Intelligent insole based on ionic hydrogel sensor array and gait recognition method

Abstract

This invention belongs to the field of flexible wearable sensing and artificial intelligence gait recognition technology, and relates to a smart insole and gait recognition method based on an ion hydrogel sensor array. The hydrogel comprises the following components in parts by weight: 26-30 parts of sulfobetaine methacrylate, 7-8 parts of acrylamide, 13-17 parts of acrylic acid, 35-42 parts of deionized water, 7-8 parts of Laponite, 2.2-2.8 parts of lithium chloride, 7-8 parts of 1-ethyl-3-methylimidazolium chloride, 0.45-0.55 parts of N,N'-methylenebisacrylamide, 0.22-0.28 parts of ammonium persulfate, and 0.22-0.28 parts of tetramethylethylenediamine. The ion hydrogel of this invention achieves excellent antifreeze, anti-dehydration and self-healing properties through dual-ion synergistic design, while maintaining good mechanical toughness and electrical stability, and can adapt to various complex environmental conditions such as low temperature and dryness.

Description

Technical Field

[0001] This invention belongs to the field of flexible wearable sensing and artificial intelligence gait recognition technology, specifically relating to smart insoles based on ion hydrogel sensor arrays and gait recognition methods. Background Technology

[0002] Human plantar pressure directly reflects key biomechanical information such as gait patterns, foot weight distribution, limb posture control, and body balance. It has irreplaceable application value in many fields, including monitoring competitive sports performance, evaluating postoperative rehabilitation effects, daily health management for the elderly, early risk warning of diabetic foot, gait screening for neurodegenerative diseases such as Alzheimer's, and smart wearable human-computer interaction. Currently, standardized clinical gait analysis mainly uses professional instruments such as force tables, long-distance pressure tracks, and optical motion capture devices. While these devices offer high accuracy and comprehensive data dimensions, they generally suffer from drawbacks such as large size, high procurement and maintenance costs, and fixed deployment locations. They can only be used in specific indoor settings such as hospitals and professional sports laboratories, failing to meet the needs for long-term, continuous dynamic plantar pressure monitoring in free settings such as home, outdoor walking, and daily activities, resulting in significant limitations in application scenarios. Addressing the lack of portability of professional equipment, the industry has gradually introduced wearable smart insoles with integrated pressure sensing units, achieving lightweight and portable plantar pressure data collection, becoming the mainstream technology for daily gait monitoring. However, existing insole sensing solutions still have several core technological shortcomings that restrict their long-term stable application:

[0003] Firstly, the mechanical properties of traditional flexible pressure sensing substrates are poor. During human walking and running, the soles of the feet continuously generate a combination of compression, shearing, and bending forces, which can easily cause interface slippage inside the sensor. After long-term use, problems such as output signal drift and sensor performance fatigue decay occur, resulting in short service life and insufficient reliability of long-term monitoring data.

[0004] Secondly, existing conductive hydrogel flexible sensing materials rely on free water and free ion migration to form conductive pathways, which have poor environmental adaptability. They are prone to water freezing and loss of sensing function in low-temperature environments. When placed at room temperature for a long time, water evaporation and drying will significantly reduce conductivity. At the same time, the ion migration rate is unstable during the stress process, causing disorder in the pressure-electrical signal correspondence and poor consistency of sensing output.

[0005] Third, most existing insole sensing devices can only collect single-point or local plantar pressure values ​​and simple resistance change raw signals. They lack multi-channel spatiotemporal pressure signal integrated analysis algorithms, cannot automatically distinguish different synchronic movements, quantify balance and load characteristics, and are difficult to convert the underlying raw electrical signals into standardized gait biomechanical information that can be directly used for health assessment and motion analysis, resulting in low intelligence.

[0006] In summary, existing technologies lack flexible ionic hydrogel sensing materials that simultaneously possess high mechanical toughness, stable conductivity over a wide temperature range, resistance to drying and freezing, good interfacial adhesion, and self-healing properties. Furthermore, there is a lack of integrated insole-type plantar pressure monitoring solutions based on such high-performance sensing materials, featuring spatially distributed sensing arrays, stable wireless data transmission, and automatic gait intelligent analysis capabilities. These technologies struggle to balance the long-term sensing stability of the materials, the portability of the devices, and the need for intelligent gait data analysis. Therefore, there is an urgent need to develop related new materials, wearable sensing devices, and gait recognition methods. Summary of the Invention

[0007] This invention provides the following technical solution: an ionic hydrogel comprising the following components in parts by weight: 26-30 parts of sulfobetaine methacrylate, 7-8 parts of acrylamide, 13-17 parts of acrylic acid, 35-42 parts of deionized water, 7-8 parts of Laponite, 2.2-2.8 parts of lithium chloride, 7-8 parts of 1-ethyl-3-methylimidazolium chloride, 0.45-0.55 parts of N,N'-methylenebisacrylamide, 0.22-0.28 parts of ammonium persulfate, and 0.22-0.28 parts of tetramethylethylenediamine. The ionic hydrogel forms a multi-layered dynamic network with coexisting chemical cross-linking, hydrogen bonding, ionic interactions, coordination, and physical cross-linking of nanofillers, exhibiting high adhesion, self-healing, antifreeze, water retention, stable electrical conductivity, and resistance to cyclic compression and shearing.

[0008] Preferably, the ionic hydrogel has a maximum adhesion strength of 28.6 kPa, an elongation at break of 340%, a Young's modulus of 368.5 kPa, and a peak ionic conductivity of 2.06 S / m.

[0009] Preferably, the components of the ionic hydrogel function as follows: sulfobetaine methacrylate constructs an amphoteric hydration layer, enhancing water retention, anti-drying properties, and ion transport stability; acrylamide provides flexible segments, enhancing tensile strength; acrylic acid introduces carboxyl groups, improving adhesion, self-healing, and energy dissipation capabilities; Laponite forms reversible physical crosslinking points, strengthening mechanical toughness and fatigue resistance; lithium chloride and 1-ethyl-3-methylimidazolium chloride synergistically achieve stable conductivity over a wide temperature range and low-temperature freeze resistance; N,N'-methylenebisacrylamide constructs a chemical crosslinking framework; and ammonium persulfate and tetramethylethylenediamine initiate the free radical polymerization of the ternary monomers.

[0010] This invention also discloses a method for preparing an ionic hydrogel, which includes the following steps:

[0011] Step A1: Add Laponite powder to deionized water and ultrasonically disperse at room temperature for no less than 30 minutes to obtain a uniform nanofiller dispersion.

[0012] Step A2: Add sulfobetaine methacrylate, acrylamide, acrylic acid, 1-ethyl-3-methylimidazolium chloride, lithium chloride, N,N'-methylenebisacrylamide, and tetramethylethylenediamine to the dispersion in sequence, stir at room temperature for no less than 30 minutes, and remove air bubbles from the precursor solution under vacuum.

[0013] Step A3: Finally, add ammonium persulfate and stir until homogeneous. Pour the mixture into a mold and seal it. Let it stand at room temperature for at least 48 hours to complete the polymerization and obtain a ternary network ionic hydrogel.

[0014] This invention also discloses a smart insole based on an ionized hydrogel sensor array. The smart insole is composed of sensing units made of the aforementioned ionized hydrogel and includes: an insole base, a multi-channel pressure sensing array, a voltage divider acquisition circuit, a microcontroller, and a wireless transmission module. The sensing units are arranged in sections across all key pressure-bearing areas of the foot, including the heel area, arch area, metatarsal area, and forefoot area. Each sensing unit acts as a variable resistor connected to the voltage divider acquisition circuit; the resistance change during pressure deformation is converted into a standardized voltage signal. The microcontroller polls and acquires the multi-channel voltage signals and performs baseline correction and normalization preprocessing. The foot pressure timing data is then transmitted to a mobile terminal via the wireless transmission module.

[0015] Preferably, the voltage divider acquisition circuit is powered on only during the ADC sampling instant, with a sampling frequency of 8Hz and a single sampling power-on duration of 100μs; the smart insole is equipped with a wireless charging module, powered by reverse Qi wireless charging from the mobile terminal; the mobile terminal can display the plantar pressure curve and two-dimensional pressure-thermal distribution map in real time.

[0016] Preferably, the ion hydrogel sensing unit is matched and assembled with a flexible FPC, conductive fabric, and silver paste flexible electrode; the wireless transmission module adopts Bluetooth, Bluetooth Low Energy, or Wi-Fi; and the microcontroller is an STM32, ESP32, or nRF series microcontroller.

[0017] This invention also discloses a gait recognition method based on smart insoles, comprising the following steps:

[0018] Step B1, Multi-channel pressure signal acquisition: Wear the smart insole, and the microcontroller synchronously acquires the voltage timing signals output by the hydrogel sensing units of each foot zone.

[0019] Step B2, Baseline Correction: Acquire baseline signals under no-load and stationary standing conditions, and perform a difference calculation between the real-time signal and the baseline to eliminate baseline offset caused by individual weight, initial sensor resistance, and wearing deviation.

[0020] Step B3, Signal preprocessing: Noise is removed by moving average, median or low-pass filtering, and the sliding window aggregates the sampling points to generate a single time window pressure feature vector.

[0021] Step B4, Gait Cycle Segmentation: Based on the timing of peak pressure in the heel and forefoot channels, divide the complete gait cycle into heel strike, full foot support, heel lift-off, and forefoot push-off.

[0022] Step B5: Construct a gait feature matrix: Construct a two-dimensional feature matrix from the 8-channel timing signals within a single cycle or single window as the input for gait recognition.

[0023] Step B6, Model Classification and Recognition: Input the feature matrix into the gait recognition model and output the gait category and recognition confidence; the gait categories include normal gait, Parkinson's abnormal gait, diabetic foot high-pressure pattern, and inversion and eversion gait.

[0024] Step B7, Terminal Visual Feedback: The mobile terminal displays the pressure distribution and gait recognition results, outputs risk warnings when abnormal gait is identified, and generates a long-term gait assessment report.

[0025] Preferably, the gait recognition model is any one or a combination of convolutional neural networks, long short-term memory networks, Transformers, support vector machines, and random forests; the convolutional neural network includes an input layer, two convolutional pooling layers, a fully connected layer, and a classification output layer, used to extract the spatial pressure correlation and temporal dynamic features of the foot.

[0026] The beneficial effects of this invention are:

[0027] 1. This invention adopts a multi-hydrogel pressure sensing unit spatial layout scheme, with sensing units correspondingly arranged in all key pressure areas of the sole, such as the heel area, arch area, metatarsal area, and forefoot area. It can simultaneously collect dynamic pressure time-series signals from multiple areas of the entire sole, effectively overcoming the shortcomings of traditional single-point pressure sensors that have a single dimension of data collection and cannot completely reproduce the force change process of the human body's complete gait. The comprehensiveness of sole pressure data collection is significantly improved, providing complete multi-dimensional biomechanical raw data support for accurate gait analysis.

[0028] 2. This invention uses an ion hydrogel sensing unit as a variable resistor connected to a voltage divider acquisition circuit, which can stably convert continuous dynamic pressure changes on the sole of the foot into a standardized voltage signal. This signal is compatible with a microcontroller for multi-channel synchronous reading and supports wireless real-time transmission of sole pressure data. This solves the problems of unstable signal conversion, difficulty in multi-channel synchronous acquisition, and the restriction of wearing activities by wired data transmission in traditional sensing methods. The signal acquisition stability and ease of use are greatly improved.

[0029] 3. This invention integrates the dual functions of visual data display and intelligent gait recognition. On the one hand, it can intuitively output the real-time pressure change curve of the sole and the two-dimensional pressure distribution heat map of the sole, making it easy for users to intuitively view the pressure state of the foot. On the other hand, it relies on the built-in gait recognition model to automatically analyze the original pressure signal and output standardized gait classification results, breaking through the limitation of traditional insoles that can only passively store and record the original pressure data, realizing the upgrade from simple data collection to autonomous intelligent analysis, and significantly improving the intelligence level of the device.

[0030] 4. The gait recognition model of this invention can accurately distinguish between normal gait, abnormal gait of Parkinson's disease, abnormal high-pressure pattern of diabetic foot, foot inversion, foot eversion, and other typical pathological and movement gaits. It covers a variety of application scenarios such as sports training monitoring, postoperative rehabilitation effect evaluation, early screening of foot lesions, and long-term home health management. The device is highly versatile and can simultaneously meet different usage needs such as sports monitoring of the general population, foot risk screening of patients with chronic diseases, and gait correction assessment of rehabilitation populations. It has a wide range of practical application value. Attached Figure Description

[0031] Figure 1 This is a graph showing the adhesion performance of the ion hydrogel insole and gait recognition method based on an ion hydrogel sensor array according to the present invention; wherein, Figure 1 a shows the strong adhesion of ionic hydrogels to various substrate surfaces; Figure 1 b shows the results of the quantitative tensile-shear adhesion strength test;

[0032] Figure 2 The diagram shows the mechanical properties of the ion-hydrogel of the present invention; wherein, Figure 2 a represents the stress-strain curves of hydrogels with different AM:AA ratios; Figure 2 b is the cyclic tensile loading-unloading curve up to 100% strain;

[0033] Figure 3 The diagram shows the ionic conductivity of the ion hydrogel of the present invention at different EmimCl ratios; wherein, Figure 3 a is the Nyquist plot; Figure 3 b is the electrochemical impedance spectroscopy;

[0034] Figure 4 The conductivity of the ion hydrogel of the present invention at different temperatures is shown in the figure.

[0035] Figure 5 This is a normalized weight retention graph of the ionic hydrogel of the present invention under environmental conditions over time.

[0036] Figure 6 The graph shows the effect of Laponite content on the viscoelastic properties of the ionic hydrogel of this invention, which demonstrates its self-healing properties.

[0037] Figure 7 This is a diagram of a sensor insole showing the distribution of the eight sensing channels (CH1-CH8) of the ion hydrogel of the present invention on the sole of the foot.

[0038] Figure 8 This is a schematic diagram of the signal acquisition circuit and duty cycle control of the gait recognition method of the present invention;

[0039] Figure 9 This is a system-level architecture diagram of the wireless sensing platform for the gait recognition method of the present invention;

[0040] Figure 10 This is a diagram of the assembled wearable system in the gait recognition method of the present invention. Detailed Implementation

[0041] The relevant technologies of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0042] like Figures 1-10 As shown, the environmentally adaptable ternary network ionic hydrogel of this embodiment is composed of the following raw materials in parts by mass:

[0043] The composition includes 28 parts of sulfobetaine methacrylate (SBMA), 7.5 parts of acrylamide (AM), 15 parts of acrylic acid (AA), 38.5 parts of deionized water, 7.5 parts of Laponite, 2.5 parts of lithium chloride (LiCl), 7.5 parts of 1-ethyl-3-methylimidazolium chloride (EmimCl), 0.5 parts of N,N'-methylenebisacrylamide (MBA), 0.25 parts of ammonium persulfate (APS), and 0.25 parts of tetramethylethylenediamine (TEMED).

[0044] In this embodiment, the roles of each component of the ion-hydrogel are as follows:

[0045] SBMA: Provides zwitterionic groups and a hydration layer; AM is used to improve the flexibility and tensile properties of hydrogel segments. AA: Provides carboxyl groups, enhancing the adhesion, hydrogen bonding, and ionic interactions of the hydrogel. Laponite: Used to form reversible physical crosslinking points, improving the mechanical strength and energy dissipation capacity of the hydrogel. LiCl is used to regulate the water state in the hydrogel, improving its antifreeze and water retention properties. EmimCl: Used to improve ion dissociation and ion migration capabilities, thereby enhancing the electrical conductivity of the hydrogel. MBA is a crosslinking agent used to form a stable chemical crosslinking network, ensuring the basic structural strength of the hydrogel after molding. APS and TEMED are initiators used to initiate monomer radical polymerization, enabling the precursor solution to form a ternary network hydrogel.

[0046] The basis or advantages for selecting the content of each component in this embodiment are as follows:

[0047] The SBMA content was selected as 28 parts. When the content is too low, the number of zwitterionic groups in the hydrogel is insufficient, making it difficult to form a stable hydration layer, resulting in insufficient water retention, anti-drying properties, and ion transport stability. When the content is too high, the electrostatic interaction between molecular chains is too strong, which may lead to an overly dense hydrogel network or increased brittleness, which is not conducive to the tensile deformation and pressure response of the material.

[0048] The AM content was selected as 7.5 parts. Acrylamide can form flexible polymer segments, improving the flexibility and tensile properties of the hydrogel. When its content is too low, there are insufficient flexible segments, and the hydrogel is prone to structural damage during repeated compression or stretching. When its content is too high, the dynamic interaction sites in the system are relatively reduced, which may reduce the material's energy dissipation capacity, adhesion properties, and pressure sensing stability.

[0049] The AA content was selected as 15 parts. The carboxyl groups in acrylic acid can form hydrogen bonds, ionic interactions, or coordination effects with other polymer chains, inorganic nanofillers, and substrate surfaces, thereby enhancing the hydrogel's adhesion, energy dissipation capacity, and self-healing ability. When the content is too low, the number of carboxyl groups is insufficient, resulting in weak adhesion and dynamic cross-linking of the hydrogel; when the content is too high, excessive carboxyl groups may lead to overly strong network interactions, causing the material to harden or affecting ion migration. Therefore, controlling the acrylic acid content at 15 parts is beneficial for improving the hydrogel's adhesion, mechanical stability, and pressure response.

[0050] In this embodiment, SBMA can be partially replaced by monomers with zwitterionic structures or strong hydration capabilities, including but not limited to one or more of carboxybenzene methacrylate, phosphocholine methacrylate, and sulfonate betaine acrylamide monomers. AM can be partially replaced by vinyl monomers with flexible segments and hydrophilicity, including but not limited to one or more of N-isopropylacrylamide, N,N-dimethylacrylamide, hydroxyethylacrylamide, and methacrylamide. AA can be partially replaced by monomers containing carboxyl groups or other monomers capable of forming dynamic interactions, including but not limited to one or more of methacrylic acid, itaconic acid, maleic acid, vinylphosphonic acid, and 2-acrylamido-2-methylpropanesulfonic acid. Laponite can be partially replaced by inorganic nanofillers capable of forming physical crosslinks or reinforcing network structures, including but not limited to one or more of montmorillonite, lithium diatomaceous earth, bentonite, nano-silica, graphene oxide, cellulose nanocrystals, and cellulose nanofibers. LiCl can be partially replaced by inorganic salts with water-retaining, antifreeze, or regulating ionic environment within the hydrogel, including but not limited to one or more of sodium chloride, potassium chloride, calcium chloride, magnesium chloride, lithium sulfate, and lithium nitrate. MBA can be made from N,N'-methylenebisacrylamide, or one or more of polyethylene glycol diacrylate, ethylene glycol dimethacrylate, gelatin methacrylamide, and alginate crosslinking systems.

[0051] Electrode materials can be copper, gold, silver, platinum, carbon materials, conductive fabrics, conductive silver paste, flexible FPC electrodes, or other flexible conductive materials. The signal acquisition module can use an STM32 microcontroller, or Arduino, ESP32, nRF series chips, or other microcontrollers with multi-channel analog signal acquisition capabilities. Wireless transmission methods can include Bluetooth, Bluetooth Low Energy, Wi-Fi, NFC, or other wireless communication methods.

[0052] Gait recognition algorithms can employ convolutional neural networks, as well as support vector machines, random forests, long short-term memory networks, Transformer networks, or other machine learning and deep learning models.

[0053] 1. This invention constructs a ternary network ionic hydrogel composed of SBMA, AM, and AA. This hydrogel also incorporates Laponite, LiCl, and EmimCl, forming a multi-layered dynamic network structure with the combined effects of chemical cross-linking, hydrogen bonding, ionic interactions, coordination, and physical cross-linking of nanofillers. This achieves a simultaneous improvement in the hydrogel's water retention, freeze resistance, and ionic conductivity.

[0054] 2. This invention uses the aforementioned environmentally adaptable ionic hydrogel to prepare a pressure sensing unit, utilizing the resistance change caused by the deformation of the hydrogel under foot pressure to detect pressure signals. This sensing unit possesses flexibility, adhesion, self-healing properties, and stable conductivity, enabling it to adapt to the complex stress environment of repeated compression, bending, and shearing in insoles.

[0055] 3. This invention distributes multiple hydrogel pressure sensing units in the insole according to the anatomical stress areas of the foot, forming a multi-channel plantar pressure sensing array. This array can simultaneously collect pressure changes in the heel, arch, metatarsal region, and forefoot region, thereby obtaining gait signals with spatial distribution and temporal variation characteristics.

[0056] 4. The present invention further combines a multi-channel hydrogel pressure sensing array with a voltage divider reading circuit, a microcontroller, a wireless transmission module and a mobile terminal to realize real-time acquisition, wireless transmission and visualization of plantar pressure signals.

[0057] The preparation method of the ion hydrogel in this embodiment is based on a one-pot method, and is specifically implemented according to the following steps:

[0058] Step A1: First, dissolve Laponite (0.2 g) in deionized water (8 mL) and sonicate at room temperature for 30 minutes.

[0059] Step A2: Subsequently, SBMA (6.0 g), AM (1.5 g), EmimCl (1.5 g), LiCl (0.5 g), MBA (0.02 g), AA (3.0 mL), and TEMED (20 μL) were dissolved in the solution. After stirring continuously at room temperature for 30 minutes, the solution was degassed to remove air bubbles.

[0060] Step A3: Dissolve 0.2 g of APS in the solution and stir at room temperature for 30 minutes. Inject the homogenized solution into a PVC tube, seal to form a hydrogel, and cool at room temperature for 48 hours. Store at room temperature for further study.

[0061] In the ionized hydrogel method of this embodiment, step A1 involves dispersing Laponite in deionized water to pre-form a uniform nanofiller dispersion system. Stirring or ultrasonic dispersion is employed, utilizing external mechanical action to overcome the agglomeration between nano-clay particles, ensuring the Laponite nanosheets are uniformly dispersed in the aqueous phase. The advantage of this step is that it allows Laponite to be uniformly distributed within the hydrogel network during subsequent polymerization, forming reversible physical cross-linking points, thus improving the hydrogel's mechanical strength, energy dissipation capacity, and cycle stability.

[0062] In step A2, SBMA, AM, and AA are added to the Laponite dispersion to form a ternary monomer mixture. SBMA contains zwitterionic groups, which can form a stable hydration layer, improving the hydrogel's water retention and ion transport stability. AM can form flexible polymer segments, improving the hydrogel's tensile strength and flexibility. AA contains carboxyl groups, which can improve the hydrogel's adhesion, energy dissipation capacity, and self-healing ability through hydrogen bonding, ionic interactions, and coordination. The hydrogel precursor solution is degassed to remove air bubbles introduced during mixing and stirring. Static degassedness, vacuum degassedness, or centrifugal degassedness methods are used to reduce the air bubble content within the precursor solution, resulting in a more uniform and complete hydrogel structure after polymerization.

[0063] In step A3, an initiator is added before gelation to initiate the free radical polymerization of the ternary monomers. Water-soluble initiators such as APS are used because their decomposition generates free radicals that initiate the polymerization of SBMA, AM, and AA, forming a ternary polymer network. The advantage of adding the initiator last before gelation is that it avoids premature polymerization of the precursor liquid before mixing, degassing, or molding, thus ensuring good flowability and moldability of the precursor liquid.

[0064] The ionized hydrogel of this embodiment was tested, and the results are as follows:

[0065] like Figure 1 As can be seen, hydrogels exhibit strong and diverse adhesion properties to various substrates (including metals, polymers, elastomers, and biological tissues). Even under slight mechanical disturbance, hydrogels maintain close contact, demonstrating their ability to establish stable interfacial coupling. Quantitative shear test ( Figure 1 (b) Further confirmation indicates that this ionic hydrogel exhibits excellent adhesion strength on various substrates, including PET, rubber, copper, steel, and biological skin, with a maximum adhesion strength of 28.6 kPa. Notably, this hydrogel maintains strong adhesion not only on smooth, rigid surfaces but also on soft or compliant substrates.

[0066] like Figure 2As shown in Figure a, the mechanical properties of the hydrogel are highly dependent on the AM / AA ratio. Increasing the AA content introduces more carboxyl groups, thereby enhancing the dynamic hydrogen bonding and ionic interactions within the network, thus improving energy dissipation and toughness. In contrast, AM mainly provides elastic segments to maintain the deformability of the material. When the AM content is too high, the network lacks sufficient dynamic interaction sites, resulting in insufficient strength and susceptibility to early fracture; while a high AA content can further improve the elongation at break, excessively dense dynamic interactions may disrupt the structural balance. Among all formulations, the hydrogel with an AM:AA ratio of 1:2 achieved the best balance between tensile strength and ductility (Young's modulus 368.5 kPa, tensile elongation 340%), indicating that its elastic support and dynamic energy dissipation are optimally coordinated.

[0067] Cyclic tensile behavior further reveals the energy dissipation mechanism and fatigue resistance of ion-hydrogel materials. For example... Figure 2 b. During repeated stretching to 100% strain, the load-unloading curves exhibit moderate and highly repeatable hysteresis loops. The presence of hysteresis indicates that reversible hydrogen bonding, ionic interactions, and the physical cross-linking structure mediated by nanoclay can all serve as sacrificial interaction mechanisms to effectively dissipate mechanical energy during deformation. The energy dissipation analysis results further highlight the excellent toughness of the ion-hydrogel material.

[0068] Figure 3 Figure a shows the Nyquist plots of hydrogels with different concentrations of EmimCl and ionized hydrogels. Adding EmimCl reduces interfacial resistance and increases conductivity. However, as the Emim+ concentration gradually increases, the conductivity of the gel decreases. We further investigated the effect of these two ion concentrations on conductivity. Figure 3 As shown in b, in the low concentration range, the conductivity of the EmimCl-doped ionic hydrogels increases with increasing concentration, reaching a peak of 2.06 S / m at 1 M, and then slightly decreasing to 1.70 S / m at 1.5 M.

[0069] In this embodiment, the conductivity of ion-hydrogels at room temperature, 0°C, and -20°C was measured. Figure 4 Although the conductivity at -20°C is only 50% of the previous value, the hydrogel still retains measurable conductivity, indicating that the ion transport channels are not completely interrupted under low-temperature conditions. This retention of conductivity suggests improved freeze resistance and provides a material basis for potential cryogenic sensing applications.

[0070] like Figure 5As shown, the water retention performance of the dual-ion hydrogel is significantly improved compared to the system lacking ion regulation. The combined effects of zwitterionic hydration, Li⁺-water coordination, and ionic liquid incorporation reduce the free water evaporation rate and stabilize the internal hydration state. After 10 days of exposure at room temperature, the weight retention rate of the dual-ion hydrogel is still close to 90%, significantly higher than the control group (67%) without dual-ion regulation. Its significantly improved dehydration resistance (improvement exceeding 20%) also ensures that the material maintains its flexibility and conductivity over time.

[0071] To further elucidate the role of Laponite in self-healing behavior, this embodiment tested the effect of Laponite content on the viscoelastic properties of the hydrogel. Figure 6 The storage modulus (G') increases monotonically with increasing Laponite content, indicating enhanced network rigidity and the formation of an effective physical cross-linked structure. This phenomenon suggests that Laponite acts as a multifunctional dynamic linker in the polymer matrix, forming reversible electrostatic and coordination interactions with carboxyl and amide groups. Notably, the loss modulus (G) exhibits a non-monotonic trend, reflecting the balance between network elasticity and energy dissipation capacity. Figure 6 The illustrations further demonstrate that the hydrogel can autonomously restore its structural integrity after being cut.

[0072] This embodiment also provides a smart insole and gait recognition method based on an environmentally adaptable ion hydrogel sensor array, which solves the problems of existing plantar pressure monitoring insoles that can only collect single-point or static pressure signals, have difficulty obtaining multi-region spatiotemporal pressure changes during continuous gait, and lack the ability to automatically analyze and recognize multi-channel plantar pressure signals.

[0073] like Figure 7 This invention forms a multi-channel plantar pressure sensing array by distributing multiple hydrogel pressure sensing units in the heel, arch, metatarsal, and forefoot areas of the insole. A voltage divider circuit converts the resistance change of the hydrogel under pressure deformation into a voltage signal. A microcontroller performs polling, baseline correction, normalization, and data encapsulation on the multi-channel voltage signals. A wireless communication module transmits the plantar pressure signals to a mobile terminal. Finally, a gait recognition model analyzes the multi-channel pressure time-series signals to output gait recognition results for normal gait, Parkinson's-like gait, diabetic foot pressure abnormality patterns, inversion, and eversion.

[0074] This invention not only enables the acquisition of plantar pressure signals, but also automatically identifies gait status based on characteristics such as pressure peak values ​​in different plantar regions, pressure center shift sequence, left-right or inner-outer pressure offset, and pressure phase difference between the forefoot and heel, thereby improving the application value of smart insoles in sports monitoring, rehabilitation assessment, and health management.

[0075] Each hydrogel pressure sensing unit acts as a pressure-sensitive variable resistor. When a person stands or walks, different areas of the sole exert pressure of varying magnitude and timing on the insole. When this pressure is applied to the corresponding hydrogel pressure sensing unit, the hydrogel undergoes compression deformation, altering the ion transport path within the hydrogel and changing the contact state between the hydrogel and the upper and lower electrodes, thus causing a change in the hydrogel's equivalent resistance. Figure 8 The detailed sensing circuitry and duty cycle acquisition scheme are demonstrated: In this configuration, each hydrogel sensor operates as a variable resistor, and its resistance changes are converted into corresponding voltage signals via a voltage divider circuit. These analog signals are then digitized. A precise excitation voltage of 3.3V is applied during ADC sampling, and the sensing circuitry is activated only briefly (approximately 100 μs) at a sampling frequency of 8Hz.

[0076] This implementation develops a fully integrated wireless gait monitoring platform capable of real-time data acquisition, wireless transmission, and visualization via a smartphone. The platform combines a distributed hydrogel sensor array with a compact printed circuit board for signal processing and digitization, a Bluetooth communication module supporting wireless data streaming, and a mobile application for spatiotemporal reconstruction of plantar pressure patterns. Eight hydrogel units embedded in anatomically relevant plantar regions serve as front-end sensors; each sensing channel modulates the ion conduction path through compression, converting local plantar load into a change in resistance. Figure 9 and 10 This diagram illustrates the system's operation and architecture. To achieve stable readings and reduce circuit complexity during dynamic gait, the sensor interface employs a voltage divider scheme to convert resistance changes into a measurable voltage output. The processed signal is sampled by an analog-to-digital converter (ADC) and then processed by an STM32 microcontroller unit (MCU), including channel polling, signal normalization, and data packet formatting. A flash memory module is also integrated to meet local buffering needs when necessary, ensuring continuous recording and minimizing data loss due to transient wireless interruptions. A key feature of this system is its power supply strategy implemented using a mobile device (smartphone). The platform utilizes Qi protocol wireless charging technology, eliminating the need for battery power. The smartphone powers the wearable module via reverse wireless charging, providing a stable power source and enhancing usability. In this configuration, the phone simultaneously serves as both a computing interface and a power supply, significantly reducing hardware complexity.

[0077] The gait recognition method of this embodiment includes the following steps:

[0078] Step B1: Multi-channel plantar pressure signal acquisition:

[0079] The smart insole is worn on the sole of the tester's foot. Eight hydrogel pressure sensing units, designated CH1 to CH8, are positioned as follows: CH1 and CH2 are located in the metatarsal head region of the forefoot; CH3 and CH5 are located in the middle of the metatarsals or the forefoot arch region; CH4 and CH8 are located in the medial and lateral heel regions; and CH6 and CH7 are located in the lateral arch or midfoot region. During standing or walking, different pressures are generated in different areas of the sole, and the multiple hydrogel pressure sensing units output corresponding resistance change signals. A microcontroller polls and collects data from each channel according to a preset sampling frequency. More preferably, the sampling frequency is 8Hz to 100Hz. Within each sampling cycle, the microcontroller sequentially reads the voltage value of each channel and converts it into a digital signal.

[0080] Step B2: Baseline Correction

[0081] Before formal identification, baseline signals for each channel are acquired under no-load or stationary standing conditions. The real-time signals of each channel during walking are compared with the corresponding baseline signals to obtain relative change signals. This step aims to eliminate baseline shifts caused by differences in initial resistance of different hydrogel sensing units, variations in insole wearing conditions, and individual weight differences, ensuring that subsequent identification relies primarily on dynamic gait characteristics rather than a single absolute voltage value.

[0082] Step B3: Signal filtering and smoothing:

[0083] The acquired multi-channel signals are filtered and smoothed to remove transient noise and abnormal spikes. Filtering methods include moving average filtering, median filtering, low-pass filtering, wavelet filtering, or combinations thereof. A sliding window is used to average the continuous sampling points of each channel, with a window length of 3–100 sampling points. Further averaging or feature summarization of 40–50 continuous sampling points yields a feature vector representing the plantar pressure distribution state within that time window.

[0084] Step B4: Gait period segmentation:

[0085] Based on the changes in peak pressure in the heel or forefoot channels, continuous plantar pressure signals are segmented into gait cycles. A complete gait cycle includes the heel strike phase, full foot support phase, heel lift-off phase, and forefoot push-off phase. During the heel strike phase, the signal in the heel channel initially increases; during the full foot support phase, pressure responses are observed in the heel, arch, and metatarsal channels; during the heel lift-off phase, the heel channel signal decreases, while the signals in the forefoot and metatarsal channels increase; during the forefoot push-off phase, the forefoot channel shows a significant peak followed by a decrease. By analyzing the order and phase difference of the peak values ​​in each channel, the dynamic process of the pressure center shifting from the heel to the forefoot can be obtained.

[0086] Step B5: Construct multi-channel gait features:

[0087] A gait feature matrix is ​​constructed from multi-channel signals within a gait cycle or a time window. With 8 sensor channels and 40 sampling points within the time window, each gait sample can be represented as an 8×40 two-dimensional matrix.

[0088] X = [S1(t), S2(t), …, SC(t)]

[0089] Where C is the number of sensing channels, which is 8; T is the number of sampling points within the time window, which is 40; and Si(t) is the normalized pressure signal of the i-th channel at time t.

[0090] Step B6: Gait recognition model classification:

[0091] The multi-channel gait features obtained in step B5 are input into the gait recognition model. The gait recognition model can be a convolutional neural network, a long short-term memory network, a support vector machine, a random forest, a K-nearest neighbor model, a Transformer model, or a combination thereof. A convolutional neural network is used to classify and recognize the multi-channel plantar pressure signals.

[0092] In one implementation, the convolutional neural network includes an input layer, a first convolutional layer, a first pooling layer, a second convolutional layer, a second pooling layer, a fully connected layer, and an output layer. The input layer receives 8-channel plantar pressure features; the convolutional layers are used to extract spatial correlations between different plantar regions and local temporal features between different sampling times; the pooling layers are used to reduce feature dimensionality and improve model robustness; the fully connected layer is used to map the extracted features to different gait categories; and the output layer outputs the probability of each gait category.

[0093] Gait categories include normal gait, Parkinsonian gait, diabetic foot stress abnormality pattern, inversion gait, and eversion gait. The final gait recognition result is determined based on the category with the highest probability in the output layer, and the corresponding recognition confidence score is also output.

[0094] Step B7: Mobile terminal display and feedback:

[0095] After receiving the gait recognition results, the mobile terminal displays a multi-channel pressure curve, a plantar pressure distribution map, gait category, and recognition confidence level. When an abnormal gait or localized high pressure state of the plantar surface is identified, the mobile terminal can output prompt information, such as "localized high pressure in the forefoot," "high pressure on the medial side of the foot," "high pressure on the lateral side of the foot," or "discontinuous pressure transfer from the heel to the forefoot." In a preferred embodiment, the mobile terminal can also record the recognition results of multiple consecutive gait cycles to generate a gait assessment report for use in sports training, rehabilitation assessment, or long-term health management.

[0096] In summary, this invention discloses an environmentally adaptable ionic hydrogel, a smart insole based on this hydrogel's sensor array, and a corresponding gait recognition method. The fabricated ionic hydrogel achieves excellent antifreeze, dehydration resistance, and self-healing properties through dual-ion synergistic design, while maintaining good mechanical toughness and electrical conductivity stability, making it adaptable to various complex environmental conditions such as low temperature and dryness. The multi-channel plantar pressure sensing smart insole constructed using this ionic hydrogel can continuously collect pressure time-series signals from different plantar areas during walking. Through a built-in signal processing flow and a deep learning gait recognition model, it can automatically identify normal gait and various common abnormal gait patterns, outputting intuitive recognition results and prompts. It can be widely applied to daily exercise monitoring, chronic disease rehabilitation assessment, and long-term foot health management, providing a new material and solution with stable performance and strong adaptability for wearable plantar pressure monitoring and gait analysis.

[0097] It should be emphasized that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any way. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention shall still fall within the scope of the technical solution of the present invention.

Claims

1. An ionic hydrogel, characterized in that, The product comprises the following components in parts by weight: 26-30 parts sulfobetaine methacrylate, 7-8 parts acrylamide, 13-17 parts acrylic acid, 35-42 parts deionized water, 7-8 parts Laponite, 2.2-2.8 parts lithium chloride, 7-8 parts 1-ethyl-3-methylimidazolium chloride, 0.45-0.55 parts N,N'-methylenebisacrylamide, 0.22-0.28 parts ammonium persulfate, and 0.22-0.28 parts tetramethylethylenediamine; The ionic hydrogel forms a multi-layered dynamic network with chemical cross-linking, hydrogen bonding, ion interaction, coordination, and physical cross-linking of nanofillers, and possesses high adhesion, self-healing, antifreeze, water retention, stable conductivity, and mechanical properties resistant to cyclic compression and shearing.

2. The ionic hydrogel according to claim 1, characterized in that, The ionic hydrogel has a maximum adhesion strength of 28.6 kPa, an elongation at break of 340%, a Young's modulus of 368.5 kPa, and a peak ionic conductivity of 2.06 S / m.

3. The ionic hydrogel according to claim 1, characterized in that, The functions of each component of the ion hydrogel are as follows: The sulfobetaine methacrylate forms an amphoteric hydration layer, which improves water retention, resistance to drying, and ion transport stability. The acrylamide provides flexible segments that enhance tensile strength; The acrylic acid incorporates carboxyl groups, which enhances adhesion, self-healing, and energy dissipation capabilities. The Laponite forms reversible physical cross-linking points, enhancing mechanical toughness and fatigue resistance; The lithium chloride and the 1-ethyl-3-methylimidazolium chloride duo ion synergistically achieve stable conductivity over a wide temperature range and low-temperature freeze resistance; The N,N'-methylenebisacrylamide forms a chemically cross-linked framework; The ammonium persulfate and the tetramethylethylenediamine initiate the free radical polymerization of the ternary monomers.

4. A method for preparing the ionic hydrogel according to any one of claims 1 to 3, characterized in that, The preparation method of the ion hydrogel includes the following steps: Step A1: Add Laponite powder to deionized water and ultrasonically disperse at room temperature for no less than 30 minutes to obtain a uniform nanofiller dispersion. Step A2: Add sulfobetaine methacrylate, acrylamide, acrylic acid, 1-ethyl-3-methylimidazolium chloride, lithium chloride, N,N'-methylenebisacrylamide, and tetramethylethylenediamine to the dispersion in sequence, stir at room temperature for no less than 30 minutes, and remove air bubbles from the precursor solution under vacuum. Step A3: Finally, add ammonium persulfate and stir until homogeneous. Pour the mixture into a mold and seal it. Let it stand at room temperature for at least 48 hours to complete the polymerization and obtain a ternary network ionic hydrogel.

5. A smart insole based on an ion hydrogel sensor array, characterized in that, The smart insole is composed of several sensing units made of ion hydrogel as described in any one of claims 1 to 3, and the smart insole includes: an insole base, a multi-channel pressure sensing array, a pressure divider acquisition circuit, a microcontroller, and a wireless transmission module; The sensing units are arranged in sections in all key pressure areas of the sole, including the heel area, arch area, metatarsal area, and forefoot area of ​​the insole. Each sensing unit is connected to the voltage divider acquisition circuit as a variable resistor. When subjected to pressure deformation, the resistance change is converted into a standardized voltage signal. The microcontroller polls and acquires multi-channel voltage signals and completes baseline correction and normalization preprocessing. The plantar pressure time series data is then sent to the mobile terminal via the wireless transmission module.

6. The smart insole based on an ion hydrogel sensor array according to claim 5, characterized in that, The voltage divider acquisition circuit is powered on only during the ADC sampling instant, with a sampling frequency of 8Hz and a single sampling power-on duration of 100μs; the smart insole is equipped with a wireless charging module, which is powered by reverse Qi wireless charging from the mobile terminal; the mobile terminal can display the plantar pressure curve and two-dimensional pressure-thermal distribution map in real time.

7. The smart insole based on an ion hydrogel sensor array according to claim 5, characterized in that, The ion hydrogel sensing unit is matched and assembled with flexible FPC, conductive fabric, and silver paste flexible electrode; the wireless transmission module adopts Bluetooth, Bluetooth Low Energy or Wi-Fi; the microcontroller is an STM32, ESP32 or nRF series single-chip microcomputer.

8. A gait recognition method based on the smart insole according to any one of claims 5 to 7, characterized in that, Includes the following steps: Step B1, Multi-channel pressure signal acquisition: Wear smart insoles, the microcontroller synchronously acquires the voltage timing signals output by the hydrogel sensing units of each foot zone; Step B2, Baseline Correction: Acquire baseline signals under no-load and stationary standing conditions, and perform a difference calculation between the real-time signal and the baseline to eliminate baseline offset caused by individual weight, initial sensor resistance, and wearing deviation; Step B3, Signal preprocessing: Noise is removed by moving average, median or low-pass filtering, and the sampling points are aggregated by sliding window to generate a single time window pressure feature vector; Step B4, Gait Cycle Segmentation: Based on the peak pressure sequence of the heel and forefoot channels, divide the complete gait cycle into heel strike, full foot support, heel lift-off, and forefoot push-off. Step B5: Constructing the gait feature matrix: Construct a two-dimensional feature matrix from the 8-channel time-series signal within a single cycle or window as the input for gait recognition; Step B6, Model Classification and Recognition: Input the feature matrix into the gait recognition model and output the gait category and recognition confidence; the gait category includes normal gait, Parkinson's abnormal gait, diabetic foot high-pressure pattern, and inversion and eversion gait; Step B7, Terminal Visual Feedback: The mobile terminal displays the pressure distribution and gait recognition results, outputs risk warnings when abnormal gait is identified, and generates a long-term gait assessment report.

9. The gait recognition method according to claim 8, characterized in that, The gait recognition model is any one or a combination of convolutional neural networks, long short-term memory networks, Transformers, support vector machines, and random forests; the convolutional neural network includes an input layer, two convolutional pooling layers, a fully connected layer, and a classification output layer, used to extract the spatial pressure correlation and temporal dynamic features of the foot.

Images