Flexible breathable pressure sensor, intelligent pillow, preparation method and application

Abstract

The invention discloses a flexible breathable pressure sensor, an intelligent pillow, a preparation method and application. The pressure sensor is made of paper-based fibers, not only has the characteristic of flexibility to adapt to compression deformation of the air bag pillow, but also has the characteristic of air permeability to meet the comfort requirement of long-time contact between the head of a human body and the pressure sensor; the pressure sensor is integrated on the surface of the air bag pillow through a single unit or a plurality of arrays, can sensitively monitor information such as magnitude, frequency and distribution of contact acting force of the head and the air bag pillow in the snoring process of a human body or under different sleeping postures, and processes multi-dimensional sensing data based on an artificial intelligence algorithm. The intelligent functions of snoring suffocation early warning and head posture recognition are achieved. According to the intelligent pillow, the flexible and breathable pressure sensor is integrated into the air bag pillow, and the obtained intelligent pillow has wide application prospects in the fields of medical monitoring, health management, medical care rehabilitation and the like.

Description

Technical Field

[0001] The present invention relates to the field of flexible sensors, and in particular to a flexible breathable pressure sensor, a smart pillow, and a preparation method and application thereof. Background Art

[0002] With the rapid development of flexible electronic technology and smart sensors, flexible sensors have been widely used in health monitoring, smart wearable devices and other fields. These sensors can monitor human health status in real time, such as physiological parameters such as movement, posture, and breathing. However, traditional flexible sensors still face problems such as insufficient sensitivity, poor stability and poor adaptability in pressure detection in complex environments, which limits their application in some high-precision scenarios. In addition, most traditional flexible sensors are built on solid polymer film substrates, which are airtight and airtight. Long-term contact with the skin hinders the micro-exchange of substances between the skin and the external environment, leading to health problems such as skin redness, swelling and inflammation. At the same time, airbag technology, as a mature pressure regulation and support technology, has been widely used in the field of medical equipment. Airbags can effectively regulate pressure and provide comfortable support, but existing airbag technologies often encounter challenges in precise air pressure control, dynamic response capabilities, and the ability to adapt to variable shapes in practical applications. In terms of application, human sleep is an important process for self-functional repair, and the total sleep time of humans accounts for one-third of life. Therefore, monitoring the state of the human body during sleep is very important, especially for patients with diseases, such as suffocation caused by improper snoring and stiff neck caused by improper head posture. Therefore, how to organically combine flexible sensors with airbag pillows to achieve efficient dynamic pressure monitoring, ensure precise air pressure control and adaptability, and intelligently judge the human body's sleep condition based on the collected sensor information data of the interaction between the human body and the airbag pillow has become an important topic in the current research field. Summary of the invention

[0003] In order to solve the above-mentioned technical problems, the present invention provides a flexible breathable pressure sensor, a smart pillow, and a preparation method and application. This application can achieve efficient dynamic pressure monitoring and precise air pressure regulation by effectively combining a flexible breathable paper-based pressure sensor with airbag pillow technology. The layout of the sensor unit or array can capture pressure changes related to health data such as the user's head posture and snoring in real time. Combined with artificial intelligence algorithms, it can accurately identify the user's snoring, head posture, and other sleep health states, providing users with more real-time and accurate health monitoring data. Compared with traditional sensors, the present invention has higher sensitivity and response speed in complex health monitoring environments, and is particularly suitable for fields such as medical monitoring and health management, further improving the accuracy and adaptability of health monitoring.

[0004] In a first aspect, the present invention provides a method for preparing a flexible breathable pressure sensor, which is achieved through the following technical solutions.

[0005] A method for preparing a flexible air-permeable pressure sensor comprises the following steps:

[0006] S1. Preparation of the sensor sensitive layer

[0007] After mixing MXene and deionized water at a mass ratio of 1:5, stir at 4000-5000 rpm for 20-30 minutes to make them evenly dispersed, and then ultrasonically disperse them for no more than 5 minutes; then spread the dust-free paper flat on a clean bottom plate, pour the dispersed MXene aqueous solution on the paper, and dip-coat it to ensure that the paper surface is evenly coated. After dipping, dry it to remove moisture to obtain the sensor sensitive layer;

[0008] S2. Preparation of sensor electrode layer

[0009] Silver paste is evenly printed on the surface of dust-free paper by screen printing technology. During the printing process, the ink passes through the fine pores of the screen template to form an interdigitated structure on the surface of the dust-free paper. The screen is then separated from the dust-free paper, and the printed electrode is cured to obtain a sensor electrode layer.

[0010] S3. Vertically stack the layers in the order of packaging layer-sensitive layer-electrode layer, wherein the packaging layer is made of dust-free paper.

[0011] Further, in step S1, the preparation method of MXene is: deionized water and concentrated hydrochloric acid are mixed, lithium fluoride is added under stirring, and after stirring for 5 minutes, Ti is added. 3 AIC 2 , stirred at 47 ° C for 24 hours to obtain a mixed solution; the mixed solution was centrifuged and washed several times with deionized water to obtain a mixed slurry, a small amount of deionized water was added to the mixed slurry, ultrasonic treatment was performed for 5 minutes, and centrifuged at 3000 rpm for 20 minutes to obtain a MXene stock solution; wherein the volume ratio of deionized water to concentrated hydrochloric acid was 1:3, and the volume ratio of lithium fluoride to Ti 3 AIC 2 The mass ratio of lithium fluoride to concentrated hydrochloric acid is 8:5, and the mass volume of lithium fluoride and concentrated hydrochloric acid is 0.1g / ml.

[0012] Furthermore, in step S1, the drying conditions are: drying at 40-60° C. for 4-6 hours.

[0013] Furthermore, in step S2, the curing conditions are: drying at 40-60° C. for 30-60 min.

[0014] In a second aspect, the present invention provides a flexible breathable pressure sensor, which is realized by the following technical solution.

[0015] A flexible breathable pressure sensor prepared by the above preparation method.

[0016] In a third aspect, the present invention provides a use of a flexible breathable pressure sensor, which is achieved through the following technical solution.

[0017] An application of the above-mentioned flexible breathable pressure sensor in sleep monitoring.

[0018] In a fourth aspect, the present invention provides a method for preparing a smart pillow, which is achieved through the following technical solutions.

[0019] A method for preparing a smart pillow comprises the following steps:

[0020] a. Design the shape and size of the upper and lower thermoplastic polyurethane films and cut them, and reserve a circular hole on one of the thermoplastic polyurethane films for connecting the inlet and outlet pipes;

[0021] b. Selecting a thermoplastic polyurethane film without reserved circular holes, screen printing an electrode thereon, printing the silver paste evenly on the thermoplastic polyurethane film, forming one or more interdigitated structures on the thermoplastic polyurethane film, and obtaining an airbag-sensor electrode;

[0022] c. Cover the above sensor sensitive layer on the printed electrode, and then cover the sensitive layer with a layer of dust-free paper;

[0023] d. Heat-bonding the upper and lower layers of thermoplastic polyurethane film and the air nozzle to the thermoplastic polyurethane film with circular holes to form a closed airbag structure.

[0024] In a fifth aspect, the present invention provides a smart pillow, which is realized through the following technical solutions.

[0025] A smart pillow prepared by the above preparation method.

[0026] In a sixth aspect, the present invention provides a use of a smart pillow, which is achieved through the following technical solutions.

[0027] An application of the above-mentioned smart pillow in sleep monitoring.

[0028] Furthermore, the sleep monitoring includes identifying sitting and lying postures, snoring patterns, and head postures.

[0029] This application has the following beneficial effects.

[0030] (1) High sensitivity: Based on the natural fiber skeleton structure of dust-free paper, it can effectively construct a conductive path, providing a relatively ideal electronic conduction platform for pressure sensors. The MXene material itself has excellent conductive properties, and its surface contains rich functional groups (such as -OH, -F and -O). These functional groups not only give MXene a strong hydrophilicity, but also make it compatible with aqueous solutions or biological environments. Because MXene has strong chemical activity, it can interact well with the external environment during the use of the sensor, further enhancing the sensor's ability to respond to small pressure changes. In addition, the metallic conductivity of MXene ensures the high efficiency of charge transfer, allowing the sensor to maintain a good linear response over a wide pressure range. Combined with the flexible characteristics of dust-free paper, these advantages make the sensor not only have high sensitivity, but also have a wide pressure range and good stability, and can be widely used in scenarios requiring high-precision pressure monitoring.

[0031] (2) Stability: The softness and elasticity of the airbag enable the sensor to work stably under different air pressure conditions. By adjusting the inflation and exhaust of the airbag, it can adapt to different working environments. Whether under high or low pressure conditions, the sensor can respond stably and provide accurate pressure data. In addition, the combination of the airbag and the sensor effectively reduces performance fluctuations caused by external interference or pressure changes, ensuring stability and reliability in long-term use.

[0032] (3) Comfort: The fibers of dust-free paper are interwoven to form a structure with tiny pores. The porosity is high, allowing air and gas molecules to pass through more easily. It is also very flexible and breathable. It avoids the discomfort that may be caused by traditional hard sensors. Even after long-term contact with the skin, the skin will not become red, swollen or inflamed. It has good biocompatibility. The airbag can adaptively change its shape and pressure distribution according to the different forms and movements of the human body, providing a more comfortable use experience. Compared with traditional wearable devices, the softness and elasticity of the airbag make it more comfortable to wear and improve comfort. It is suitable for health management applications that require long-term monitoring.

[0033] (4) Intelligence: This application can not only realize real-time monitoring of the sitting and lying postures of the human body, but also effectively identify snoring patterns. By analyzing the pressure signal of the airbag-sensor unit, it can accurately identify the transition between the sitting state and the lying state, and provide posture adjustment prompts through real-time data feedback. According to the amplitude and duration of the snoring waveform, it can effectively distinguish between normal and abnormal snoring patterns, and issue a warning signal in time to remind the sleeper to adjust the sleeping posture or seek medical help. By arranging a sensor array on the airbag and combining the use of deep learning to process the time series data collected by the flexible breathable sensor, it is possible to effectively extract key features from the sensor's waveform signal, thereby achieving accurate sleep posture recognition. This application has broad application prospects, especially in the fields of health monitoring and sleep management. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a flow chart of MXene preparation of the present invention;

[0035] Figure 2 is a flow chart of electrode preparation of the present invention;

[0036] Figure 3 is a flow chart for preparing the sensitive layer of the sensor of the present invention;

[0037] Figure 4 It is a sensor packaging flow chart of the present invention;

[0038] Figure 5 This is a physical picture and characteristic characterization picture of the screen-printed electrode of the present invention;

[0039] Figure 6 The physical image and SEM image of the sensitive layer of the sensor of the present invention are shown;

[0040] Figure 7 This is a physical picture of the sensor monomer of the present invention;

[0041] Figure 8 This is a bending test diagram of the sensor of the present invention;

[0042] Fig. 9 This is a diagram showing the moisture permeability test results of the sensor of the present invention;

[0043] Fig.10 is a diagram of the biocompatibility test results of the sensor of the present invention;

[0044] Fig.11 It is a sensitivity variation curve diagram of the sensor of the present invention;

[0045] Fig.12 is a response time curve graph of the sensor of the present invention;

[0046] Fig.13 This is a test diagram of the cyclic stability of the sensor of the present invention;

[0047] Fig.14 is a flow chart of the preparation of the airbag of the present invention;

[0048] Fig.15 This is a PCB diagram of the airbag pressure detection board of the present invention;

[0049] Fig.16 This is a diagram of the preparation of the airbag-sensor electrode of the present invention;

[0050] Fig.17 This is a physical picture of the airbag-sensor package of the present invention;

[0051] Fig.18 This is a diagram of the preparation of the airbag-sensor array electrode of the present invention;

[0052] Fig.19 This is a physical picture of the airbag-sensor array of the present invention;

[0053] Fig. 20 This is the airbag-sensor signal acquisition diagram of the present invention;

[0054] Fig.21 It is a sensitivity test curve diagram of the airbag of the present invention under different air pressures;

[0055] Fig. 22 It is a test curve diagram of airbag-sensor response time of the present invention;

[0056] Fig.23 It is a curve diagram of the airbag-sensor cycle stability test of the present invention;

[0057] Fig.24 This is a schematic diagram of the sitting and lying posture recognition of the present invention;

[0058] Fig.25 is a schematic diagram of snoring pattern recognition according to the present invention;

[0059] Fig.26 It is a diagram of the internal structure of the deep neural network of the present invention;

[0060] Fig. 27 is a mapping diagram of the training data of the present invention in a two-dimensional feature space;

[0061] Fig.28 It is a combination diagram of different characteristic waveforms of three head postures obtained by the airbag-sensor array of the present invention;

[0062] Fig.29 It is a confusion matrix diagram of the head posture recognition classification results of the present invention. DETAILED DESCRIPTION

[0063] The invention is further described below with reference to the accompanying drawings and embodiments. Unless otherwise specified, the experimental method used in the present invention is a conventional method, and the experimental equipment, materials, reagents, etc. used can be purchased from relevant material sales companies.

[0064] like Figure 1 As shown: First weigh the medicine with weighing paper, take 0.8g of lithium fluoride, Ti 3 AIC 2 0.5g is reserved, and then 2.5ml of deionized water is taken out with a 1-5ml pipette into the material bottle, and 7.5ml of concentrated hydrochloric acid is taken out and slowly added into the bottle in small amounts several times. The bottle is placed on a magnetic stirrer, and then lithium fluoride is added (stirred for 5 minutes), and then Ti is slowly added. 3 AIC 2 , 47℃, stir for 24 hours. After stirring, centrifuge and pour the prepared medicine evenly into four centrifuge tubes, add deionized water to 3 / 4 of the centrifuge tube, put the medicine symmetrically, centrifuge for 5 minutes at 10,000 rpm, take out the sample, pour out the supernatant, add deionized water again, shake well, and centrifuge it in a centrifuge. Repeat 8 times. After pouring out the supernatant for the last time, add a little deionized water, shake the sample evenly, pour it into a centrifuge tube, and place it in an ultrasonic cleaner for 5 minutes. After the end, divide the sample in the bottle evenly into 2 centrifuge tubes, centrifuge once, 3000 rpm, 20 minutes. After the centrifugation, pour the supernatant into the sample bottle, ultrasonically evenly for 10 minutes, take 2ml of the sample into a polytetrafluoroethylene load-bearing box with a 1-5ml test tube, and dry the box in a vacuum drying oven at 60℃. After drying, weigh the MXene mass, calculate the concentration (mg / ml), and attach a concentration label.

[0065] like Figure 2 As shown: Screen-printed electrodes (SPE) use screen printing technology to print conductive paste layer by layer on an inert solid flat substrate, use a screen or hollow template to make an electrode pattern, and dry to remove the solvent in the electrode paste to make a solid electrode. During the screen printing process, a rubber knife is used to pass the ink through the open area of ​​the screen to form the desired design pattern on the surface of the substrate. The screen is then separated from the substrate, leaving the paste in the desired design. Subsequently, the printed electrode is cured in different states and dried at 60°C for 30 minutes. Different types of grids can be used for printing depending on the application requirements of the electrode.

[0066] like Figure 3As shown: In order to prepare the sensitive layer of the flexible pressure sensor, MXene and deionized water are first mixed in a mass ratio of 1:5. After mixing MXene with water, use a mechanical stirrer to stir for 20 minutes to ensure uniform dispersion, while avoiding damage to the MXene layer due to excessive stirring speed. Then ultrasonically disperse for 5 minutes. If necessary, the time can be appropriately extended. Pay attention to temperature control to prevent overheating from damaging the MXene structure. Then spread the dust-free paper flat on a clean custom acrylic board to ensure that the paper is flat and dust-free. Pour the dispersed MXene aqueous solution onto the paper and dip-coat it for 10 seconds to ensure that the paper surface is evenly coated. After dipping, remove the moisture by vacuum drying. The drying temperature is controlled at 50°C and the drying time is about 5 hours to obtain the MXene fiber substrate.

[0067] like Figure 4 As shown: After the sensitive layer and silver paste electrode of the pressure sensor are prepared, the layers are stacked vertically in the order of "encapsulation layer-sensitive layer-electrode layer". The electrode layer is bonded to the sensitive layer using TPU hot melt film to ensure the stability of the sensor. The encapsulation layer uses dust-free paper, and the encapsulation layer covers the sensitive layer.

[0068] like Figure 5 As shown: Silver paste (silver paste using BASE-SCD2 from Mifang Technology) is evenly printed on the surface of dust-free paper through screen printing technology to ensure the accuracy and uniformity of the electrode pattern, as shown Figure 5 As shown in (a), during the printing process, the silver paste is transferred to the surface of the dust-free paper through the fine pores of the screen template, forming a delicate and flat interdigital structure. This method not only improves production efficiency, but also ensures high precision and good conductivity of the electrode layer. The flatness and stability of the interdigital electrode are crucial to improving the sensitivity and response speed of the sensor, and help ensure the stability and high efficiency of the sensor during operation. It is also suitable for large-scale production and meets the manufacturing needs of flexible electronic devices. Characterization of cross-sectional micromorphology Figure 5 (b) and the silver (Ag) element mapping distribution diagram (c), the silver paste is evenly printed on the three-dimensional skeleton of the dust-free paper during the screen printing process, forming a continuous and dense electrode structure. These silver paste particles are filled in the three-dimensional pore network of the dust-free paper, fully contacting the fibers to form a stable electrode layer. Due to the good conductivity of the silver paste, it can provide an efficient current conduction path between the sensitive layer and the electrode, ensuring the conductivity and response speed of the sensor. The continuity of the silver paste layer also ensures the overall stability of the electrode and avoids possible poor contact or current leakage problems. This electrode structure formed by depositing silver paste on the fabric frame is of great significance to the long-term stable operation and reliability of the sensor.

[0069] like Figure 6As shown in the figure, the dust-free paper has a natural three-dimensional porous structure, good support and structural stability, and is soft and breathable, which facilitates the attachment of MXene sheets, which plays a key role in improving the sensitivity of the sensor and becomes the core component of the sensor. In order to more intuitively observe the micro-nanostructure on the sensitive layer, the surface morphology was characterized in detail using a scanning electron microscope (SEM). Figure 6 (a) shows a real picture of dust-free paper after being soaked in MXene solution. The real picture shows that the paper is black, indicating that MXene has been successfully integrated into the paper substrate. Figure 6 (b) Scanning electron microscopy image further shows that the MXene film forms a tightly wrinkled coating around the paper fibers. Figure 6 (c) Further magnification of the surface of the sensitive layer shows the flaky MXene structure. The three-dimensional porous skeleton of the dust-free paper has a rich pore structure and a large surface area, which provides excellent conditions for the attachment of MXene. Its porous framework not only provides a stable support for the MXene flakes, but also effectively adsorbs and disperses MXene, thereby enhancing its bonding with the dust-free paper. This three-dimensional structure of dust-free paper can also ensure that MXene is evenly distributed on the surface, avoid flake agglomeration, and ensure the efficient conductivity and excellent mechanical properties of the sensitive layer. The actual picture of the sensor monomer is as follows Figure 7 shown.

[0070] like Figure 8 As shown: The assembled flexible pressure sensor has excellent flexibility due to the use of flexible paper as the substrate, which allows the sensor to be bent and twisted without damage, thus laying the foundation for the attachment between the sensor and human skin and the detection of sensing signals. The resistance after bending was measured with a bending radius of 2 mm. The results showed that after bending 1,000 times, the resistance changed very little (within 2.4%), indicating that its resistance is always stable and can be used frequently and flexibly.

[0071] like Fig. 9 Shown: To demonstrate the air permeability of paper-based sensors, beakers filled with water were covered with sensors and different films for several days at the same room temperature and 44% humidity, and the remaining moisture content was measured. The open beaker had the largest water loss rate, which is the most breathable and moisture-permeable case. The moisture permeability of dust-free paper ranked second, indicating that dust-free paper is breathable due to its inherent fabric pores. After dip coating with MXene, the water loss rate of dust-free paper is still similar to that of the original dust-free paper, indicating that its good air permeability is well preserved. In contrast, beakers covered with impermeable polyimide (PI) and thermoplastic polyurethane (TPU) films (traditional polymer substrates widely used in the development of flexible electronics) are almost impermeable.

[0072] like Fig.10As shown: After 10 days of long-term attachment of the sensor, there was no redness, swelling, inflammation or allergic reaction on the skin surface, which shows that the sensor material has good biocompatibility and can be safely in contact with the skin and used for a long time. In contrast, after ten days of attachment, the PDMS material showed obvious signs of redness on the skin, which may be due to the slight irritation or allergic reaction caused by the material during long-term contact with the skin. This comparison result highlights the advantages of the tested sensor material in biocompatibility, which can effectively reduce irritation or damage to the skin, showing its wide potential and safety.

[0073] like Fig.11 As shown: The sensor sensitivity curve can be roughly divided into three sections: In the low pressure range of 0-20kPa, the sensor sensitivity is 16.7kPa -1 , in the medium pressure range of 21-65kPa, the sensor sensitivity is 10.6kPa -1 , in the high pressure range of 66-110kPa, the sensor sensitivity is 6.1kPa -1 . This is because when the paper fibers are first compressed, the pore space between them is compressed, and the distribution and contact points of the MXene flakes between the paper fibers change significantly. In this process, the conductive paths (such as contact points) between the MXene flakes undergo significant deformation and adjustment, resulting in the most sensitive resistance change. Because the overlap and contact conditions of the conductive paths vary greatly, a small external force (pressure or strain) will cause a large resistance change, thereby maximizing the sensitivity. Then, as the pressure increases, the pore space is gradually completely compressed, the contact between the MXene flakes becomes closer, and the network deformation tends to saturation. At this point, re-pressurization mainly leads to the compression of the paper fibers and the reduction of the overall thickness. Although the compression of the paper fibers will continue to affect the conductive network, since most of the pores have been closed, the deformation of the conductive paths is no longer significant, so the resistance changes more slowly and the sensitivity is relatively low. Finally, when the compression force reaches its limit, the overall geometric deformation of the paper tends to saturation, causing the further change of the resistance response to become slow, or even stable or saturated, and the sensitivity will be significantly reduced.

[0074] like Fig.12 As shown in the figure: The sensor is loaded and unloaded with a 15kPa low pressure to study its response time. Before the test, the test frequency of the LCR meter is set to 1kHz and the sampling interval is 0.25ms. The experimental results show that when the sensor is loaded and unloaded with a 15kPa low pressure, the response time is 25ms and the recovery time is 50ms. Fig.12As shown, the sensor's real-time monitoring capability under rapid pressure changes is demonstrated, indicating that it can accurately respond to changes in external pressure in a short period of time and recover in time. This feature makes the sensor have wide application potential in practical applications, especially in situations requiring fast feedback and high-precision measurement, such as smart wearable devices, medical monitoring, and industrial sensing.

[0075] like Fig.13 As shown in the figure: The service life of the sensor is closely related to its working stability. Sensors with strong working stability can maintain the accuracy and reliability of test results under changing environmental conditions. To verify this, a press was used to perform 10,000 compression-release cycle experiments on the sensor. During the experiment, the relative current change amplitude and waveform of the sensor remained basically consistent. Fig.13 As shown, the results show that the pressure sensor exhibits excellent cycling stability.

[0076] like Fig.14 As shown: First, the airbag is modeled in the CAD software, and the upper and lower layers of the airbag structure are accurately drawn. The size of the airbag prepared in this application is 250×150mm. In order to ensure that there is sufficient process margin for subsequent production, additional dimensional space is appropriately added during the design. Next, a laser cutting machine is used to accurately cut the shape of the airbag, and special attention is paid to reserving a circular hole with a radius of 3mm on the lower layer of the airbag TPU film for subsequent connection of the inlet and outlet pipes. The design of the circular hole ensures the smooth connection between the airbag and the external pipe and ensures the flow of gas in and out. After the cutting is completed, a laser cutting machine is used to cut on the entire layer of TPU film to ensure that the cutting shape is consistent with the CAD model and to ensure accuracy to avoid material waste and dimensional errors. Then, a heat sealer is used to heat-seal the airbag. First, the air nozzle layer is made by a heat sealer, and the pre-designed air nozzle is fixed on the lower layer of TPU film with a reserved circular hole to ensure the accurate docking of the air nozzle position. Next, the upper and lower layers of TPU film are heat-melted, and the upper and lower layers of TPU film of the airbag are firmly connected together through high temperature and pressure to form a closed airbag structure. The temperature and pressure during the heat-sealing process must be strictly controlled to ensure the firmness and airtightness of the connection, thereby ensuring the safety and reliability of the airbag during use.

[0077] like Fig.15 As shown in the figure: After the airbag is made, the air pressure acquisition board is used to monitor the internal air pressure in real time. The air pressure acquisition board can accurately collect the air pressure changes inside the airbag and transmit the data to the host computer system for further processing. Fig.15The physical picture of the air pressure acquisition board and the corresponding host computer program interface are shown (the airbag internal pressure data acquisition adopts the principle of communicating vessels. The airbag, trachea, US6330-006-S air pressure sensor and airbag inflation and deflation control system together constitute a closed system, thereby ensuring that the sensor can accurately measure the pressure inside the airbag. The real-time monitoring system of the airbag internal pressure receives the pressure data sent by the controller through the serial port, and unpacks the data on the host computer side, and converts the received raw data into the airbag internal pressure value. Through the relationship curve between the air pressure sensor and the internal pressure, the system can calculate the actual internal pressure value. The system collects data every 100 milliseconds to ensure real-time monitoring of the airbag internal pressure value). The air pressure acquisition board continuously collects data on the internal pressure of the airbag through the built-in sensor. These data include the real-time value and fluctuation of the air pressure, which provides an important basis for the performance evaluation of the airbag. After the acquisition is completed, the data is displayed on the computer screen through the host computer program, and the changes in the airbag internal pressure can be viewed in real time to ensure the stability and reliability of the airbag under different working conditions. This process provides basic data and technical support for further research on the subsequent airbag-sensor combination, and can provide an important reference for the performance analysis and optimization of the combination of airbags and sensors, ensuring its accurate monitoring and control in practical applications.

[0078] like Fig.16 As shown in the previous production process, after the airbag is made, the airbag is first completely deflated, and then the electrode is screen printed. Fig.16 As shown in (a), this approach has some problems in actual operation, which are mainly reflected in two aspects: when the airbag is deflated, the surface of the airbag becomes loose and unstable, making it difficult to accurately fix the printing position of the electrode, which in turn affects the alignment accuracy of the electrode; secondly, because there are inflation and deflation nozzles on the airbag, the surface of these areas is uneven, which may cause breakage or discontinuity during electrode printing, affecting the conductivity and overall performance of the electrode, thereby reducing the qualified rate and production efficiency of the finished product. Therefore, improvements are made to address the above problems. The new process is to first print the electrode on a single-layer TPU film, such as Fig.16 As shown in (b), the side without the inflation and discharge nozzle is selected. The surface of the TPU film is very flat and there is no interference from the nozzle, which can provide a more stable and uniform printing substrate. Screen printing the electrode on this flat film can not only ensure the flatness of the electrode, but also greatly improve the continuity of the electrode, avoiding the previous electrode discontinuity problem. Because this printing method provides better substrate support, the uniformity and adhesion of the electrode layer are also significantly improved, so that the electrode can be firmly fixed on the surface of the airbag, laying the foundation for subsequent sensor integration.

[0079] like Fig.17As shown in the figure: After the preparation of the sensor electrode is completed, the airbag-sensor is packaged as a whole to ensure the stability and reliability of the sensor during use. During the packaging process, the sensitive layer is first accurately placed on top of the printed electrode to form a good contact with the electrode. A layer of dust-free paper of the same material as the sensitive layer is then placed on top of the sensitive layer to avoid direct contact between the sensor sensitive layer and the skin, thus ensuring the long-term stability and accuracy of the sensor. Fig.17 shown.

[0080] like Fig.18 As shown: First, the above-mentioned airbag sensor preparation process is still used to print the electrodes on a single-layer TPU film to ensure the accuracy and stability of the electrode pattern. Fig.18 As shown, the nozzle layer and the upper and lower layers of TPU are then heat-melted to check air tightness to ensure there is no gas leakage. The area of ​​the electrode's interdigital part is 10×10mm2, and its overall width is 130mm. This design ensures that the size of the electrode array is within the effective contact area of ​​the airbag and can fully cover the force-bearing area on the airbag surface, thereby ensuring that the sensor can effectively capture pressure changes during the deformation of the airbag. The sensor part is then packaged. The actual airbag-sensor picture is as shown below. Fig.19 shown.

[0081] like Fig. 20 As shown in the figure: After completing the preparation of the airbag-sensor, the design of data collection and analysis work began. The data collection part is collected by an 8-channel pressure sensor board (Jiangsu Nengsida Electronic Technology Flexible Pressure Film Sensor MY2901) to ensure that real-time data can be obtained from multiple sensor nodes at the same time. The PCB board is as follows Fig. 20 As shown in (a), the obtained data is displayed on the computer by the host computer, such as Fig. 20 (b) as shown.

[0082] like Fig.21 As shown in the figure: After the preparation of the airbag-sensor monomer was completed, the performance of the airbag under different air pressures was analyzed. First, the relative current change of the airbag under different air pressure conditions was tested. Fig.21 As shown in the figure, when the pressure inside the airbag is low, the airbag material is softer and the overall deformation amplitude is larger, resulting in a lower response sensitivity of the sensor and a smaller relative current change. As the pressure inside the airbag gradually increases, the airbag becomes harder, the deformation amplitude decreases, and the resistance change of the sensor becomes more obvious, resulting in increased sensitivity and an increase in the current change amplitude. This shows that the sensitivity of the airbag-sensor monomer is positively correlated with the air pressure. The increase in air pressure can effectively improve the sensitivity of the sensor, enabling it to have higher detection accuracy and response capabilities under high-pressure environments.

[0083] like Fig. 22As shown: Before the test, the LCR meter test frequency was set to 1kHz and the sampling interval was set to 0.25ms. Fig. 22 As shown, the experimental results show that due to the influence of the soft texture of the airbag, the response time of the sensor becomes more linear, and due to the elastic properties of the airbag material, the recovery time of the sensor increases. This indicates that the softness of the airbag has delayed its recovery to its initial state to a certain extent. This may be because it takes a longer time for the airbag to rebound to its original shape after deformation, thus affecting the recovery process of the sensor.

[0084] like Fig.23 As shown in the figure, the service life of the sensor is closely related to its working stability. Sensors with strong working stability can maintain the accuracy and reliability of test results under changing environmental conditions. To verify this, a press was used to perform 10,000 compression-release cycles on the sensor. During the experiment, the relative current change amplitude and waveform of the airbag-sensor were basically consistent with the performance of the sensor unit, as shown in the figure. Fig.23 The experimental results show that despite a large number of compression-release cycles, the performance of the sensor has hardly declined significantly, showing good repeatability and stability. This shows that the pressure sensor has excellent cycle stability, can maintain reliable performance during long-term use, and can adapt to complex application environments.

[0085] like Fig.24 As shown: Next, the sensor is used to identify the sitting and lying postures. When the subject is in a sitting position, since no pressure is applied, the sensor does not receive obvious pressure changes, so the relative current signal remains stable with almost no fluctuations, showing a stable baseline. When the subject turns to a lying position, the weight of the head will act evenly on the sensor surface, causing the airbag sensor to be compressed, resulting in corresponding pressure changes. The results are shown in Fig.24 As shown in the figure, at this time, the relative current of the sensor will change significantly, showing a trend of fluctuation and increase, reflecting the pressure difference caused by the change of posture. By real-time monitoring of the change of current signal, it is possible to accurately distinguish whether the subject is in a sitting or lying state, thereby achieving effective recognition of sitting and lying postures.

[0086] like Fig.25 As shown in the figure: When performing snoring detection, different snoring patterns can be effectively distinguished according to the amplitude and duration of the snoring waveform. Normal snoring patterns usually show regular snoring, with each snoring lasting about 2 seconds, followed by a rest period of about 2 seconds. Fig.25(a) This snoring pattern is usually relatively stable and does not seriously affect the sleeper's breathing. However, abnormal snoring patterns are characterized by snoring sounds that last for a long time, usually close to 5 seconds or even longer, such as Fig.25 (b) This type of snoring is often accompanied by long pauses or irregular breathing pauses, which is very likely to cause suffocation or serious breathing disorders, increasing the risk of sleep apnea syndrome. Through real-time monitoring of snoring waveforms, the sensor can identify abnormal snoring patterns and promptly issue a warning signal to remind the sleeper to adjust their posture or seek medical help. In this way, the sensor can not only effectively distinguish between normal and abnormal snoring behaviors, but also issue an alarm in time when abnormal snoring that endangers health occurs, helping sleepers avoid potential safety risks and thus ensuring their sleep quality and health.

[0087] like Fig.26 As shown in the figure: Deep learning is an important research direction in the field of artificial intelligence, which originated from the research of artificial neural networks (ANN). Its core idea is to imitate the neural network structure of the human brain, learn the high-level features of data through multi-level nonlinear transformation, and then achieve efficient pattern recognition and prediction tasks. Fig.26 As shown, deep learning is particularly suitable for complex and nonlinear data processing tasks, and can automatically extract meaningful features from large-scale data without relying on manually designed features. In this application, a 1DCNN model is used to process time series data collected by a flexible breathable sensor, which can effectively extract key features from the waveform signal of the sensor, thereby achieving accurate sleep posture recognition. This application designs a smart pillow that monitors the head posture of the sleeper in real time by integrating five flexible breathable sensors on the surface of the pillow. The sensor can detect slight pressure changes in the head during lying down, and when the head position changes, the signal output by the sensor will also change significantly. Through a multi-channel signal acquisition system, the signals of the five sensors are collected at the same time and sent to the 1D CNN model for processing and analysis. The 1D CNN model can effectively identify different head postures through training, including different sleeping postures such as lying on the left, lying on the right, and lying on the right. Each sleeping posture will produce different pressure distribution patterns on the sensor, which are gradually extracted by CNN through multiple convolutional layers and finally classified into different posture types. In this way, the smart pillow can not only monitor the user's sleeping posture in real time, but also effectively distinguish different head posture changes, providing accurate sleep posture monitoring function. The core goal of the research is to train the CNN model to identify the impact of different sleeping positions on sensor signals and achieve efficient posture classification. This technology can be applied to smart sleep monitoring devices to provide more accurate data support for sleep quality assessment and help users improve their sleep health.

[0088] like Fig. 27 As shown: The mapping of training data in the two-dimensional feature space is as follows Fig. 27 As shown in the figure, the high-dimensional features collected by the five sensors are mapped to a two-dimensional plane through dimensionality reduction technology, so as to better understand how the model distinguishes different head postures (lying on the left, lying straight, and lying on the right). As can be seen from the figure, different head postures form different clusters in the two-dimensional space, indicating that the model can extract discernible features through sensor signals.

[0089] like Fig.28 As shown: The characteristic waveform combinations obtained by the five sensors under three different head postures (lying left, lying straight, lying right) are as follows Fig.28 As shown in the figure. In each posture, the signal output by the sensor shows obvious time series characteristics, which change with the different postures of the head. By processing these signals through a one-dimensional convolutional neural network (1DCNN), local features that are helpful for classification can be effectively extracted. The figure clearly shows the waveform differences under various postures, and further understands how the sensor signals show different patterns under different head postures. These waveform differences provide important information for the training of subsequent models, enabling the model to recognize and distinguish different head postures, effectively improving the classification accuracy.

[0090] like Fig.29 As shown: By showing the confusion matrix of the head posture recognition classification results Fig.29 As shown, we can intuitively see the classification results of the model in three different postures and whether the model has misclassified. The analysis of the confusion matrix reveals the recognition accuracy of the model on the test set. The vast majority of test samples are accurately classified, and the number of misclassified samples is very small. In this experiment, the accuracy of the model reached 98%. This result shows that the model performs well in distinguishing left-lying, front-lying, and right-lying postures, and has high classification accuracy. The experimental results further verify the effectiveness and reliability of the deep learning model, especially when processing sensor waveform data, the convolutional neural network (CNN) shows a strong feature extraction and classification ability. This result not only proves the high efficiency of CNN in this application, but also provides strong support for the sensor-based sleep posture monitoring system.

[0091] The embodiments of this specific implementation method are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for preparing a flexible gas permeable pressure sensor, characterized in that: The following steps are involved: S1. Preparation of the sensor sensitive layer After mixing MXene and deionized water at a mass ratio of 1:5, stir at 4000-5000 rpm for 20-30 minutes to achieve uniform dispersion, and then ultrasonically disperse for no more than 5 minutes; Then, the dust-free paper is laid flat on a clean base plate, and the dispersed MXene aqueous solution is poured onto the paper for dip coating to ensure that the paper surface is evenly coated. After dip coating, the paper is dried to remove moisture to obtain the sensor sensitive layer. S2. Preparation of sensor electrode layer Silver paste is evenly printed on the surface of dust-free paper by screen printing technology. During the printing process, the ink passes through the fine pores of the screen template to form an interdigitated structure on the surface of the dust-free paper. The screen is then separated from the dust-free paper, and the printed electrode is cured to obtain a sensor electrode layer. S3. Vertically stack the layers in the order of packaging layer-sensitive layer-electrode layer, wherein the packaging layer is made of dust-free paper.

2. The method for preparing a flexible breathable pressure sensor according to claim 1, characterized in that: In step S1, the preparation method of MXene is: mixing deionized water and concentrated hydrochloric acid, adding lithium fluoride under stirring, stirring for 5 minutes, and then adding Ti3AIC2, stirring at 47°C for 24 hours to obtain a mixed solution; the mixed solution is centrifuged and washed several times with deionized water to obtain a mixed slurry, a small amount of deionized water is added to the mixed slurry, ultrasonically treated for 5 minutes, and centrifuged at 3000rpm for 20 minutes to obtain a MXene stock solution; wherein the volume ratio of deionized water to concentrated hydrochloric acid is 1:3, the mass ratio of lithium fluoride to Ti3AIC2 is 8:5, and the mass volume of lithium fluoride to concentrated hydrochloric acid is 0.1g / ml.

3. The method for preparing a flexible breathable pressure sensor according to claim 1, characterized in that: In step S1, the drying conditions are: drying at 40-60°C for 4-6 hours.

4. The method for preparing a flexible breathable pressure sensor according to claim 1, characterized in that: In step S2, the curing conditions are: drying at 40-60°C for 30-60 minutes.

5. A flexible breathable pressure sensor prepared by the preparation method according to any one of claims 1 to 4.

6. Use of the flexible breathable pressure sensor according to claim 5 in sleep monitoring.

7. A method for preparing a smart pillow, characterized in that: The following steps are involved: a. Design the shape and size of the upper and lower thermoplastic polyurethane films and cut them, and reserve a circular hole on one of the thermoplastic polyurethane films for connecting the inlet and outlet pipes; b. Selecting a thermoplastic polyurethane film without reserved circular holes, screen printing an electrode thereon, printing the silver paste evenly on the thermoplastic polyurethane film, forming one or more interdigitated structures on the thermoplastic polyurethane film, and obtaining an airbag-sensor electrode; c. Covering the sensitive layer of the sensor according to claim 1 on top of the printed electrode, and then covering the sensitive layer with a layer of dust-free paper; d. Heat-bonding the upper and lower layers of thermoplastic polyurethane film and the air nozzle to the thermoplastic polyurethane film with circular holes to form a closed airbag structure.

8. A smart pillow prepared by the preparation method according to claim 7.

9. Application of the smart pillow according to claim 8 in sleep monitoring.

10. The use according to claim 9, characterized in that: The sleep monitoring includes identifying sitting and lying postures, snoring patterns, and head postures.

Images