexible strain sensor, the process is as follows: (1) Clean and dry the rigid substrate; (2) preparing vertical graphene by chemical vapor deposition on the rigid substrate of step (1); the obtained vertical graphene nanosheets have a height of 4 to 6 μm; (3) Physical vapor deposition of a gold film on the vertical graphene in step (2) with a thickness of 30 to 100 nm to obtain a VG-Au conductive structure; (4) The VG-Au conductive structure is transferred to the solidified lower PDMS substrate by peeling, and the PDMS liquid is covered on the VG-Au conductive structure and solidified to form an upper PDMS layer; (5) Cut the flexible sensing layer into appropriate sizes, form silver paste electrodes at both ends of the device, and then use a medical breathable film for secondary packaging.

[0014] Furthermore, the rigid substrate in step (1) is carbon paper, silicon wafer, alumina ceramic substrate, quartz wafer, etc. The rigid substrate needs to be cleaned and dried before use, and the thickness of the rigid substrate is 100~500μm.

[0015] Preferably, the rigid substrate in step (1) is ultrasonically cleaned in acetone, isopropyl alcohol and deionized water in sequence for 5 to 15 minutes each time, and then dried on a hot plate at 100 to 150° C. for later use.

[0016] Furthermore, in step (2), the process conditions of chemical vapor deposition are as follows: raw material gas: methane (CH4), gas flow rate is 5~10 sccm; chamber pressure is 6~12 Pa; RF power is 1000~2000 W; chamber temperature is 600~900℃; growth time is 5~40 minutes.

[0017] Preferably, the process conditions of the chemical vapor deposition are as follows: raw material gas: methane (CH4), gas flow rate is 5~8 sccm; chamber pressure is 8~10 Pa; RF power is 1000~1500 W; chamber temperature is 600~900℃; growth time is 5~40 minutes.

[0018] Furthermore, in step (3), the physical vapor deposition process conditions are as follows: working gas: argon (Ar), gas flow rate is 10~20 sccm; chamber pressure is 0.5~2 Pa; RF power is 100~200 W; chamber temperature is 100±10℃; deposition time is 10~50 seconds.

[0019] Preferably, the physical vapor deposition process conditions are as follows: working gas: argon (Ar), gas flow rate is 10~20sccm; chamber pressure is 0.5~2 Pa; RF power is 120 W; chamber temperature is 100±10℃; deposition time is 20~30 seconds.

[0020] Furthermore, in step (4), the PDMS pre-curing conditions are: heating at 90-110°C for 5-15 minutes, and the thickness of the upper and lower PDMS layers is 100-500 μm.

[0021] Furthermore, in step (5), silver paste leads are printed on both ends of the device and thermally cured to form silver paste electrodes. The curing temperature of the silver paste electrodes is 60-80° C., and the curing time is 10-30 minutes.

[0022] The Au-VG flexible strain sensor is prepared by the above preparation method.

[0023] The above-mentioned Au-VG flexible strain sensor is used in wearable devices and / or human-computer interaction fields.

[0024] The aforementioned Au-VG flexible strain sensor is used in gesture recognition. Six Au-VG flexible strain sensors are attached to the backs of the five fingers of a glove and the wrist, and connected to an integrated circuit board with wires for signal collection, forming a wearable signal acquisition system.

[0025] Gesture Data Collection and Classification: Five volunteers aged 22 to 30 demonstrated 26 alphabetical hand gestures while wearing smart gloves, repeating each gesture 10 times. This yielded 26 × 10 × 5 = 1,300 sets of multi-channel electrical signal samples. These samples were divided into a training set (780 samples), a validation set (260 samples), and a test set (260 samples) in a 6:2:2 ratio.

[0026] Data preprocessing: A 3-second raw signal (300 samples × 6 channels) is extracted from each sample. The mean of the first 100 samples in each channel is calculated as the baseline value. Subsequently, the corresponding baseline is subtracted from each channel in the 300 × 6 signal matrix to perform static offset correction. The corrected signal is transposed to a 6 × 300 channel-time format and added to the sample set. Finally, all samples are iterated to obtain the global maximum absolute value, which is then used to normalize each data point, uniformly mapping the signal amplitude to the range [–1, 1].

[0027] Gesture recognition model training and deployment: The training model used is an Inception-BiLSTM network that combines multi-scale convolution with a bidirectional temporal network. The Inception module extracts features from the input signal through multi-scale parallel convolution, capturing local dynamic features across different temporal receptive fields and concatenating them in the channel dimension to form a high-dimensional temporal feature map. The bidirectional LSTM module performs forward and backward transfers in time to model the complete temporal dependency. The model was trained using preprocessed data and subsequently achieved 99.6% classification accuracy for 26 letters in the test set.

[0028] This invention provides a flexible strain sensor product whose core functional layer is a vertical graphene-metal composite conductive network, which is ultimately transferred and cured onto a flexible polydimethylsiloxane (PDMS) substrate. During device fabrication, plasma-enhanced chemical vapor deposition (PECVD) is used to in-situ grow vertical graphene (VG) on a rigid substrate such as carbon paper. The height of the vertical graphene (VG) is controlled to be 4-6 μm, creating a three-dimensional conductive framework. Subsequently, a gold (Au) film with a nominal thickness of 50±10 nm is deposited on the VG surface using radio frequency physical vapor deposition (RF-PVD) to create a crack-tunneling synergistic conductive structure and enhance microstrain response. After the conductive sensitive layer is constructed, the entire VG / Au layer is transferred onto a pretreated PDMS substrate, achieving high adhesion and integration between the VG / Au layer and the PDMS. Lead areas are formed at the device edge by printing silver paste or metal evaporation, and are connected to a signal acquisition device via metal conductors. The flexible sensing unit is fixed on the backs of the five fingers and the wrist, forming a multi-channel distributed sensing network that fits the skin and is used to collect local strain signals caused by gesture movements.

[0029] The flexible strain sensor of the present invention can be applied to a variety of physiological signal monitoring. A wearable gesture recognition system based on this sensor can recognize gesture signals in real time. It is also suitable for human-computer interaction systems to complete the operation of drones. It is suitable for the following applications: 1. Gesture and sign language recognition: This product can be attached to the back of gloves, wrists, and other parts of the body to monitor the bending state of fingers and joints in real time. It can identify complex dynamic gestures such as American Sign Language (ASL) through strain signals, and can be used for communication assistance for the hearing-impaired or virtual reality control.

[0030] 2. Wearable Human-Computer Interaction System: By integrating with wireless communication modules, the sensor enables interactive motion control with mobile devices, robots, or drones. This low-power, high-accuracy, natural interaction capability makes it suitable for remote operation and augmented reality scenarios. Furthermore, the sensor can be integrated into motion analysis and human-computer interaction systems. A sensor glove worn on the hand with an additional sensor at the elbow can be used to identify and analyze shooting motions. The continuous change in sensor resistance throughout the entire shooting process, from the ball launch, lifting, wrist compression, to the release, can be used to identify shooting form and assist in optimizing training strategies to improve hit rate. In table tennis, the sensor glove can identify the differences in mechanical response between forehand and backhand movements during light blocks and heavy spikes, enabling movement judgment and technique analysis. The glove is also suitable for grasping analysis, effectively identifying variations in grip strength due to differences in stiffness, weight, or shape.

[0031] 3. Physiological Signal Monitoring: Attachable to the pharynx, eyelids, tendons, joints, and other areas, the sensor can detect subtle deformations such as speech, swallowing, blinking, and muscle vibrations, as well as larger deformations in areas like the knuckles and knees. This technology is useful in a variety of applications, including fatigue detection, rehabilitation assessment, silent input, and human-computer interaction. Regarding physiological signal monitoring, the sensor attached to the chest area can monitor strain response during normal and deep breathing, demonstrating significant differences in resistance changes. It can effectively distinguish respiratory rate and rhythm during standing, sitting, talking, and different gaits (e.g., slow and fast walking). Attached to the cheek or upper eyelid, the sensor can sensitively respond to facial expressions and eye movements, such as changes in resistance caused by subtle movements like smiling, frowning, blinking, and yawning, making it suitable for emotion recognition and neuromuscular disorder assessment.

[0032] In terms of joint movement monitoring, by sticking the sensor on the knee, it can accurately identify the strain response signals corresponding to different bending amplitudes, from knee extension (180°) to slight bending (120°), moderate bending (60°) and full squat state (0°); by sticking it on the ankle, it can be used to monitor the resistance changes caused by different forces (light, medium and heavy) during the accelerator and brake process in a simulated driving environment, which is helpful for driving behavior analysis and fatigue driving warning.

[0033] The present invention proposes to construct a crack-tunneling cooperative structure by in-situ depositing a metal film on the vertical graphene (VG) surface, thereby achieving the following key goals: Improved microstrain resolution: linear response R in the 3.3–13.3% strain range 2 >0.98, with an average gauge factor (GF) exceeding 240, significantly outperforming existing VG solutions. This allows for distinguishing subtle differences when used to collect different physiological signals, helping to improve gesture recognition. Reduce sheet resistance to match low-power system requirements: initial sheet resistance is controlled at ≤500Ω / sq, supporting direct reading of wireless communication modules such as Bluetooth; Improved cycle stability and response rate: After 10,000 cycles, ΔR / R0 drift is <3%, and response time is <100ms, meeting dynamic recognition requirements. Enhanced system adaptability: Sensors were integrated into gloves to collect six-channel signals from the fingertips and wrists. Combined with the Inception-BiLSTM deep learning model, this significantly improved the accuracy of ASL 26-letter recognition to 99.6%, validating its practicality.

[0034] In summary, the present invention addresses the key issues of existing flexible strain sensing technology such as "insufficient micro-strain sensitivity, high surface resistance, unstable response and complex manufacturing, and recognition defects when integrated into sign language recognition systems". It proposes a new solution path that coordinates the optimization of material structure, conductive mechanism and system integration, which has good academic innovation and engineering practical prospects.

[0035] The core concept of this invention is to construct a three-dimensional conductive network structure with synergistic enhancement of crack-tunneling, so as to significantly improve the sensitivity, linearity and stability of flexible strain sensors in the microstrain range. The sensor can also be integrated into wearable devices and combined with deep learning models to achieve accurate recognition of gestures. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 Schematic diagram of the structure of the sensor of the present invention; Figure 2 is the SEM image of vertical graphene (VG) in Example 1; Figure 3 is the resistance change rate ΔR / R0 of the sensor in Example 1 under different strains; Figure 4 The response time and recovery time of the sensor in Example 1 during loading and unloading; Figure 5 The significant resistance response changes of the sensor in Example 1 under different stretching displacements (1–9 mm); Figure 6 The cyclic sensing performance of the sensor in Example 1 during loading and unloading processes. After 10,000 cycles, no significant performance degradation was monitored. Figure 7 : This is the confusion matrix of the wearable glove including the sensor of Example 1 for the classification of 26 English letter gestures. DETAILED DESCRIPTION

[0037] The following examples are intended only to provide those of ordinary skill in the art with a complete disclosure and description of how to make and evaluate the compounds, compositions, articles, devices and / or methods described and claimed herein, and are intended to be illustrative only and not to limit the scope of what the inventors regard as their invention. Efforts have been made to ensure accuracy with respect to numbers (e.g., amounts, temperatures, etc.), but some errors and deviations should be accounted for.

[0038] Plasma Enhanced Chemical Vapor Deposition Equipment (PECVD) Used for in-situ growth of vertical graphene (VG) on rigid carbon paper substrates. The equipment must independently control RF power, gas flow, temperature, and chamber pressure, and is compatible with methane (CH4). Requirements include an RF power adjustment range of 1000-2000 W, an operating temperature of 600-900°C, precise gas control of 1-20 sccm, and chamber vacuum control capability of <5 Pa.

[0039] Radio Frequency Physical Vapor Deposition Equipment (RF-PVD) Used to deposit a 50 nm thick metal film on the VG surface. The equipment should be equipped with a high-purity gold target, 100~200 W RF power control function, a cavity temperature controllable to about 100°C, and a high vacuum system (background pressure <1×10 -3 Pa). Argon gas (10~20 sccm) can be stably introduced as the working atmosphere.

[0040] Six-channel signal acquisition control board and Bluetooth module This invention uses a custom circuit board to connect sensor electrodes and collect strain gauge signals. The control board includes a six-channel transimpedance amplifier (TIA), an analog multiplexer, a signal bias amplifier, a 32-bit MCU (with built-in ADC), and a Bluetooth module supporting BLE 5.0 or higher. Its communication power consumption is ≤5 mW. It is used to transmit multi-channel signals serially to a mobile phone or other terminal. This creates a low-power gesture recognition system with six channels of synchronous acquisition. The core utilizes a high-performance 8-bit microcontroller that simultaneously acquires analog signals from multiple finger bend sensors. After conditioning with dual high-precision operational amplifiers, the signals are digitized in parallel by the main control chip's 12-bit ADC. Only six of these channels are used, and the processed results are transmitted via Bluetooth low energy. Powered by a 3.7V lithium battery, it supports USB charging and a regulated 3.3V output. Equipped with a hardware power switch, it provides an integrated solution for synchronous multi-channel signal acquisition, high-precision processing, and wireless transmission.

[0041] Example 1 A preparation method based on Au-VG flexible strain sensor, the process is as follows: (1) Preparation of the lower substrate PDMS: Sylgard 184 PDMS prepolymer and curing agent were selected and mixed in a mass ratio of 10:1. The mixture was stirred at 300 rpm for 5 minutes. The mixture was placed in a vacuum environment for degassing for at least 10 minutes to remove bubbles. The degassed PDMS was poured into a clean mold, with a film thickness of approximately 200 μm. The film was thermally cured at 80°C for 2 hours to obtain a flexible lower substrate PDMS film.

[0042] (2) Growth of vertical graphene (VG) skeleton: Conductive carbon paper with a thickness of 150-200 μm was selected as a temporary rigid carrier, and ultrasonically cleaned in acetone, isopropyl alcohol and deionized water for 10 minutes each time. After drying, it was dried on a hot plate at 120°C for 10 minutes. It was placed in a plasma enhanced chemical vapor deposition (PECVD) chamber, and the process parameters were set to methane (CH4). It was grown for about 20 minutes under the conditions of methane flow rate of 5 sccm, chamber pressure of 10 Pa, RF power of 1000 W and temperature of 800°C. A three-dimensional vertical graphene network with a sheet height of about 5 μm can be grown in situ on the surface of the carbon paper. The SEM image is shown as follows: Figure 2 Figure 2 (a) shows an SEM image of vertical graphene (VG) prepared according to the present invention at 7kx magnification. It can be seen that VG grows uniformly on the substrate surface, forming a dense, highly vertical lamellar structure. The lamellar layers interweave to form multi-scale pore channels, creating an overall uniform three-dimensional network morphology. Figure 2 (b) shows an SEM image of VG at 40kx magnification. The lamellar edges are sharp, with wavy wrinkles and nanometer-scale thickness. The surface is smooth and has three-dimensional undulations. The gaps between the lamellar layers can be identified as submicron-scale pores at high magnification.

[0043] (3) Deposition of a 50nm gold film: After growth, the gold film was directly transferred to a physical vapor deposition (PVD) chamber without air exposure for gold film deposition. Using a high-purity gold target, sputtering was performed for 30 seconds at an argon flow rate of 15sccm, a chamber pressure of 1Pa, an RF power of 120W, and a temperature of 100°C to obtain an approximately 50nm thick Au film. This layer, in conjunction with the VG network, synergizes to form a crack sensing and conductive compensation channel, improving device sensitivity and conductive stability.

[0044] (4) Transfer and bonding of the entire structure: The Au-VG layer is transferred to the cured lower PDMS substrate by mechanical peeling. During the transfer process, the surface is kept flat and bonded to avoid curling or crack extension. This bonded structure constitutes the lower three-layer sensing core of the device.

[0045] (5) Upper PDMS encapsulation: Repeat the PDMS mixing and degassing process in step 1, and evenly cover the PDMS liquid on the surface of the Au layer, controlling the thickness to 200 μm. After standing at room temperature for 15 minutes for pre-curing, place it in an 80°C environment for 2 hours to form a soft upper encapsulation structure. The structure is as follows: Figure 1 As shown, it includes a lower substrate PDMS, a vertical graphene layer thickness, a gold film layer and an upper substrate PDMS.

[0046] (6) Device cutting and lead packaging: Laser cutting technology is used to cut the entire sensor into 35 mm × 5 mm. The middle 30 mm area is used as the effective strain sensing area. Silver paste leads are printed on both ends of the device and heat-cured. Then, a medical breathable film is used for secondary packaging. This packaging effectively prevents the Au layer from oxidation and sweat corrosion, improving the stability of the device.

[0047] The sensor prepared in Example 1 was subjected to step-by-step deformation by a stretching platform within the strain range of 0-17%, and the corresponding resistance changes were recorded and GF was calculated. The results are shown in FIG. Figure 3 As shown in the figure, the results show that the strain range is 3.3%~13.3% and the linearity is high (R 2 >0.98), GF reached 232.5, and the average GF exceeded 240.

[0048] The sensor of Example 1 was subjected to periodic stretching-releasing at a 1 Hz cycle, and the resistance variation curve was recorded. The results are shown in Figure 4 ,Depend on Figure 4 It can be seen that the response time and recovery time are both less than 100ms.

[0049] Figure 5 The sensor of Example 1 demonstrates the relative resistance change (ΔR / R0) at different strain amplitudes (1mm, 3mm, 5mm, 7mm, and 9mm). The peak signal shows a significant increasing trend as the displacement increases. Furthermore, multiple consecutive loads were applied at each displacement level, and the peaks and valleys of the waveforms overlapped significantly. The baseline after stretching and release consistently returned to near zero, with no drift or hysteresis observed.

[0050] Figure 6 The sensor of Example 1 was subjected to 10,000 stretch-release cycles at 3% strain, and the change in initial resistance and resistance after the cycles was monitored. The results showed that the drift after 10,000 stretch-release cycles was less than 3%.

[0051] Six sensors were attached to the backs of five fingers and the wrist of a glove (such as a commercially available disposable PVC glove). Wires were connected to an integrated circuit board for signal collection. Each gesture was recorded with a fixed time step of 300 seconds (3 seconds, 100 Hz sampling frequency). The dataset was collected by five people wearing the gloves, each performing 26 gestures 10 times, for a total of 10 * 5 * 26 = 1300 data sets. The dataset was partitioned into training, validation, and test sets in a 6:2:2 ratio. The collected data undergoes data preprocessing. First, the raw data for the first 3 seconds (300 sampling points) of each channel is extracted from the data. Then, the average value of the first 100 sampling points in the signal segment on each channel is taken as the baseline. The baseline is then subtracted from the entire 300×6 original signal matrix point by point to achieve static offset correction. The corrected signal is then transposed to a 6×300 channel-time format and collected into a sample list. After processing all samples, the global maximum absolute value is found and each value is normalized to map the signal distribution to the [–1, 1] interval, thereby obtaining a standardized tensor that removes the baseline offset and unifies the amplitude range. After the preprocessed data is input into the designed Inception-BiLSTM model for training, high-precision recognition of the test set data is achieved, as shown in the following results. Figure 7 As shown. Figure 7 It can be seen that the wearable sensing system of the present invention combines inception-BiLSTM to process, recognize and classify the gesture signals of 26 English letters, and finally the accuracy rate in the test set reaches 99.6%. The performance parameters of the sensor prepared in Example 1 of the present application and existing sensors are shown in Table 1.

[0052] Table 1: Comparison of performance indicators between the Au-VG sensor of Example 1 and the existing VG sensor

[0053] Example 2 The difference from Example 1 lies in the optimization of the vertical graphene growth process: the CH4 gas flow rate range is recommended to be precisely controlled within 5-8 sccm; the chamber pressure is recommended to be controlled within 8-10 Pa, the VG sheet height is concentrated within 4-6 μm, the structure is more uniform, and the conductive channel is more stable; the RF power in the range of 1000-1500 W can improve the verticality of the sheet structure.

[0054] Although preferred embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative work shall fall within the scope of protection of the present invention.

Claims

1. A method for preparing a flexible strain sensor based on Au-VG, characterized in that: The process is as follows: (1) Clean and dry the rigid substrate; (2) preparing vertical graphene by chemical vapor deposition on the rigid substrate of step (1); the obtained vertical graphene nanosheets have a height of 4 to 6 μm; (3) Physical vapor deposition of a gold film on the vertical graphene in step (2) with a thickness of 30 to 100 nm to obtain a VG-Au conductive structure; (4) The VG-Au conductive structure is transferred to the solidified lower PDMS substrate by peeling, and the PDMS liquid is covered on the VG-Au conductive structure and solidified to form an upper PDMS layer; (5) Cut the flexible sensing layer into appropriate sizes, form metal electrodes at both ends of the device, and then use a medical breathable film for secondary packaging.

2. The method for preparing the Au-VG flexible strain sensor according to claim 1, characterized in that: In step (2), the process conditions of chemical vapor deposition are as follows: raw material gas: methane (CH4), gas flow rate is 5~10 sccm; chamber pressure is 6~12 Pa; RF power is 1000~2000 W; chamber temperature is 600~900℃; growth time is 5~40 minutes.

3. The method for preparing the Au-VG flexible strain sensor according to claim 1, characterized in that: In step (3), the physical vapor deposition process conditions are as follows: working gas: argon (Ar), gas flow rate is 10~20 sccm; chamber pressure is 0.5~2 Pa; RF power is 100~200 W; chamber temperature is 100±10℃; deposition time is 10~50 seconds.

4. The method for preparing the Au-VG flexible strain sensor according to claim 1, characterized in that: In step (4), the PDMS pre-curing conditions are: heating at 90~110℃ for 5~15 minutes, and the thickness of the upper and lower PDMS layers is 100~500μm.

5. The method for preparing the Au-VG flexible strain sensor according to claim 1, characterized in that: In step (5), silver paste is printed on both ends of the device and thermally cured to form silver paste electrodes. The curing temperature of the silver paste electrode is 60~80℃ and the curing time is 10~30 minutes.

6. The method for preparing the Au-VG flexible strain sensor according to claim 2, characterized in that: The process conditions of the chemical vapor deposition are as follows: raw material gas: methane (CH4), gas flow rate is 5~8 sccm; chamber pressure is 8~10Pa; RF power is 1000~1500 W; chamber temperature is 600~900℃; and growth time is 5~40 minutes.

7. The method for preparing the Au-VG flexible strain sensor according to claim 3, characterized in that: The physical vapor deposition process conditions are as follows: working gas: argon (Ar), gas flow rate is 10~20 sccm; chamber pressure is 0.5~2 Pa; RF power is 120 W; chamber temperature is 100±10℃; deposition time is 20~30 seconds.

8. A flexible strain sensor based on Au-VG, characterized in that: The method is prepared according to any one of claims 1 to 7.

9. An application of an Au-VG flexible strain sensor in the field of wearable devices and / or human-computer interaction, characterized in that: The Au-VG flexible strain sensor according to claim 8 is used.

10. The use according to claim 9, characterized in that The application of the Au-VG flexible strain sensor in gesture recognition and sign language.

Images