Flexible triboelectric sensor for human body vital sign detection, preparation method of flexible triboelectric sensor and intelligent detection method of flexible triboelectric sensor

Abstract

The invention discloses a flexible triboelectric sensor for human body vital sign detection and a preparation method and an intelligent detection method thereof, and belongs to the technical field of flexible triboelectric sensors. The sensor comprises a semitransparent silica gel film layer, a multi-walled carbon nanotube composite abrasive paper modified silica gel layer, a foam separation layer, a conductive fabric electrode layer, a PET substrate layer and a metal shielding layer which are sequentially arranged in an attached mode, the electrode layer is of a fabric structure, and the multi-walled carbon nanotube composite abrasive paper modified silica gel layer is an upper friction layer of the flexible friction electric sensor. The electrode layer is a lower friction layer of the flexible friction electric sensor and is provided with a rectangular through hole, the size of the rectangular through hole is the same as that of the multi-walled carbon nanotube composite abrasive paper modified silica gel layer, and the multi-walled carbon nanotube composite abrasive paper modified silica gel layer is located on the inner side of the rectangular through hole. The sensor is provided with the flexible electrode layer and the high-performance friction layer, the precision, the sensitivity and the response speed can be improved, and the posture, breath, gesture and pulse of a user can be recognized.

Description

Technical Field

[0001] This application relates to the technical field of flexible triboelectric sensors, and particularly to a flexible triboelectric sensor for detecting human vital signs, its preparation method, and an intelligent detection method. Background Art

[0002] Wearable flexible electronic devices have made remarkable progress in recent years in fields such as motion monitoring and human-computer interaction due to their unique advantages in flexibility and portability. Such devices can be directly attached to the skin surface or fixed to different parts of the body through force rings or straps, etc., so as to achieve real-time collection of various body information. This information can not only be used to judge an individual's motion state but also effectively evaluate their health status. The core component of wearable flexible electronic devices is a flexible sensor, among which flexible triboelectric sensors are particularly important. This kind of sensor can directly convert many mechanical signals generated by the human body into corresponding electrical pulse signals, thus having unique advantages in the extraction of physiological information. By analyzing the intensity of the electrical signal and its characteristic peaks, it can deeply explore the human body posture and physiological information contained in the signal, so it has become one of the indispensable key components in wearable flexible electronic devices. With the continuous progress of technology, the application scope of these devices is also constantly expanding, and it is expected to play a more important role in multiple fields such as medical treatment, sports, and health monitoring in the future.

[0003] CN104779832A discloses a method of using a fluorocarbon plasma treatment process to form micro-nano structures on the polymer surface to increase the roughness of the friction material, thereby improving the electrical output performance of the triboelectric sensor. CN104167949A discloses a heating and embossing treatment technology to obtain a polymer thin film layer with an uneven structure.

[0004] However, the sensing performance of the friction layer of the above sensors is not high, resulting in low accuracy, poor sensitivity, and slow response speed of the flexible triboelectric sensor. Moreover, the above technical solutions lack the processing and analysis of output data, which is not conducive to their application in the fields of human body posture perception, respiration, and pulse monitoring. Summary of the Invention

[0005] In view of this, this application provides a flexible triboelectric sensor for detecting human vital signs, its preparation method, and an intelligent detection method, which have a flexible electrode layer and a high-performance friction layer, can improve the accuracy, sensitivity, and response speed of the flexible triboelectric sensor, and can analyze the data collected by the flexible triboelectric sensor to identify the user's posture, respiration, gesture, and pulse.

[0006] Specifically, it includes the following technical solutions: In a first aspect, the present application provides a flexible triboelectric sensor for detecting human vital signs. The flexible triboelectric sensor includes a semi-transparent silicone film layer, a multi-walled carbon nanotube composite sandpaper modified silicone layer, a foam spacer layer, a conductive fabric electrode layer, a PET substrate layer, and a metal shielding layer that are sequentially laminated. The electrode layer has a fabric structure. The multi-walled carbon nanotube composite sandpaper modified silicone layer is the upper friction layer of the flexible triboelectric sensor, and the electrode layer is the lower friction layer of the flexible triboelectric sensor. The foam spacer layer and the electrode layer are on the same plane and have rectangular through-holes. The size of the rectangular through-holes is the same as the size of the multi-walled carbon nanotube composite sandpaper modified silicone layer. The multi-walled carbon nanotube composite sandpaper modified silicone layer is located inside the rectangular through-holes.

[0007] In some embodiments, the length of the multi-walled carbon nanotube composite sandpaper modified silicone layer is 1 - 3 cm, the width is 1 - 1.5 cm, and the thickness is 60 - 120 μm.

[0008] In some embodiments, the material of the conductive fabric electrode layer is a fabric soaked in any one of Au, Ag, and Cu conductive metal slurries. The fabric is any one of polyester, cotton, and cotton - spandex blend. The length of the conductive fabric electrode layer is 2 - 6 cm, the width is 1 - 2 cm, and the thickness is 80 - 130 μm.

[0009] In some embodiments, the lengths of the semi-transparent silicone film layer, the PET substrate layer, and the metal shielding layer are all 2 - 6 cm, and the widths are all 1 - 2 cm; The thickness of the semi-transparent silicone film layer is 60 - 120 μm; The length of the foam spacer layer is 2 - 6 cm, the width is 1 - 2 cm, and the length of the rectangular through-hole is 1 - 3 cm, and the width is 1 - 1.5 cm; The thickness of the PET substrate layer is 180 μm - 220 μm; The thickness of the metal shielding layer is 80 μm - 120 μm.

[0010] In some embodiments, the flexible triboelectric sensor further includes a PVC waterproof layer. The PVC waterproof layer is laminated on the outside of the metal shielding layer, and its size completely follows that of the metal shielding layer, generally with a length of 2 - 6 cm and a width of 1 - 2 cm.

[0011] In some embodiments, the centers of the semi-transparent silicone film layer, the multi-walled carbon nanotube composite sandpaper modified silicone layer, the electrode layer, the PET substrate layer, the metal shielding layer, and the PVC waterproof layer are all on the same straight line.

[0012] In some embodiments, the flexible triboelectric sensor has a response time of 42 ms.

[0013] In a second aspect, the present application provides a method for preparing a flexible triboelectric sensor for detecting human vital signs, the method comprising the following steps: Step 1, drop silica gel with a hardness of 0 degrees on the center of the spin disk of the spin coater, and perform spin coating. The rotation speed of the spin coater is set to three revolutions per second, and the spin coating is performed for 30 to 40 seconds. After drying for 3 to 3.5 hours, the semi-transparent silica gel film layer is cut to obtain the semi-transparent silica gel film layer; Step 2, placing the fabric electrode substrate under the screen of a screen printer, and applying a conductive metal slurry on the screen with a drop amount of 1 mL each time, and then performing the next drop application after the conductive metal slurry fully penetrates into the fiber gaps of the fabric electrode substrate each time, and the drop application is repeated 3 to 6 times to obtain an electrode layer; Step 3, adding 0.2g~0.4g of multi-walled carbon nanotubes into 20g~40g of silica gel with a hardness of 20 degrees and stirring to obtain a silica gel solution mixed with the multi-walled carbon nanotubes; then sticking one side of sandpaper with a particle size of 1200~3600 meshes to the center of the upper surface of the rotating disk of the spin coater, dripping 3~4mL of the silica gel solution mixed with the multi-walled carbon nanotubes on the upper surface of the sandpaper, and then setting the rotation speed of the spin coater to 3~5 turns per second, and spinning for 30~40 seconds; finally, removing the rotating disk, drying for 4~4.5 hours, and then cutting to obtain a multi-walled carbon nanotube composite sandpaper modified silica gel layer; Step 4, sequentially laminate the translucent silicone film layer, the multi-walled carbon nanotube composite sandpaper-modified silicone layer, the foam partition layer, the conductive fabric electrode layer, the PET base layer, the metal shielding layer and the PVC waterproof layer, and connect one side of the conductive fabric electrode layer and one end of the DuPont line with conductive silver glue to obtain a flexible triboelectric sensor.

[0014] In a third aspect, the present application provides an intelligent detection method. The intelligent detection method is applied to an intelligent detection system including a flexible triboelectric sensor for detecting human vital signs as described in the first aspect, wherein the intelligent detection system also includes a signal analysis module. The intelligent detection method includes: The flexible triboelectric sensor for detecting human vital signs collects vital sign signals including the user's main body joint movements, five finger joint movements, breathing movements and pulse heart rate, and sends the vital sign signals to the signal analysis module; The signal analysis module receives the vital sign signal and determines the user's posture, gesture, breathing pattern and pulse rate based on the vital sign signal.

[0015] In some embodiments, determining the user's posture, gesture, breathing pattern, and pulse rate based on the vital sign signals includes: Inputting the preprocessed vital sign signals into an identification network, and outputting an identification result including the user's posture, gesture, breathing pattern, and pulse rate. The identification network has an MLP-Mixer architecture, including a one-dimensional convolutional layer, a normalization layer, a mixed spatial information layer, a mixed position feature layer, a global average pooling layer, and a fully connected layer.

[0016] The beneficial effects of the technical solution provided by this application at least include: This application provides a flexible triboelectric sensor for human vital sign detection, its preparation method, and an intelligent detection method. The flexible triboelectric sensor includes a semi-transparent silicone film layer, a multi-walled carbon nanotube composite sandpaper modified silicone layer, a foam spacer layer, a conductive fabric electrode layer, a PET substrate layer, and a metal shielding layer that are sequentially laminated. The electrode layer has a fabric structure. The multi-walled carbon nanotube composite sandpaper modified silicone layer is the upper friction layer of the flexible triboelectric sensor, and the electrode layer is the lower friction layer of the flexible triboelectric sensor and also the output electrode of the flexible triboelectric sensor. The foam spacer layer and the multi-walled carbon nanotube composite sandpaper modified silicone layer are on the same plane and have rectangular through-holes. The size of the rectangular through-holes is the same as the size of the multi-walled carbon nanotube composite sandpaper modified silicone layer. The multi-walled carbon nanotube composite sandpaper modified silicone layer is located inside the rectangular through-holes. By setting an electrode structure with fabric micro-protrusions and a high-performance friction layer, the accuracy, sensitivity, and response speed of the flexible triboelectric sensor can be improved. The flexible triboelectric sensor has the characteristics of self-power supply, non-invasive monitoring, and high stability, and can identify the user's posture, breathing, gesture, and pulse. Description of the Drawings

[0017] To more clearly illustrate the technical solutions in the embodiments of this application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of this application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0018] Figure 1 It is a schematic structural diagram of the flexible triboelectric sensor for human vital sign detection provided by the embodiment of this application; Figure 2(a) is a schematic diagram of the electrical characteristics of the flexible triboelectric sensor for human vital sign detection provided by the embodiment of this application; Figure 2(b) is a schematic diagram of the wearing position of the flexible triboelectric sensor for human vital sign detection provided by the embodiment of this application when detecting breathing behavior; Figure 2(c) is a schematic diagram of the response and recovery time of the flexible triboelectric sensor for human vital sign detection provided by the embodiment of the present application; Figure 2(d) is a schematic diagram of the output voltage distribution range of the sensor corresponding to nine respiratory behaviors detected by the flexible triboelectric sensor for human vital sign detection provided by the embodiment of the present application during the respiratory behavior test; Figure 2(e) is a schematic diagram of the sensor output signal waveform corresponding to nine respiratory behaviors detected by the flexible triboelectric sensor for human vital sign detection provided by the embodiment of the present application during the respiratory behavior test; Figure 2(f) is a schematic diagram of the specific joint positions to be detected by the flexible triboelectric sensor for human vital sign detection provided by the embodiment of the present application during the detection of human major joint movements; Figure 2(g) is a schematic diagram of the gestures and the corresponding output waveforms of the sensor when the flexible triboelectric sensor for human vital sign detection provided by the embodiment of the present application performs different gesture number tests; Figure 3(a) is a schematic diagram of the 10s radial artery pulse signal detected by the flexible triboelectric sensor for human vital sign detection provided by the embodiment of the present application during the pulse test; Figure 3(b) is a schematic diagram of integrating 8 sensors into an insole-type sensor array by the flexible triboelectric sensor for human vital sign detection provided by the embodiment of the present application; Figure 3(c) is a schematic diagram of the pressure distribution felt by the sensors of the insole-type sensing array composed of the flexible triboelectric sensors provided by the embodiment of the present application when the subject is walking normally; Figure 4 It is a flowchart of a preparation method of a triboelectric sensor provided by the embodiment of the present application; Figure 5(a) is a schematic diagram of the structure of the intelligent detection method provided by the embodiment of the present application; Figure 5(b) is a confusion matrix of the intelligent detection method provided by the embodiment of the present application on the training set when classifying 9 respiratory states; Figure 5(c) is a confusion matrix of the intelligent detection method provided by the embodiment of the present application on the test set when classifying 9 respiratory states; Figure 5(d) is a precision change curve of the intelligent detection method provided by the embodiment of the present application when classifying and training 9 respiratory states; Figure 5(e) is a loss function change curve of the intelligent detection method provided by the embodiment of the present application when classifying and training 9 respiratory states; Figure 6(a) is a confusion matrix of the intelligent detection method provided by the embodiment of the present application on the training set when classifying 11 joint movement features; Figure 6 (b) is the confusion matrix of the intelligent detection method provided by the embodiment of the present application when classifying 11 joint motion features on the test set; Figure 6 (c) is the accuracy change curve of the intelligent detection method provided by the embodiment of the present application when training the classification of 11 joint motion features; Figure 6 (d) is the loss function change curve of the intelligent detection method provided by the embodiment of the present application when training the classification of 11 joint motion features; Figure 7 (a) is the confusion matrix of the intelligent detection method provided by the embodiment of the present application when classifying ten gesture digits on the training set; Figure 7 (b) is the confusion matrix of the intelligent detection method provided by the embodiment of the present application when classifying ten gesture digits on the test set; Figure 7 (c) is the accuracy change curve of the intelligent detection method provided by the embodiment of the present application when training the classification of ten gesture digits; Figure 7 (d) is the loss function change curve of the intelligent detection method provided by the embodiment of the present application when training the classification of ten gesture digits; Figure 8 (a) is the schematic diagram of the prediction accuracy of the intelligent detection method provided by the embodiment of the present application for the heart rate HR; Figure 8 (b) is the schematic diagram of the prediction accuracy of the intelligent detection method provided by the embodiment of the present application for the time difference PPT between the first characteristic peak and the third characteristic peak of the pulse wave; Figure 8 (c) is the schematic diagram of the prediction accuracy of the intelligent detection method provided by the embodiment of the present application for the pulse reflection index RI; Figure 8 (d) is the schematic diagram of the prediction accuracy of the intelligent detection method provided by the embodiment of the present application for the upstroke time UT of the cardiac cycle; Figure 8 (e) is the loss function curve of the intelligent detection method provided by the embodiment of the present application when training the prediction of four characteristic physiological signals of the pulse; Figure 9 (a) is the confusion matrix of the intelligent detection method provided by the embodiment of the present application when identifying eight walking postures of people on the training set; Figure 9 (b) is the confusion matrix of the intelligent detection method provided by the embodiment of the present application when identifying eight walking postures of people on the test set; Figure 9 (c) is the accuracy change curve of the intelligent detection method provided by the embodiment of the present application when training the identification of eight walking postures of people; Figure 9 (d) is the loss function change curve of the intelligent detection method provided by the embodiment of the present application when training the identification of eight walking postures of people.

[0019] The reference numerals in the figures are respectively represented as: 1 - Semi - transparent silicone film layer, 2 - Multi - walled carbon nanotube composite sandpaper modified silicone layer, 3 - Foam spacer layer, 4 - Conductive fabric electrode layer, 5 - PET base layer, 6 - Metal shielding layer, 7 - PVC waterproof layer. Specific embodiments

[0020] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0021] To make the technical solutions and advantages of the present application clearer, the embodiments of the present application will be further described in detail below in conjunction with the accompanying drawings.

[0022] As Figure 1 shown, in the first aspect of the present application, a flexible triboelectric sensor for detecting human vital signs is provided. The flexible triboelectric sensor includes a semi - transparent silicone film layer 1, a multi - walled carbon nanotube composite sandpaper modified silicone layer 2, a foam spacer layer 3, a conductive fabric electrode layer 4, a PET base layer 5, and a metal shielding layer 6 that are sequentially and conformally arranged. The electrode layer 4 has a fabric structure. The multi - walled carbon nanotube composite sandpaper modified silicone layer 2 is the upper friction layer of the flexible triboelectric sensor, and the electrode layer 4 is the lower friction layer of the flexible triboelectric sensor and also the output electrode of the flexible triboelectric sensor. The foam spacer layer 3 and the multi - walled carbon nanotube composite sandpaper modified silicone layer are on the same plane and have rectangular through - holes. The size of the rectangular through - holes is the same as the size of the multi - walled carbon nanotube composite sandpaper modified silicone layer. The multi - walled carbon nanotube composite sandpaper modified silicone layer is located inside the rectangular through - holes.

[0023] The semi - transparent silicone film layer 1 has the characteristics of being flexible and skin - friendly, facilitating the direct close contact between the flexible friction sensor and the user's skin through the semi - transparent silicone film, enhancing the detection accuracy, and improving the user experience.

[0024] The foam spacer layer 3 surrounding the outside of the multi - walled carbon nanotube composite sandpaper modified silicone layer plays a protective role for the electrode layer 4 and makes the multi - walled carbon nanotube composite sandpaper modified silicone layer 2 and the electrode layer 4 closely fit, thus better demonstrating the friction effect.

[0025] In some embodiments, the multi - walled carbon nanotube composite sandpaper modified silicone layer 2 can be prepared by physical doping, magnetic stirring, and spin - coating methods. The length of the multi - walled carbon nanotube composite sandpaper modified silicone layer 2 is 1 - 3 cm, the width is 1 - 1.5 cm, and the thickness is 60 - 120 μm.

[0026] The multi-walled carbon nanotube composite sandpaper modified silica gel layer 2 located below the semi-transparent silica gel film layer 1 is prepared by physical doping, magnetic stirring and spin coating methods, and has the characteristics of ultra-thin, homogeneous and dense. Compared with the pure silica gel film, the multi-walled carbon nanotube composite sandpaper modified silica gel layer 2 has a stronger triboelectric effect.

[0027] The PET base layer 5 has high transparency and excellent mechanical properties, which is beneficial to the triboelectric performance of the flexible triboelectric sensor.

[0028] An electrode layer 4 with a fabric structure is arranged on the PET base layer 5, which serves as the lower friction layer of the flexible triboelectric sensor and also acts as the output electrode of the flexible triboelectric sensor. The unique surface micro-convex structure of the fabric is used to improve the triboelectric performance, which is convenient for subsequent monitoring of various joint movements and pulse conditions of users.

[0029] By setting the metal shielding layer 6, the flexible triboelectric sensor can be resistant to electromagnetic interference and ensure the stable operation of the flexible triboelectric sensor.

[0030] In some embodiments, the material of the electrode layer 4 is a fabric soaked in any one of the conductive metal slurries of Au, Ag and Cu. The fabric is any one of polyester, cotton and cotton-spandex blend. The length of the conductive fabric electrode layer 4 is 2 - 6 cm, the width is 1 - 2 cm, and the thickness is 80 - 130 μm. It should be noted that the fabric soaked in the conductive metal slurries of Au, Ag and Cu as the material of the electrode layer 4 has the best conductivity and can reduce the loss of weak signals during the sensing process.

[0031] In some embodiments, the lengths of the semi-transparent silica gel film layer 1, the PET base layer 5 and the metal shielding layer 6 are all 2 - 6 cm, and the widths are 1 - 2 cm; the thickness of the semi-transparent silica gel film layer 1 is 60 - 120 μm. The length of the foam spacer layer 3 is 2 - 6 cm, the width is 1 - 2 cm, and the length of the rectangular through-hole is 1 - 3 cm, the width is 1 - 1.5 cm; the thickness of the PET base layer 5 is 180 μm - 220 μm; the thickness of the metal shielding layer 6 is 80 μm - 120 μm.

[0032] In some embodiments, the flexible triboelectric sensor further includes a PVC waterproof layer 7. The PVC waterproof layer 7 is adhesively disposed between the PET base layer 5 and the metal shielding layer 6, and the size of the PVC waterproof layer 7 is the same as that of the metal shielding layer 6. The length of the PVC waterproof layer 7 is 2 - 6 cm, the width is 1 - 2 cm, and the thickness is 130 μm - 150 μm.

[0033] By setting the PVC waterproof layer 7, the flexible triboelectric sensor has a waterproof function, and when the user is exposed to humid, rainy and other environments, the flexible triboelectric sensor can also perform high-performance detection.

[0034] In some embodiments, the centers of the translucent silicone film layer 1, the multi-walled carbon nanotube composite sandpaper modified silicone layer 2, the electrode layer 4, the PET substrate layer 5, the metal shielding layer 6, and the PVC waterproof layer 7 are all on the same straight line.

[0035] In some embodiments, the fastest response time of the flexible triboelectric sensor is 42 ms.

[0036] With such a setting, the structure of the flexible triboelectric sensor is small and compact. By utilizing the friction between the electrode layer 4 with a fabric structure and the multi-walled carbon nanotube composite sandpaper modified silicone layer 2, and through the cooperation of the various layers of the flexible triboelectric sensor, the flexible triboelectric sensor has high precision, high sensitivity, and a fast response speed, and has the characteristics of self-power supply, skin-friendly, waterproof, and anti-wear.

[0037] During the test, the flexible triboelectric sensor is divided into two parts. The translucent silicone film layer 1, the multi-walled carbon nanotube composite sandpaper modified silicone layer 2, and the foam spacer layer 3 are used as the first part, and the electrode layer 4, the PET substrate layer 5, and the metal shielding layer 6 are used as the second part. The first part is slapped on the second part at a certain frequency and pressure, so that the multi-walled carbon nanotube composite sandpaper modified silicone layer 2 as the upper friction layer and the electrode layer 4 as the lower friction layer generate electricity by friction. Fig. 2(a) is a schematic diagram of the electrical characteristics of the flexible triboelectric sensor for detecting human vital signs provided by the embodiment of the present application. Taking the material of the electrode layer 4 as Ag and cotton cloth as an example, the maximum output voltage between the electrode layer and the multi-walled carbon nanotube composite sandpaper modified silicone layer 2 can be about 175 V, and the current is about 27 μA. As shown in Fig. 2(b), this figure shows the wearing method and wearing position of the sensor during the breathing behavior test. As shown in Fig. 2(c), in the above situation, the response and recovery times of the flexible triboelectric sensor are 42 ms and 97 ms respectively. As can be seen from Fig. 2(d), in the nine breathing modes of deep fast (DF), deep moderate (DM), deep slow (DS), moderate fast (MF), moderate moderate (MM), moderate slow (MS), shallow fast (SF), shallow moderate (SM), and shallow slow (SS), the breathing intensities of deep, moderate, and shallow are significantly different. Fig. 2(e) gives the output waveforms of the flexible triboelectric sensor in the aforementioned nine breathing modes. Fig. 2(f) gives the specific parts to be detected when the flexible triboelectric sensor detects the movements of the main joints of the human body. Fig. 2(g) shows the typical signal waveforms that the flexible triboelectric sensor can output when it is used for gesture digit detection.

[0038] In some embodiments, the flexible triboelectric sensor can be used to monitor the user's pulse signal. As shown in Fig. 3(a), the flexible triboelectric sensor monitors the radial artery pulse wave of a 31-year-old tester for 10 s, and it can be seen that the flexible triboelectric sensor can capture the tester's pulse beats.

[0039] In some embodiments, after being integrated into an insole sensor array, the flexible triboelectric sensor can detect a person's walking state. As shown in Fig. 3(b), this is a cross-sectional view of the sensor array, and a total of 8 flexible triboelectric sensors are integrated inside. Fig. 3(c) shows the pressure distribution felt by the insole sensor array of a 31-year-old tester in a normal walking state. It can be seen that the flexible triboelectric sensors at different positions can clearly distinguish the distribution of pressure magnitudes.

[0040] In summary, the flexible triboelectric sensor for human vital sign detection provided in the embodiments of the present application has a flexible electrode layer and a high-performance friction layer, which can improve the accuracy, sensitivity, and response speed of the flexible triboelectric sensor, and can identify the user's posture, breathing, gestures, and pulse.

[0041] In a second aspect, as Figure 4 shown, the embodiments of the present application provide a preparation method for the flexible triboelectric sensor of the first aspect of the present application. The method includes the following steps: Step 1: Drop silicone rubber with a hardness of 0 degrees at the center of the spinning disk of a spin coater, and perform spin coating. The rotation speed of the spin coater is set at three revolutions per second, and spin coating is performed for 30 - 40 s. After drying for 3 - 3.5 h, it is cut to obtain a semi-transparent silicone rubber thin film layer 1.

[0042] The rotation speed of the spin coater has a decisive influence on the final thickness of the silicone rubber thin film. The semi-transparent silicone rubber thin film layer 1 obtained in Step 1 has excellent mechanical properties, which is beneficial to the triboelectric performance of the flexible triboelectric sensor.

[0043] Step 2: Place the fabric electrode substrate under the screen of a screen printer, and drop the conductive metal slurry on the upper part of the screen at a dropwise coating amount of 1 mL each time. After each conductive metal slurry fully penetrates into the fiber gaps of the fabric electrode substrate, the next dropwise coating is carried out. The number of dropwise coating repetitions is 3 - 6 times to obtain an electrode layer 4.

[0044] In some embodiments, choosing cotton cloth as the base fabric of the fabric electrode can obtain better triboelectric performance.

[0045] In some embodiments, the conductive metal slurry in (2) can be Ag nanoparticle ink.

[0046] Step 3: Incorporate 0.2 g to 0.4 g of multi-walled carbon nanotubes into 20 g to 40 g of silica gel with a hardness of 20 degrees and stir to obtain a silica gel solution mixed with multi-walled carbon nanotubes. Then, stick one side of sandpaper with a particle size number of 1200 to 3600 meshes to the center of the upper surface of the spinning disk of the spin coater. Drop 3 to 4 mL of the silica gel solution mixed with multi-walled carbon nanotubes onto the upper surface of the sandpaper. After that, set the rotation speed of the spin coater to 3 to 5 revolutions per second and spin coat for 30 to 40 s. Finally, remove the spinning disk, dry for 4 to 4.5 h, and then cut it to obtain the multi-walled carbon nanotube composite sandpaper modified silica gel layer 2.

[0047] With such settings, the relevant parameters for preparing the mixed silica gel can make the obtained multi-walled carbon nanotube composite sandpaper modified silica gel layer 2 homogeneous and dense, and it has a more significant triboelectric effect compared to a pure silica gel film without surface modification.

[0048] Step 4: Bond the semi-transparent silica gel film layer 1, the multi-walled carbon nanotube composite sandpaper modified silica gel layer 2, the foam spacer layer 3, the conductive fabric electrode layer 4, the PET substrate layer 5, the metal shielding layer 6, and the PVC waterproof layer 7 in sequence. Connect one side of the conductive fabric electrode layer 4 and one end of a Dupont wire with conductive silver glue to obtain a flexible triboelectric sensor.

[0049] In some embodiments, the preparation method may specifically include: (1) Drop 0-degree silica gel at the center of the spinning disk of the spin coater. Set the rotation speed of the spin coater to three revolutions per second, spin coat for 30 s, dry for 3 h, and then cut it to obtain the semi-transparent silica gel film layer 1.

[0050] (2) Place the fabric electrode substrate under the screen of the screen printing machine. Drop the conductive metal slurry on the upper side of the screen at a dropping amount of 1 mL each time. Wait until each time the conductive metal slurry fully penetrates into the fiber gaps of the fabric electrode substrate before the next dropping. The dropping is repeated 3 times to obtain the electrode layer 4.

[0051] In some embodiments, the specific implementation of (2) can be: Use metal Ag nanoparticle ink with a surface tension of 26 to 29 mN / m and a screen printing machine to fabricate the conductive layer on a cotton cloth with a thickness of 70 to 120 μm. After completion, place the above conductive layer in a 60°C vacuum drying oven for 1 h to obtain the final electrode layer 4.

[0052] In some embodiments, for the convenience of detection or connection with other devices and instruments, connect a Dupont wire to one side of the electrode layer 4 with conductive silver glue to lead out the detection end.

[0053] (3) Incorporate 0.2 g of multi-walled carbon nanotubes into 20 g of silica gel with a hardness of 20 degrees and stir to obtain a silica gel solution mixed with multi-walled carbon nanotubes. Then, stick one side of a sandpaper with a particle size of 1200 mesh to the center of the upper surface of the turntable of a spin coater, drop 3 mL of the silica gel solution mixed with multi-walled carbon nanotubes on the upper surface of the sandpaper, and then set the rotation speed of the spin coater at three revolutions per second and spin coat for 30 seconds. Finally, remove the turntable and place it in a ventilated and dry place to stand for 4 hours to obtain a multi-walled carbon nanotube composite sandpaper modified silica gel layer 2.

[0054] (4) Laminate a semi-transparent silica gel film layer, a multi-walled carbon nanotube composite sandpaper modified silica gel layer, a foam spacer layer, a conductive fabric electrode layer, a PET substrate layer, a metal shielding layer, and a PVC waterproof layer in sequence, and connect one side of the conductive fabric electrode layer and one end of a Dupont wire with conductive silver glue.

[0055] In the above preparation method, the specific implementation manner of (4) can be: use a foam spacer layer 3 to isolate the periphery of the electrode layer 4; coat a layer of waterproof glue on the upper surface of the foam spacer layer 3; first stick the side of the multi-walled carbon nanotube composite sandpaper modified silica gel layer 2 that is not modified by sandpaper to the center of the semi-transparent silica gel film layer 1, and then laminate the semi-transparent silica gel film with the foam spacer layer 3; stick the electrode layer 4 to the center of the PET substrate layer 5, and then laminate the metal shielding layer 6 outside the PET substrate; finally, stick a PVC waterproof layer 7 on the lower surface of the metal shielding layer 6, and connect one side of the conductive fabric electrode layer 4 and one end of a Dupont wire with conductive silver glue to obtain a flexible triboelectric sensor.

[0056] In some embodiments, the metal shielding layer 6 can be brass foil, aluminum foil, or tin foil.

[0057] In summary, the preparation method of the flexible triboelectric sensor for human vital sign detection provided by the embodiments of the present application can endow the flexible triboelectric sensor with a flexible electrode layer and a high-performance friction layer, which can improve the accuracy, sensitivity, and response speed of the flexible triboelectric sensor, so that the flexible triboelectric sensor can identify the user's posture, breathing, gesture, and pulse.

[0058] Thirdly, in order to analyze and extract the user's motion posture, breathing pattern, and pulse physiological information, the embodiments of the present application provide an intelligent detection method. As Figure 4 shown, this method is applied to an intelligent detection system including a flexible triboelectric sensor for human vital sign detection as described in the first aspect, and the intelligent detection system further includes a signal analysis module; The flexible triboelectric sensor for human vital sign detection collects vital sign signals including the user's main body joint movements, five-finger joint movements, breathing movements, and pulse heart rate, and sends the vital sign signals to the signal analysis module; The signal analysis module receives the vital sign signals and determines the user's posture, gesture, breathing pattern, and pulse rate based on the vital sign signals.

[0059] In some embodiments, determining the user's posture, gesture, breathing pattern, and pulse rate based on the vital sign signals includes: inputting the preprocessed vital sign signals into an identification network, and outputting an identification result including the user's posture, gesture, breathing pattern, and pulse rate. Referring to FIG. 5(a), the identification network has an MLP-Mixer architecture, including a one-dimensional convolutional layer, a normalization layer, a mixed spatial information layer, a mixed position feature layer, a global average pooling layer, and a fully connected layer.

[0060] It should be noted that the identification network can perform classification tasks for the user's breathing pattern, posture, and gesture, and can perform an identification task for the pulse rate, so as to obtain an identification result including the user's posture, gesture, breathing pattern, and pulse rate.

[0061] In the task of classifying nine breathing patterns (i.e., the classification task of the user's breathing pattern), each breathing pattern has 200 sets of complete data, and one set of data represents a complete breathing action. Therefore, there are a total of 1800 sets of data. When training the identification network, all the data is divided into a training set and a test set in a ratio of 7:3. After 50 training rounds, the confusion matrix of the identification network on the training set is shown in FIG. 5(b), and the accuracy reaches 100%. The confusion matrix on the test set is shown in FIG. 5(c), and the accuracy is 99.26%. FIGS. 5(d) and 5(e) show the accuracy change curve and the loss function change curve during the classification training for nine breathing states. It can be seen from the figure that after 25 training rounds, the accuracy of the identification network on the training set and the test set has reached a very high level. Similarly, the loss function has also reached a very low level. It can be seen that the specificity of this identification network for this type of classification task is very high. Subsequently, the classification tasks and regression prediction tasks for the user's posture, gesture, and pulse rate are appropriately modified based on this network.

[0062] In the task of classifying 11 joint motion features (i.e., the classification task of the user's posture), 200 sets of data were collected for the motion of each joint. One set of data represents a complete motion cycle of this joint. Therefore, there are a total of 2200 sets of data. After appropriately modifying the input channel dimension of the aforementioned recognition network capable of classifying breathing patterns, it can be used to classify 11 joint motion features. All data is still divided into a training set and a test set in a 7:3 ratio. After 50 training rounds, the confusion matrix of the recognition network on the training set is shown in Figure 6(a), and the accuracy reaches 95.7%. The confusion matrix on the test set is shown in Figure 6(b), and the accuracy is 97.69%. Figures 6(c) and 6(d) show the accuracy change curve and loss function change curve of the recognition network during the classification training of 11 joint motion features. It can be seen from the figure that after 20 rounds of training this time, the accuracy of the recognition network on the training set and the test set has reached a very high level. Similarly, the loss function has also reached a very low level.

[0063] In the task of classifying 10 gesture digits (i.e., the classification task of the user's gestures), 200 sets of data were collected for the finger motions corresponding to each gesture digit. One set of data represents a complete motion cycle of this gesture digit. Therefore, there are a total of 2000 sets of data. After appropriately modifying the input channel dimension of the aforementioned recognition network, it can be used to classify 10 gesture digits. All data is still divided into a training set and a test set in a 7:3 ratio. After 50 training rounds, the confusion matrix of the recognition network on the training set is shown in Figure 7(a), and the accuracy reaches 100%. The confusion matrix on the test set is shown in Figure 7(b), and the accuracy is 99.83%. Figures 7(c) and 7(d) show the accuracy change curve and loss function change curve of the recognition network during the classification training of 10 gesture digits. It can be seen from the figure that after 8 rounds of training this time, the accuracy of the network on the training set and the test set has reached a very high level. Similarly, the loss function has also reached a very low level.

[0064] In the task of extracting physiological information from pulse signals (i.e., the recognition task of the user's pulse rate), a total of 220 sets of pulse data from three people were collected. Each set of data is a complete pulse cycle. After revising the aforementioned recognition network into a prediction regression type, all data is divided into a training set and a test set in a 7:3 ratio. Finally, the network is for heart rate HR, forward wave (P 1 ), and dicrotic wave (P 3Predictive analysis was performed on the time difference PPT, reflection index RI, and systolic upslope time UT between (). Figures 8(a) to 8(d) show the prediction accuracy of the recognition network for HR, PPT, RI, and UT. All data points are closely distributed around the line X = Y. MAE represents the mean absolute error, and RMSE represents the root mean square error. Except that the error of HR is greater than 1, the errors of the other three parameters are all less than 0.2. It can be seen that the recognition network is very effective in deeply mining cardiovascular information. Figure 8(e) shows the loss function curve of the recognition network during training when predicting four characteristic physiological signals of the pulse. It can be seen from the figure that after 5 rounds of training, the loss functions of the recognition network on the training set and the test set reach a very low level.

[0065] When classifying eight walking states of a person (i.e., the classification task of the user's posture), 200 groups of data were collected for each walking state. One group of data represents a complete action cycle of this state. Therefore, there are a total of 1600 groups of data. After appropriately modifying the input channel dimension of the aforementioned recognition network, the classification of eight walking states can be performed. All data is still divided into a training set and a test set in a ratio of 7:3. After 50 training rounds, the confusion matrix of the recognition network on the training set is shown in Figure 9(a), and the accuracy reaches 97.51%. The confusion matrix on the test set is shown in Figure 9(b), and the accuracy is 93.44%. Figures 9(c) and 9(d) show the accuracy change curve and the loss function change curve of the recognition network during the classification training of eight walking states. It can be seen from the figure that after 20 rounds of training, the accuracy of the recognition network on the training set and the test set has reached a very high level. Similarly, the loss function has also reached a very low level.

[0066] In summary, it can be known that the intelligent detection method provided by the embodiments of the present application enables the flexible triboelectric sensor to recognize the user's posture, breathing, gestures, and pulse. The flexible triboelectric sensor has the characteristics of self-power supply, non-invasive monitoring, and high stability. Cooperating with the recognition network can improve the recognition accuracy.

[0067] In the present application, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance. The term "plurality" means two or more, unless otherwise clearly defined.

[0068] Those skilled in the art will readily think of other embodiments of the present application after considering the specification and practicing the present application disclosed herein. The present application is intended to cover any variations, uses, or adaptations of the present application, which follow the general principles of the present application and include common general knowledge or conventional technical means in the technical field not disclosed in the present application. The specification and examples are only considered exemplary.

[0069] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present application is only limited by the appended claims.

Claims

1. Flexible triboelectric sensor for detecting human vital signs, characterized in that: The flexible triboelectric sensor comprises a translucent silicone film layer (1), a multi-walled carbon nanotube composite sandpaper modified silicone layer (2), a foam barrier layer (3), a conductive fabric electrode layer (4), a PET substrate layer (5) and a metal shielding layer (6) which are sequentially laminated and arranged, wherein the conductive fabric electrode layer (4) has a cross-vertical and cross-horizontal structure of a fabric, the multi-walled carbon nanotube composite sandpaper modified silicone layer (2) is an upper friction layer of the flexible triboelectric sensor, the conductive fabric electrode layer (4) is a lower friction layer of the flexible triboelectric sensor, the foam barrier layer (3) and the multi-walled carbon nanotube composite sandpaper modified silicone layer (2) are located on the same plane, and have a rectangular through hole, the size of the rectangular through hole is the same as the size of the multi-walled carbon nanotube composite sandpaper modified silicone layer (2), and the multi-walled carbon nanotube composite sandpaper modified silicone layer (2) is located inside the rectangular through hole.

2. The flexible triboelectric sensor for detecting human vital signs according to claim 1 is characterized in that: The multi-walled carbon nanotube composite sandpaper modified silica gel layer (2) has a length of 1 to 3 cm, a width of 1 to 1.5 cm, and a thickness of 60 to 120 μm.

3. The flexible triboelectric sensor for detecting human vital signs according to claim 1, characterized in that: The conductive fabric electrode layer (4) is made of a fabric soaked in any one of Au, Ag and Cu conductive metal slurries, the fabric is any one of polyester, cotton and a cotton-spandex blend, and the conductive fabric electrode layer (4) is 2 to 6 cm in length, 1 to 2 cm in width and 80 to 130 μm in thickness.

4. The flexible triboelectric sensor for detecting human vital signs according to claim 1, characterized in that: The length of the translucent silicone film layer (1), the PET base layer (5) and the metal shielding layer (6) are all 2 to 6 cm, and the width is 1 to 2 cm; The thickness of the translucent silicone film layer (1) is 60-120 μm; The foam insulation layer (3) has a length of 2 to 6 cm and a width of 1 to 2 cm, and the rectangular through hole has a length of 1 to 3 cm and a width of 1 to 1.5 cm; The thickness of the PET base layer (5) is 180 μm to 220 μm; The thickness of the metal shielding layer (6) is 80 μm to 120 μm.

5. The flexible triboelectric sensor for detecting human vital signs according to claim 1, characterized in that: The flexible triboelectric sensor further comprises a PVC waterproof layer (7); The PVC waterproof layer (7) is arranged between the PET base layer (5) and the metal shielding layer (6), and the size of the PVC waterproof layer (7) is the same as that of the metal shielding layer (6).

6. The flexible triboelectric sensor for detecting human vital signs according to claim 1, characterized in that: The center of the translucent silicone film layer (1), the center of the multi-walled carbon nanotube composite sandpaper modified silicone layer (2), the center of the conductive fabric electrode layer (4), the center of the PET base layer (5), the center of the metal shielding layer (6) and the center of the PVC waterproof layer (7) are all on the same straight line.

7. The flexible triboelectric sensor for detecting human vital signs according to claim 1, characterized in that: The fastest response time of the flexible triboelectric sensor is 42ms.

8. A method for preparing a flexible triboelectric sensor for detecting human vital signs as claimed in any one of claims 1 to 7, characterized in that: The method comprises the following steps: Step 1, drop silica gel with a hardness of 0 degrees on the center of the spin disk of a spin coater, and perform spin coating. The spin coater's rotation speed is set to three revolutions per second, and the spin coating is performed for 30 to 40 seconds. After drying for 3 to 3.5 hours, the semi-transparent silica gel film layer (1) is obtained. Step 2, placing the fabric electrode substrate under the screen of a screen printer, and applying a conductive metal slurry on the screen with a drop amount of 1 mL each time, and performing the next drop application after the conductive metal slurry fully penetrates into the fiber gaps of the fabric electrode substrate each time, and the drop application is repeated 3 to 6 times to obtain an electrode layer (4); Step 3, adding 0.2g~0.4g of multi-walled carbon nanotubes into 20g~40g of silica gel with a hardness of 20 degrees and stirring to obtain a silica gel solution mixed with the multi-walled carbon nanotubes; then sticking one side of sandpaper with a particle size of 1200~3600 mesh to the center of the upper surface of the rotating disk of the spin coater, dripping 3~4mL of the silica gel solution mixed with the multi-walled carbon nanotubes on the upper surface of the sandpaper, then setting the rotation speed of the spin coater to 3~5 revolutions per second, and spin coating for 30~40s; finally, removing the rotating disk, drying for 4~4.5h, and then cutting to obtain a multi-walled carbon nanotube composite sandpaper modified silica gel layer (2); Step 4, laminating the translucent silicone film layer (1), the multi-walled carbon nanotube composite sandpaper modified silicone layer (2), the foam isolation layer (3), the conductive fabric electrode layer (4), the PET substrate layer (5), the metal shielding layer (6) and the PVC waterproof layer (7) in sequence, connecting one side of the conductive fabric electrode layer (4) and one end of the DuPont wire with conductive silver glue, to obtain a flexible triboelectric sensor.

9. An intelligent detection method, characterized in that: The intelligent detection method is applied to an intelligent detection system including a flexible triboelectric sensor for detecting human vital signs as described in any one of claims 1 to 7, wherein the intelligent detection system also includes a signal analysis module. The intelligent detection method includes: The flexible triboelectric sensor for detecting human vital signs collects vital sign signals including the user's main body joint movements, five finger joint movements, breathing movements and pulse heart rate, and sends the vital sign signals to the signal analysis module; The signal analysis module receives the vital sign signal and determines the user's posture, gesture, breathing pattern and pulse rate based on the vital sign signal.

10. The intelligent detection method according to claim 9, characterized in that: Based on the vital sign signal, determining the user's posture, gesture, breathing pattern, and pulse rate includes: The preprocessed vital sign signal is input into the recognition network, and the recognition results including the user's posture, gesture, breathing pattern and pulse rate are output. The recognition network has an MLP-Mixer architecture, including a one-dimensional convolutional layer, a normalization layer, a mixed spatial information layer, a mixed position feature layer, a global average pooling layer and a fully connected layer.

Images