Multi-channel electromyographic strain bimodal electronic skin and neural signal decomposition method

Abstract

The invention provides a multichannel electromyographic strain bimodal electronic skin and a neural signal decomposition method, and belongs to the technical field of sensors. The multi-channel electromyographic strain bimodal electronic skin comprises six layers of composite structures, and the six layers of composite structures are a stretchable flexible substrate layer, a multi-channel strain sensing layer, a dielectric isolation layer, a stretchable circuit, a multi-channel electromyographic sensing layer and a viscous packaging layer in sequence. The device has the flexible stretchable characteristic, reliable contact between the electronic skin and the skin can be guaranteed during large-amplitude motion, dual-mode detection of human body electromyographic signal-resistance type stress changes can be achieved, a motion-artifact-free fusion signal processing algorithm based on multi-channel dual-mode signals is further provided, and multi-mode fusion signals without motion artifacts are obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of sensors, and in particular, to a multi-channel electromyogram strain dual-mode electronic skin and a method for decomposing nerve signals. Background Art

[0002] Surface electromyogram (sEMG) is a bioelectrical signal guided and recorded by electrodes placed on the surface of muscles. It reflects the electrophysiological characteristics of the neuromuscular system during activity, is closely related to the activity and functional state of muscles, and is a key tool for tracking neuromuscular activity in the fields of health monitoring and disease treatment.

[0003] As the core component for measuring sEMG signals, the development of electromyogram electrodes is crucial. Traditional sEMG acquisition devices usually use commercial silver / silver chloride electrodes. However, such electrodes have many limitations and cannot achieve long-term high-fidelity sEMG signal acquisition. Moreover, due to their non-stretchable characteristics, when the skin deforms due to human activities, interface delamination easily occurs between the electrodes and the skin due to the significant difference in modulus, making it difficult to meet the high requirements of practical applications. In recent years, electromyogram electrodes have gradually developed towards array and flexibility, but even so, during large-deformation movements, existing electromyogram electrodes are still easily interfered by motion artifacts, affecting the accuracy of signals. Patent CN118319315A discloses a honeycomb multi-channel flexible electromyogram electrode, in which the electrode is made of PI and metal, and its honeycomb geometric design endows it with certain planar bending and stretching capabilities, but it cannot meet the requirements of large-deformation movements, and the electrode has no adhesiveness and poor adhesion to the skin. In addition, this electromyogram electrode can only collect single-mode signals of electromyogram and cannot directly provide skin strain information caused by muscle contraction. Relying solely on electromyogram signals often has problems such as insufficient information, being easily interfered by motion artifacts, or inaccurate intention recognition, which limits the depth and breadth of its application.

[0004] In the cutting-edge field of flexible electronic skin, the electrophysiological and strain dual-mode sensing technology is of great significance. It can simultaneously capture human bioelectrical signals and mechanical deformation information, providing strong data support for the precise control of intelligent prosthetics and the scientific evaluation of sports rehabilitation. The present invention focuses on the deficiencies of existing electronic skin in stretchability and dual-mode integration, and proposes an innovative modular hierarchical preparation strategy. Through key process breakthroughs such as the construction of a stretchable flexible substrate, the step-by-step preparation of a multi-channel electromyogram / strain sensing layer, and the optimization of a sticky encapsulation layer, the independent transmission of multi-channel dual-mode signals is successfully achieved, and reliable connection with an external acquisition device is realized through a stretchable interconnection circuit and a terminal pluggable interface, achieving low-impedance connection with the external acquisition device, opening up a new path for the dual-mode integration of flexible electronic skin, and strongly promoting the development of fields such as intelligent prosthetics and sports rehabilitation. Summary of the Invention

[0005] To solve the problem that traditional high-density myoelectric e-skins can only detect single signals, the present invention invented a multi-channel myoelectric strain dual-modal e-skin and a neural signal decomposition method, aiming to solve the dual-modal detection of physiological electrical signals and muscle strain of the wearer.

[0006] To achieve the above object, the technical solutions adopted by the present invention are as follows:

[0007] A multi-channel myoelectric strain dual-modal e-skin, the multi-channel myoelectric strain dual-modal e-skin includes a six-layer composite structure, and the six-layer composite structure is in turn: a stretchable flexible base layer 7, a multi-channel strain sensing layer 5, a dielectric isolation layer 4, a stretchable circuit 3, a multi-channel myoelectric sensing layer 2, and a viscous encapsulation layer 1.

[0008] The stretchable flexible base layer 7 is composed of a highly crosslinked polydimethylsiloxane elastomer and is used to carry the other five layers.

[0009] Further, the thickness of the stretchable flexible base layer 7 is 50μm - 100μm, the elongation at break is 100% - 120%, and the Young's modulus is the same as that of human skin, which can fit the human skin and effectively prevent interface peeling.

[0010] Further, the preparation method of the highly crosslinked polydimethylsiloxane elastomer: Mix the polydimethylsiloxane (PDMS) prepolymer with the curing agent, fully stir and then perform defoaming treatment, and then spin-coat on the surface of the polyethylene terephthalate (PET) support layer, and obtain the highly crosslinked polydimethylsiloxane elastomer after thermal curing treatment and peeling; Further, the mass ratio of the polydimethylsiloxane (PDMS) prepolymer to the curing agent is 10 - 20:1, the temperature of the thermal curing treatment is 80°C - 100°C, and the time of the thermal curing treatment is 0.5 hour - 1 hour.

[0011] Preferably: Both the polydimethylsiloxane (PDMS) prepolymer and the curing agent are Dow Corning SYLGARD184.

[0012] Further, the surface of the stretchable flexible base layer 7 needs to be treated with oxygen plasma to turn the hydrophobic surface of the stretchable flexible base layer 7 into hydrophilic, thereby effectively enhancing the stability of the conductive ink on the stretchable flexible base layer 7, making the conductive ink adhere more firmly, and at the same time can also well prevent the occurrence of interlayer separation when the six-layer composite structure is stacked, ensuring the stability of the multi-layer structure.

[0013] The multi-channel strain sensing layer 5 is used to convert the skin deformation signal into a resistance signal and includes a stretchable interconnection circuit 6 and a strain sensitive unit array.

[0014] The stretchable interconnect circuit 6 is a wire structure formed by dispensing and printing a conductive ink with a liquid metal content of 60 wt% - 80 wt%.

[0015] Furthermore, the thickness of the stretchable interconnect circuit 6 is 30 μm - 50 μm; the line width of the wire structure is 50 μm - 200 μm, the resistance change rate is 2% - 10% at a 40% tensile strain rate, and the resistance change rate is 2% - 5% under 20% cyclic stretching (strain rate 10% / s), and the fatigue life is at least 10 4 times.

[0016] The strain - sensitive unit array is composed of identical strain - sensitive units, and the strain - sensitive units are formed by dispensing and printing a conductive ink with a liquid metal content of 20 wt% - 40 wt%; the strain - sensitive units are slender rectangles, and both ends in the length direction of each strain - sensitive unit are connected to the plug - and - play interface at the end of the electrode through the stretchable interconnect circuit 6.

[0017] Furthermore, the strain - sensitive unit array contains 8 - 64 strain - sensitive units, the width of each strain - sensitive unit is 50 μm - 100 μm, and the thickness is 30 μm - 50 μm; the strain - sensitive units have dynamic detection performance, can detect a planar displacement of 0.1 mm - 5 mm, the resolution is 10 μm - 50 μm, the sensitivity is at least 10 Ω / mm, and is synchronized with the myoelectric signal at the 50 - millisecond level.

[0018] Furthermore, the conductive ink in the stretchable interconnect circuit 6 and the strain - sensitive unit array is a mixed system of conductive silver paste and gallium - indium liquid metal alloy. The conductive ink with a liquid metal content of 60 wt% - 80 wt% means that the proportion of gallium - indium liquid metal alloy in the conductive ink is 60 wt% - 80 wt%; the proportion of gallium - indium liquid metal alloy in the conductive ink with a liquid metal content of 20 wt% - 40 wt% is 60 wt% - 80 wt%; the mass ratio of gallium to indium in the used gallium - indium liquid metal alloy is 3 - 4:1.

[0019] The dielectric isolation layer 4 is used to achieve electrical isolation and mechanical decoupling between the multi - channel myoelectric sensing layer 2 and the multi - channel strain sensing layer 5, and is composed of a highly cross - linked polydimethylsiloxane elastomer.

[0020] Furthermore, the thickness of the dielectric isolation layer 4 is 20 μm - 30 μm.

[0021] Furthermore, the preparation method of the highly cross - linked polydimethylsiloxane elastomer is the same as that of the stretchable flexible base layer 7.

[0022] The manufacturing process and materials used for the stretchable circuit 3 are the same as those of the stretchable interconnection circuit 6 of the multi-channel strain sensing layer 5. However, the shape is different from that of the stretchable interconnection circuit 6 of the multi-channel strain sensing layer 5. The stretchable circuit 3 includes a circuit array and a wire structure that connects the circuit array to the pluggable interface at the electrode end.

[0023] Further, the thickness of the stretchable circuit 3 is 30μm - 50μm.

[0024] The multi-channel electromyogram sensing layer 2 is used to collect electromyogram signals on the skin surface and is an electrode array composed of the same electrode units; the electrode array is connected to the pluggable interface at the electrode end through the stretchable circuit 3 to achieve low-impedance connection with an external acquisition device.

[0025] Further, the preparation method of the electrode unit: Mix the polydimethylsiloxane (PDMS) prepolymer and the curing agent in a mass ratio of 30 - 50:1, add silver powder with a particle size of 1μm - 10μm to form a mixed solution, and the mass ratio of the silver powder to the polydimethylsiloxane (PDMS) prepolymer is 2.5 - 3.5:1; Dot-print the mixed solution to form the electrode unit.

[0026] Further, the spacing between the electrode units in the electrode array is 5mm - 10mm, the equivalent diameter of each electrode unit is 3mm - 5mm, the number of electrode units is 8 - 64, and the relationship between the skin contact impedance of the electrode unit and the AC frequency is that the contact impedance is 4.4kΩ - 5kΩ at 1kHz. The thickness of the electrode array is 50μm - 80μm.

[0027] Further, the spatial mapping relationship between the stretchable circuit 3 and the multi-channel electromyogram sensing layer 2: The size and position of the circuit array of the stretchable circuit 3 are the same as those of the electrode array in the multi-channel electromyogram sensing layer 2.

[0028] Further, the spatial mapping relationship between the multi-channel electromyogram sensing layer 2 and the multi-channel strain sensing layer 5: The centers at both ends in the length direction of each strain sensitive unit in the multi-channel strain sensing layer 5 correspond to the center positions of two adjacent electrode units in the multi-channel electromyogram sensing layer 2.

[0029] The viscous encapsulation layer 1 is used to bond the multi-channel electromyogram and strain dual-mode electronic skin to the skin and is composed of a low-crosslinked polydimethylsiloxane elastomer.

[0030] Further, the adhesion strength of the viscous encapsulation layer 1 is 0.6N / cm - 0.7N / cm, and the thickness is 20μm - 30μm;

[0031] Furthermore, a method for preparing a low-crosslinked polydimethylsiloxane elastomer: Mix a polydimethylsiloxane (PDMS) prepolymer with a curing agent, perform defoaming treatment after sufficient stirring, and spin-coat it on the multi-channel electromyography sensing layer 2 using the spin-coating method, and obtain it through heat curing treatment; Preferably: the mass ratio of the polydimethylsiloxane (PDMS) prepolymer to the curing agent is 30-50:1. The temperature of the heat curing treatment is 80°C - 100°C, and the time of the heat curing treatment is 0.5 hour - 1 hour.

[0032] The above-mentioned method for decomposing nerve signals of a multi-channel electromyography strain dual-modal electronic skin is a motion-artifact-free fusion signal processing method based on multi-channel dual-modal signals, specifically as follows:

[0033] The multi-channel electromyography strain dual-modal electronic skin is pasted on the human skin and connected to a multi-channel electromyography acquisition device and a multi-channel resistance signal acquisition device through a pluggable interface at the electrode end. The sampling frequency of the electromyography signal is 1kHz - 2kHz, and the sampling frequency of the resistance signal is 200Hz - 400Hz. Then divide the resistance signal by the sensitivity to obtain the strain signal of the skin, and finally obtain the electromyography-strain dual-modal signal of human muscle activity, and process the electromyography-strain dual-modal signal of human muscle activity according to the following steps:

[0034] 1) Signal preprocessing: Perform band-pass filtering and power frequency notch filtering on the electromyography signal; perform Kalman filtering on the strain signal.

[0035] 2) Motor unit localization: Use a blind source separation algorithm to decompose the preprocessed electromyography signal to obtain the motor unit activity sequence, and use the spike-triggered averaging method to analyze the motor unit activity sequence to obtain the initial motor unit position and motor unit waveform, achieving the purpose of motor unit localization;

[0036] 3) Motion artifact elimination: According to the preprocessed strain signal, construct a motion artifact spatial distribution map, and use the motion artifact spatial distribution map to correct the motor unit position, realizing the elimination of the motion noise component related to the strain signal in the electromyography signal, and further eliminating the motion artifact.

[0037] 4) Multi-modal fusion output: Align the motor unit activity sequence, motor unit waveform, corrected motor unit position and strain signal in the time domain, and output a multi-modal fusion signal composed of the electromyography signal, motor unit activity sequence, corrected motor unit position, motor unit waveform, and strain signal.

[0038] Compared with the prior art, the beneficial effects of the present invention are:

[0039] 1) The present invention has flexible and stretchable characteristics, and at the same time, the encapsulation layer has adhesiveness, which can not only better fit the skin but also ensure reliable contact between the electrodes and the skin during large-amplitude movements.

[0040] 2) The present invention can not only detect multi-channel human myoelectric signals but also achieve dual-mode detection of human myoelectric signals-resistive stress changes. In addition, the present invention also includes a multi-channel dual-mode signal motion-artifact-free fusion signal processing algorithm, which can obtain motion-artifact-free multi-modal fusion signals while realizing dual-mode detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings forming a part of this application are used to provide a further understanding of this application. The schematic embodiments and descriptions thereof of this application are used to explain this application and do not constitute an improper limitation of this application.

[0042] Figure 1 It is a schematic diagram of the overall structure of a multi-channel myoelectric strain dual-mode electronic skin according to an embodiment;

[0043] Figure 2 It is a schematic diagram of the hierarchical structure of a multi-channel myoelectric strain dual-mode electronic skin according to an embodiment;

[0044] Figure 3 It is a curve showing the change of the resistance value of the wire structure of the stretchable interconnection circuit of a multi-channel myoelectric strain dual-mode electronic skin according to an embodiment with the stretching rate;

[0045] Figure 4 It is a curve showing the change of the resistance value of the strain-sensitive unit of a multi-channel myoelectric strain dual-mode electronic skin according to an embodiment with the stretching rate;

[0046] Figure 5 It is a test diagram of the skin contact impedance of the electrode unit of a multi-channel myoelectric strain dual-mode electronic skin according to an embodiment;

[0047] Figure 6 It is a test diagram of the adhesion strength of the adhesive encapsulation layer of a multi-channel myoelectric strain dual-mode electronic skin according to an embodiment;

[0048] Figure 7 It is a flowchart of the multi-modal biological signal data processing method of a multi-channel myoelectric strain dual-mode electronic skin according to an embodiment;

[0049] Figure 8 It is the myoelectric signal collected by 16 channels according to an embodiment;

[0050] Figure 9 It is the strain signal collected by 14 channels according to an embodiment;

[0051] Figure 10 It is the preprocessed myoelectric signal according to an embodiment;

[0052] Figure 11 The preprocessed strain signal of the embodiment;

[0053] Figure 12 All the motor unit activity sequences obtained by decomposing using the blind source separation algorithm of the embodiment;

[0054] Figure 13 The motor unit waveform of one of the motor units obtained by using the spike-triggered averaging method of the embodiment;

[0055] Figure 14 The initial motor unit position of one of the motor units obtained by using the spike-triggered averaging method of the embodiment;

[0056] Figure 15 The spatial distribution diagram of motion artifacts of the embodiment;

[0057] Figure 16 For Figure 14 The corrected motor unit position of the motor unit.

[0058] In the figure: 1 Viscous encapsulation layer, 2 Multi-channel electromyography sensing layer, 3 Stretchable circuit, 4 Dielectric isolation layer, 5 Multi-channel strain sensing layer, 6 Stretchable interconnection circuit, 7 Stretchable flexible substrate layer. Detailed implementation manners

[0059] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0060] It should be noted that the following detailed description is illustrative and is intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present application belongs.

[0061] Embodiment 1:

[0062] In this embodiment, a multi-channel electromyography-strain dual-modal electronic skin is disclosed. As shown in Figure 1 and Figure 2 , the multi-channel electromyography-strain dual-modal electronic skin includes a six-layer composite structure arranged in sequence from the bottom layer to the top layer. The six-layer composite structure is in sequence: a stretchable flexible substrate layer 7, a multi-channel strain sensing layer 5, a dielectric isolation layer 4, a stretchable circuit 3, a multi-channel electromyography sensing layer 2, and a viscous encapsulation layer 1. The preparation steps are as follows:

[0063] Step 1. Preparation of the stretchable flexible substrate layer 7:

[0064] Mix the Dow Corning SYLGARD 184 polydimethylsiloxane (PDMS) prepolymer and the curing agent in a mass ratio of 10:1. After thorough stirring, conduct defoaming treatment, and then spin-coat it on the surface of the polyethylene terephthalate (PET) support layer at a spin-coating speed of 1000 rpm. Conduct thermal curing treatment at 80 °C for 0.5 hours to obtain a stretchable flexible base layer 7 with a thickness of 100 μm, an elongation at break of 100%, and a Young's modulus of 2.5 MPa.

[0065] Step 2. Preparation of the multi-channel strain sensing layer 5:

[0066] S2.1. Mix the conductive silver paste and the gallium-indium liquid metal alloy in a mass ratio of 3:7, where the mass ratio of gallium to indium in the gallium-indium liquid metal alloy is 75.5:24.5. After mixing, mechanically stir at a speed of 500 rpm for 30 minutes to obtain the conductive ink for the stretchable interconnected circuit 6.

[0067] S2.2. Conduct oxygen plasma treatment on the surface of the stretchable flexible base layer 7 for 70 seconds with a radio frequency power of 30 w.

[0068] S2.3. Dot-print the stretchable interconnected circuit 6 on the stretchable flexible base layer 7. The line width of the wire structure of the stretchable interconnected circuit 6 is 200 μm, the resistance change rate is 10% at a 40% tensile strain rate, the resistance change rate is 5% under 20% cyclic stretching (strain rate 10% / s), and the fatigue life is 10 4 times. The curve of the resistance value of the wire structure of the stretchable interconnected circuit 6 changing with the stretching rate is as Figure 3 shown. It can be observed that the resistance change of the wire structure of the stretchable interconnected circuit 6 is extremely small when stretched and can be stretched by more than 45%, having excellent performance.

[0069] S2.4. Mix the conductive silver paste and the gallium-indium liquid metal alloy in a mass ratio of 6:4, where the mass ratio of gallium to indium in the gallium-indium liquid metal alloy is 75.5:24.5. After mixing, mechanically stir at a speed of 500 rpm for 30 minutes to obtain the conductive ink for the strain-sensitive unit.

[0070] S2.5. Use the conductive ink of the strain-sensitive unit to connect the ends of the stretchable interconnected circuit 6 pairwise by dot-printing to manufacture a strain-sensitive unit array composed of 14 7×2 strain-sensitive units, thereby obtaining the multi-channel strain sensing layer 5; the strain-sensitive unit is in the shape of an elongated rectangle with a width of 100 μm and a thickness of 50 μm; the strain-sensitive unit has dynamic detection performance and can detect a planar displacement of 0.1 mm - 5 mm with a resolution of 50 μm and a sensitivity of 11.6 Ω / mm. Figure 4Shows the change of the resistance value of the above-mentioned strain-sensitive unit with the stretching rate. It can be clearly observed that its resistance value changes greatly under continuous stretching, demonstrating the extremely high sensitivity of the multi-channel strain sensing layer 5 to tensile strain.

[0071] Step 3. Preparation of the dielectric isolation layer 4, stretchable circuit 3, and multi-channel electromyogram sensing layer 2:

[0072] S3.1. Mix the Dow Corning SYLGARD184 polydimethylsiloxane (PDMS) prepolymer and the curing agent at a mass ratio of 10:1, stir well, and then perform defoaming treatment. Spin-coat it on the surface of the multi-channel strain sensing layer 5 at a speed of 3000 rpm. When spin-coating, use a silicone mask to protect the connection position between the stretchable interconnect circuit 6 and the electrode terminal pluggable interface. Cure it thermally at 80 °C for 0.5 hours to obtain the dielectric isolation layer 4 with a thickness of 30 μm.

[0073] S3.2. Perform oxygen plasma treatment on the surface of the dielectric isolation layer 4 for 70 seconds with a radio frequency power of 30 w.

[0074] S3.3. Prepare the stretchable circuit 3 using the same materials and processes as the stretchable interconnect circuit 6 of the multi-channel strain sensing layer 5, including a circuit array and a wire structure connecting the circuit array to the electrode terminal pluggable interface. The thickness of the stretchable circuit 3 is 50 μm.

[0075] S3.4. Multi-channel electromyogram sensing layer 2: Mix the polydimethylsiloxane (PDMS) prepolymer and the curing agent at a mass ratio of 50:1, add 5-μm flaky silver powder to form a mixed solution, and the mass ratio of the silver powder to the polydimethylsiloxane (PDMS) prepolymer is 3:1. Dot-print the mixed solution to form electrode units. The spacing between the electrode units is 10 mm, the shape of the electrode units is a circle with a diameter of 3 mm, and the number is 16, forming an 8×2 electrode array with a thickness of 80 μm. Figure 5 Shows the relationship between the impedance of the above-mentioned electrode unit and the AC frequency. The contact impedance is 4.4 kΩ at 1 kHz. Spatial mapping relationship between the stretchable circuit 3 and the multi-channel electromyogram sensing layer 2: The circuit array of the stretchable circuit 3 and the size and position of the electrode array in the multi-channel electromyogram sensing layer 2 are the same. Spatial mapping relationship between the multi-channel electromyogram sensing layer 2 and the multi-channel strain sensing layer 5: The centers at both ends of each strain-sensitive unit in the multi-channel strain sensing layer 5 in the length direction correspond to the center positions of two adjacent electrode units in the multi-channel electromyogram sensing layer 2.

[0076] Step 4. Preparation of the adhesive encapsulation layer 1:

[0077] S4.1. Use a silicone mask to protect the connection position between the stretchable circuit 3 and the electrode terminal pluggable interface and the electrode array.

[0078] S4.2. Mix the polydimethylsiloxane (PDMS) prepolymer and the curing agent in a mass ratio of 40:1, perform defoaming treatment after sufficient stirring, then spin-coat it on the multi-channel electromyogram sensing layer 2 at a spin-coating speed of 3000 rpm, and perform thermal curing treatment at 80 °C for 0.5 hours. Remove the silicone mask to obtain the viscous encapsulation layer 1 with a thickness of 30 μm. Figure 6 The peel force of the viscous encapsulation layer with a width of 24 mm and a length of 100 mm is shown, and its adhesion strength is calculated to be 0.71 N / cm.

[0079] Example 2:

[0080] Use silicone glue to stick the pluggable interface at the electrode end on the stretchable flexible substrate layer 7 of the multi-channel electromyogram strain dual-mode electronic skin prepared in Example 1, and connect it to the stretchable circuit 3 and the stretchable interconnection circuit 6.

[0081] Paste the multi-channel electromyogram strain dual-mode electronic skin on the human skin, connect it to a multi-channel electromyogram acquisition device (the multi-channel electromyogram acquisition device can use a wireless high-density electroencephalogram and electromyogram hybrid synchronous acquisition and monitoring system provided by Patent CN119679431A to acquire electromyogram signals) and a multi-channel resistance signal acquisition device, acquire electromyogram signals and resistance signals, set the sampling frequency of the electromyogram signals to 2 kHz, the sampling frequency of the resistance signals to 200 Hz, then divide the resistance signals by the sensitivity to obtain the strain signals of the skin, and finally obtain the electromyogram-strain dual-mode signals of human muscle activities. And process the electromyogram-strain dual-mode signals of human muscle activities according to the following steps:

[0082] 1) Signal preprocessing: Perform band-pass filtering (20 Hz - 500 Hz) and power frequency notch filtering on the electromyogram signals; perform Kalman filtering on the strain signals to extract the dynamic deformation amount of 0 Hz - 100 Hz.

[0083] 2) Motor unit localization: Use the blind source separation algorithm to decompose the preprocessed electromyogram signal to obtain the motor unit activity sequence. Use the spike-triggered averaging method to analyze the motor unit activity sequence to obtain the initial motor unit positions and motor unit waveforms, achieving the purpose of motor unit localization. Preferably, the blind source separation algorithm can recover the original independent signals from the mixed signals without any prior knowledge of the mixing process or source signals. The spike-triggered averaging method is a classical method for analyzing the relationship between neuron discharges (spikes) and external stimuli or behavioral signals. Its core idea is to extract the potential laws or features related to neuron activities by aligning and averaging specific signals before and after the occurrence of spikes (action potentials) multiple times. A motor unit is the "smallest functional unit" of muscle contraction, including a motor neuron and all the skeletal muscle fibers it innervates. The motor unit activity sequence contains the firing information of the motor unit, directly showing the moment when the neural command is issued and conveyed, characterized as the 0-1 firing sequence of the motor unit. 3) Motion artifact elimination: Based on the preprocessed strain signal, construct a motion artifact spatial distribution map, and use the motion artifact spatial distribution map to correct the motor unit positions, realizing the elimination of the motion noise components related to the strain signal in the electromyogram signal, and thus eliminating the motion artifacts.

[0084] 4) Multimodal fusion output: Align the motor unit activity sequence, motor unit waveforms, corrected motor unit positions, and strain signal in the time domain, and output a multimodal fusion signal composed of the electromyogram signal, motor unit activity sequence, corrected motor unit positions, motor unit waveforms, and strain signal.

[0085] Figures 7 to 16 Shows the flow chart of the neural signal decomposition method and the effect diagram during the process of the multi-channel electromyogram-strain dual-modal e-skin.

Claims

1. A multi-channel electromyogram strain dual-modal electronic skin, characterized in that, The multi-channel electromyogram strain dual-mode electronic skin includes a six-layer composite structure, which are, in sequence: a stretchable flexible substrate layer (7), a multi-channel strain sensing layer (5), a dielectric isolation layer (4), a stretchable circuit (3), a multi-channel electromyogram sensing layer (2), and a viscous encapsulation layer (1); The stretchable flexible substrate layer (7) is composed of a highly crosslinked polydimethylsiloxane elastomer and is used to carry the other five layers; The multi-channel strain sensing layer (5) is used to convert skin deformation signals into resistance signals and includes a stretchable interconnection circuit (6) and a strain-sensitive unit array; The stretchable interconnection circuit (6) is a wire structure formed by dispensing and printing a conductive ink with a liquid metal content of 60wt%-80wt%; The strain-sensitive unit array is composed of the same strain-sensitive units, and the strain-sensitive units are formed by dispensing and printing a conductive ink with a liquid metal content of 20wt%-40wt%; the strain-sensitive units are slender rectangles, and both ends in the length direction of each strain-sensitive unit are connected to the electrode end pluggable interface through the stretchable interconnection circuit (6); The dielectric isolation layer (4) is used to achieve electrical isolation and mechanical decoupling between the multi-channel electromyogram sensing layer (2) and the multi-channel strain sensing layer (5), and is composed of a highly crosslinked polydimethylsiloxane elastomer; The preparation process and materials used for the stretchable circuit (3) are the same as those of the stretchable interconnection circuit (6) of the multi-channel strain sensing layer (5); however, the shape is different from that of the stretchable interconnection circuit (6) of the multi-channel strain sensing layer (5). The stretchable circuit (3) includes a circuit array and a wire structure connecting the circuit array to the electrode end pluggable interface; The multi-channel electromyogram sensing layer (2) is used to collect electromyogram signals on the skin surface and is an electrode array composed of the same electrode units; the electrode array is connected to the electrode end pluggable interface through the stretchable circuit (3) to achieve a low-impedance connection with an external acquisition device; The viscous encapsulation layer (1) is used to attach the multi-channel electromyogram strain dual-mode electronic skin to the skin and is composed of a low-crosslinked polydimethylsiloxane elastomer.

2. The multi-channel electromyogram strain dual-modal electronic skin according to claim 1, wherein The thickness of the stretchable flexible substrate layer (7) is 50μm - 100μm, the elongation at break is 100% - 120%, and the Young's modulus is the same as that of human skin; the surface of the stretchable flexible substrate layer (7) needs to be treated with oxygen plasma to transform the hydrophobic surface of the stretchable flexible substrate layer (7) into a hydrophilic surface; Preparation method of the highly crosslinked polydimethylsiloxane elastomer: Mix the polydimethylsiloxane PDMS prepolymer and the curing agent, fully stir and then perform defoaming treatment, then spin-coat on the surface of a polyethylene terephthalate PET support layer, and peel off after heat curing treatment to obtain the highly crosslinked polydimethylsiloxane elastomer; the mass ratio of the polydimethylsiloxane PDMS prepolymer to the curing agent is 10 - 20:1, the temperature of the heat curing treatment is 80℃ - 100℃, and the time of the heat curing treatment is 0.5 hour - 1 hour.

3. A multi-channel electromyogram strain dual-modal electronic skin according to claim 1, characterized in that, The thickness of the stretchable interconnect circuit (6) is 30 μm - 50 μm; the wire width of the wire structure is 50 μm - 200 μm, the resistance change rate is 2% - 10% under a 40% tensile strain rate, and the resistance change rate is 2% - 5% under 20% cyclic stretching, and the fatigue life is at least 10 4 times; the strain-sensitive unit array contains 8 - 64 strain-sensitive units, the width of each strain-sensitive unit is 50 μm - 100 μm, and the thickness is 30 μm - 50 μm; the strain-sensitive unit has dynamic detection performance, can detect a planar displacement of 0.1 mm - 5 mm, the resolution is 10 μm - 50 μm, the sensitivity is at least 10 Ω / mm, and is synchronized with the electromyogram signal at the 50 millisecond level.

4. The multi-channel electromyogram strain dual-modal electronic skin according to claim 1, wherein The conductive ink in the stretchable interconnect circuit (6) and the strain-sensitive unit array adopts a mixed system of conductive silver paste and gallium-indium liquid metal alloy. The conductive ink with a liquid metal content of 60wt%-80wt% means that the proportion of gallium-indium liquid metal alloy in the conductive ink is 60wt%-80wt%; the conductive ink with a liquid metal content of 20wt%-40wt% means that the proportion of gallium-indium liquid metal alloy in the conductive ink is 60wt%-80wt%; the mass ratio of gallium to indium in the gallium-indium liquid metal alloy used is 3-4:

1.

5. A multi-channel electromyogram strain dual-modal electronic skin according to claim 1, characterized in that The thickness of the dielectric isolation layer (4) is 20 μm-30 μm; the preparation method of the highly cross-linked polydimethylsiloxane elastomer is the same as the preparation method of the stretchable flexible base layer (7).

6. The multi-channel electromyogram strain dual-modal electronic skin according to claim 1, characterized in that The thickness of the stretchable circuit (3) is 30 μm-50 μm; the spacing between the electrode units in the electrode array is 5 mm-10 mm, the equivalent diameter of each electrode unit is 3 mm-5 mm, the number of electrode units is 8-64, and the relationship between the skin contact impedance of the electrode unit and the AC frequency is that the contact impedance is 4.4 kΩ-5 kΩ at 1 kHz; the thickness of the electrode array is 50 μm-80 μm.

7. A multi-channel electromyogram strain dual-modal electronic skin according to claim 1, characterized in that, The electrode unit is prepared by mixing polydimethylsiloxane (PDMS) prepolymer and curing agent in a mass ratio of 30-50:1, adding 1-10 μm silver powder to form a mixed solution, wherein the mass ratio of silver powder to polydimethylsiloxane (PDMS) prepolymer is 2.5-3.5:1; and dispensing and printing the mixed solution to form the electrode unit.

8. The multi-channel electromyogram strain dual-modal electronic skin according to claim 1, characterized in that, The spatial mapping relationship between the stretchable circuit (3) and the multi-channel electromyographic sensing layer (2) is as follows: the circuit array of the stretchable circuit (3) and the electrode array in the multi-channel electromyographic sensing layer (2) are the same in size and position; the spatial mapping relationship between the multi-channel electromyographic sensing layer (2) and the multi-channel strain sensing layer (5) is as follows: the centers of the two ends of the length direction of each strain sensitive unit in the multi-channel strain sensing layer (5) correspond to the center positions of two adjacent electrode units in the multi-channel electromyographic sensing layer (2).

9. A multi-channel electromyogram strain dual-modal electronic skin according to claim 1, characterized in that, The adhesive encapsulation layer (1) has an adhesive strength of 0.6N / cm-0.7N / cm and a thickness of 20μm-30μm; a method for preparing a low-crosslinked polydimethylsiloxane elastomer comprises mixing a polydimethylsiloxane (PDMS) prepolymer with a curing agent, stirring the mixture thoroughly and then performing a defoaming treatment, and then applying the mixture on a multi-channel electromyographic sensing layer (2) by spin coating, and then performing a heat curing treatment; the mass ratio of the polydimethylsiloxane (PDMS) prepolymer to the curing agent is 30-50:1; the heat curing treatment temperature is 80°C-100°C, and the heat curing treatment time is 0.5 hour-1 hour.

10. A method for decomposing nerve signals of a multi-channel electromyogram strain dual-modal electronic skin according to any one of claims 1-9, characterized in that, The details are as follows: The multi-channel electromyogram (EMG)-strain dual-modal electronic skin is pasted on the human skin and connected to a multi-channel EMG acquisition device and a multi-channel resistance signal acquisition device through a pluggable interface at the end of the electrode. The sampling frequency of the EMG signal is 1 kHz - 2 kHz, and the sampling frequency of the resistance signal is 200 Hz - 400 Hz. Then, the resistance signal is divided by the sensitivity to obtain the strain signal of the skin, and finally, the EMG-strain dual-modal signal of human muscle activity is obtained. The EMG-strain dual-modal signal of human muscle activity is processed according to the following steps: 1) Signal preprocessing: Perform band-pass filtering and power frequency notch filtering on the EMG signal; perform Kalman filtering on the strain signal; 2) Motor unit localization: Use the blind source separation algorithm to decompose the preprocessed EMG signal to obtain the motor unit activity sequence, and use the spike-triggered averaging method to analyze the motor unit activity sequence to obtain the initial motor unit position and motor unit waveform, so as to achieve the purpose of motor unit localization; 3) Motion artifact elimination: According to the preprocessed strain signal, construct a motion artifact spatial distribution map, and use the motion artifact spatial distribution map to correct the motor unit position, so as to eliminate the motion noise component related to the strain signal in the EMG signal, and then eliminate the motion artifact; 4) Multi-modal fusion output: Align the motor unit activity sequence, motor unit waveform, corrected motor unit position and strain signal in the time domain, and output a multi-modal fusion signal composed of the EMG signal, motor unit activity sequence, corrected motor unit position, motor unit waveform and strain signal.

Images