Intelligent artificial limb mimicry nerve pressure sensing system based on synaptic transistor

Abstract

The invention relates to the field of artificial limb pressure sensing, in particular to an intelligent artificial limb mimic nerve pressure sensing system based on synaptic transistors. The system comprises a receptor, a voltage-to-frequency circuit, a synaptic transistor, a current-to-voltage circuit and a palm-like electromagnetic actuator receptor which are electrically connected in sequence, and the receptor detects pressure through a resistive flexible pressure sensor and converts a pressure signal into a voltage signal; the voltage-to-frequency circuit converts the voltage signal into an electric pulse with a corresponding frequency; the synaptic transistor converts the electric pulse into post-synaptic current; the current-to-voltage circuit converts the post-synaptic current into a voltage signal after synaptic processing; a resistive flexible pressure sensor is installed at the palm or finger pulp position of the palm-like electromagnetic actuator and is opened and closed in response to the voltage signal after synapse processing, and the opening and closing degree is in positive correlation with the voltage signal after synapse processing. According to the invention, the self-adaptive grabbing of the palm-like electromagnetic actuator on the object is realized under the condition of no CPU (Central Processing Unit) intervention.

Description

Technical Field

[0001] The present invention relates to the field of prosthetic limb pressure perception, and in particular to an intelligent prosthetic limb mimetic neural pressure perception system based on synaptic transistors. Background Art

[0002] Traditional prosthetic pressure sensing systems have limitations in functionality and user experience.

[0003] First, these systems typically rely on rigid electronic sensors (such as strain gauges or capacitive pressure sensors), whose mechanical properties do not match the compliance of human soft tissue, resulting in poor wearing comfort and easy signal distortion due to deformation during dynamic use.

[0004] Secondly, the signal processing of traditional sensors relies on a centralized von Neumann computing architecture, which requires frequent transmission of sensor data to an external processor for analysis. This not only introduces delays but also limits battery life due to high power consumption.

[0005] Furthermore, existing systems lack the adaptive capabilities of biological nervous systems. They cannot dynamically adjust sensitivity or learn pressure patterns for different tasks like natural tactile receptors, resulting in coarse and unnatural feedback. For example, when grasping fragile objects, traditional prosthetic limbs may be unable to distinguish between slight slips and stable grips in real time, increasing the risk of manipulation failure.

[0006] These limitations make it difficult for prosthesis users to perform fine operations or obtain intuitive tactile feedback, seriously restricting the functional recovery of prostheses and the improvement of quality of life. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to overcome the defects of the existing technology and provide an intelligent prosthetic limb mimetic neural pressure sensing system based on synaptic transistors, which combines synaptic transistors with resistive flexible pressure sensors to achieve adaptive grasping of objects by palm-like electromagnetic actuators without the need for CPU intervention.

[0008] In order to solve the above technical problems, the technical solution of the present invention is: an intelligent prosthetic limb mimicking neural pressure sensing system based on synaptic transistors, comprising:

[0009] A sensor, which detects pressure through its resistive flexible pressure sensor and is used to convert the pressure signal into a voltage signal;

[0010] a voltage-to-frequency circuit, connected to the sensor, for converting the voltage signal into an electrical pulse of corresponding frequency;

[0011] a synaptic transistor, connected to the voltage-to-frequency circuit, for converting the electrical pulse into a postsynaptic current;

[0012] a current-to-voltage circuit, connected to the synaptic transistor, for converting the post-synaptic current into a voltage signal after synaptic processing;

[0013] A palm-like electromagnetic actuator has the resistive flexible pressure sensor installed at the palm or fingertips, which is connected to the current-to-voltage circuit and is configured to open and close in response to the voltage signal after synaptic processing, and the degree of opening and closing is positively correlated with the voltage signal after synaptic processing.

[0014] Furthermore, the sensor is a Wheatstone bridge, and the resistive flexible pressure sensor is a bridge arm of the Wheatstone bridge.

[0015] Furthermore, the resistive flexible pressure sensor includes an upper PI flexible substrate, an upper interdigital electrode, a pressure-sensitive film having a micro-cone array, a lower interdigital electrode and a lower PI flexible substrate stacked sequentially from top to bottom.

[0016] Furthermore, the preparation method of the resistive flexible pressure sensor includes:

[0017] The steps for preparing a pressure-sensitive film with a microcone array are as follows: graphene oxide and multi-walled carbon nanotubes are poured into a glass beaker containing dimethylformamide and stirred, and then ultrasonicated using an ultrasonic wall breaking machine. The resulting suspension is poured into a custom template with a microcone array and baked in an oven to evaporate the organic solvent, thereby obtaining a pressure-sensitive film with a microcone array.

[0018] Preparation of interdigital electrodes: Printing interdigital electrodes on a polyimide substrate using a microelectronic printer, followed by hot-plate annealing in an atmospheric atmosphere, wherein the printing material is silver paste;

[0019] The steps of preparing the packaging material are as follows: polydimethoxysilane and a curing agent are mixed and stirred thoroughly, and then vacuumed in a vacuum dish to remove bubbles to obtain a packaging material;

[0020] Packaging step: using packaging materials to package the two interdigitated electrodes and the pressure-sensitive film with the micro-cone array into a resistive flexible pressure sensor with a sandwich structure; wherein the two interdigitated electrodes are distributed on both sides of the pressure-sensitive film with the micro-cone array.

[0021] Furthermore, the synaptic transistor is a flexible structure formed by inkjet printing.

[0022] Furthermore, the preparation method of the synaptic transistor includes:

[0023] Step A: inkjet printing a bottom gate electrode on a PI substrate using silver ink, followed by a gradient temperature post-treatment; wherein the silver ink is prepared by dissolving silver acetate in a mixed solution of ammonium hydroxide and formic acid and adding 2,3-butanediol;

[0024] Step B: Continue to inkjet print the zirconium oxide gate dielectric layer with zirconium oxide ink, then cure it on a hot stage, repeat this process two to four times to reach the desired thickness, and then hot stage anneal the zirconium oxide gate dielectric layer; wherein the zirconium oxide ink is prepared by dissolving zirconium oxychloride in acetonitrile and ethylene glycol in a precursor ink and filtering it with a filter;

[0025] Step C, continuing to inkjet print the indium gallium zinc oxide channel layer with IGZO ink and performing hot stage annealing; wherein the IGZO ink is obtained by mixing indium nitrate hydrate, gallium nitrate hydrate, and zinc nitrate hydrate in a molar ratio of 9:1:2 and filtering;

[0026] Step D: Continue to use the silver ink in step A to inkjet print the top source and drain electrodes.

[0027] Furthermore, the palm-like electromagnetic actuator includes stacked fabric layers and a composite film, and the composite film is formed by hot-pressing a hot-pressed film material and a copper mesh.

[0028] After adopting the above technical solution, the present invention has the following beneficial effects:

[0029] (1) The flexible pressure sensor of the present invention fits the human body better, which can increase the wearing comfort of prosthetic users and reduce signal distortion during dynamic use;

[0030] (2) The synaptic transistor achieves a biologically similar wide dynamic range (0Pa-35kPa) and adaptive response by simulating the electrical pulse frequency modulation and weight-adjustable characteristics of the synapse, restoring the dynamic adaptability of biological pressure perception. Nonlinear pulse coding (the pulse frequency changes with the pressure index) can simultaneously realize micro-pressure detection and strong pressure detection.

[0031] (3) Traditional prosthetic pressure sensing systems rely on independent ADCs and processors, resulting in high latency and energy consumption. The intelligent prosthetic limb mimetic neural pressure sensing system based on synaptic transistors of the present invention can reduce data transmission requirements through integrated storage and computing, thereby reducing power consumption and enhancing the endurance of intelligent prostheses.

[0032] (4) This system adopts a neural pulse signal encoding mechanism to directly convert pressure information into biocompatible electrical pulse signals, which not only greatly reduces power consumption but also seamlessly connects with residual peripheral nerves to provide a more natural tactile perception experience.

[0033] These breakthrough advantages make this invention promising to bring a truly bionic, intelligent and low-power next-generation prosthetic solution to amputees. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1This is a principle block diagram of the intelligent prosthetic limb mimicking neural pressure sensing system based on synaptic transistors of the present invention;

[0035] Figure 2 This is a workflow diagram of the synaptic transistor-based intelligent prosthetic neural pressure sensing system of the present invention;

[0036] Figure 3 A circuit diagram of a sensor and a non-inverting amplifier circuit of the present invention;

[0037] Figure 4 A circuit diagram of an integral operation circuit, a hysteresis comparator circuit, and a zero-crossing comparator circuit of the present invention;

[0038] Figure 5 It is an output curve diagram of the integral operation circuit, the hysteresis comparator circuit and the zero-crossing comparator circuit of the present invention;

[0039] Figure 6 An exploded view of the resistive flexible pressure sensor of the present invention;

[0040] Figure 7 A diagram showing the relationship between pressure and current of the resistive flexible pressure sensor of the present invention;

[0041] Figure 8 A diagram showing the relationship between the pressure applied to the resistive flexible pressure sensor of the present invention and the output frequency of the voltage-to-frequency circuit;

[0042] Figure 9 is a diagram showing the internal structure of the synaptic transistor of the present invention;

[0043] Figure 10 is a response curve diagram of the synaptic transistor of the present invention to a single interval pulse;

[0044] Figure 11 These are characteristic curves of the synaptic transistor of the present invention under 3V, 20ms interval pulse, 3V, 380ms interval pulse and 3V, 850ms interval pulse respectively.

[0045] Figure 12 is a graph of postsynaptic current of the synaptic transistor of the present invention under different pressures;

[0046] Figure 13 is a circuit diagram of a current-to-voltage circuit of the present invention;

[0047] Figure 14 is a schematic diagram of a palm-like electromagnetic actuator of the present invention;

[0048] Figure 15 Schematic diagram of the opening and closing degree of the palm-like electromagnetic actuator of the present invention under different voltages.

[0049] Figure 16 A diagram showing the relationship between the pressure applied to the resistive flexible pressure sensor of the present invention and the closing angle of the palm-like electromagnetic actuator;

[0050] In the picture:

[0051] 1. Sensor; 11. Resistive flexible pressure sensor; 111. Upper PI flexible substrate; 112. Upper interdigital electrode; 113. Pressure-sensitive film with microcone array; 114. Lower interdigital electrode; 115. Lower PI flexible substrate;

[0052] 2. Voltage-to-frequency circuit; 21. Non-inverting amplifier circuit; 22. Integral operation circuit; 23. Hysteresis comparator circuit; 24. Zero-crossing comparator circuit;

[0053] 3. Synaptic transistor; 31. Source; 32. Drain; 33. Amorphous indium gallium zinc oxide metal oxide channel layer; 34. Zirconium oxide gate dielectric layer; 35. Gate; 36. Polyimide film substrate;

[0054] 4. Current to voltage circuit;

[0055] 5. Palm-like electromagnetic actuator;

[0056] Figure 14 (a) Top view of the palm-like electromagnetic actuator;

[0057] Figure 14 (b) is a bottom view of the palm-like electromagnetic actuator. DETAILED DESCRIPTION

[0058] In order to make the contents of the present invention more clearly understood, the present invention is further described in detail below based on specific embodiments in conjunction with the accompanying drawings.

[0059] like Figures 1 to 16 As shown, an intelligent prosthetic limb mimicking neural pressure sensing system based on synaptic transistors includes:

[0060] The bionic trap-jaw ant's sensor 1 includes a Wheatstone bridge in which one arm is a resistive flexible pressure sensor 11, which is used to convert a pressure signal into a voltage signal;

[0061] The voltage-to-frequency circuit 2 is connected to the sensor 1 and the bionic nerve fiber and is used to convert the voltage signal into an electric pulse of corresponding frequency;

[0062] a synaptic transistor 3 connected to the voltage-to-frequency circuit 2 for converting electrical pulses into postsynaptic currents;

[0063] a current-to-voltage circuit 4 connected to the synaptic transistor 3 and configured to convert the post-synaptic current into a voltage signal after synaptic processing;

[0064] The palm-like electromagnetic actuator 5 has a resistive flexible pressure sensor 11 installed at the palm or fingertips, which is connected to the current-to-voltage circuit 4 and is configured to open and close in response to the voltage signal after synaptic processing, and the degree of opening and closing is positively correlated with the voltage signal after synaptic processing.

[0065] The gate 35 of synaptic transistor 3 serves as the transistor's port signal input, simulating the presynaptic membrane of a biological synapse. The conductance between source 31 and drain 32 simulates the biological synaptic weight. Depending on the amplitude, frequency, and duration of the pulse input to gate 35, the synaptic transistor can generate different excitatory postsynaptic currents. The output of voltage-to-frequency circuit 2 is connected to gate 35 of synaptic transistor 3; drain 32 outputs current through a series resistor connected to ground to generate a voltage; drain 32 is connected to the input of current-to-voltage circuit 4.

[0066] Specifically, the synaptic transistor 3 can simulate the plasticity of biological neural synapses, achieving dynamic learning and adaptive capabilities similar to those of the human tactile system, enabling the prosthesis to adjust its sensitivity in real time according to the usage scenario, such as automatically adjusting pressure feedback when grasping different objects. In addition, the system adopts a neural pulse signal encoding mechanism to directly convert pressure information into biocompatible electrical pulse signals, which not only greatly reduces power consumption but also seamlessly connects with residual peripheral nerves to provide a more natural tactile perception experience. In addition, thanks to the integration of flexible electronic technology, the system can be made into an ultra-thin, stretchable sensor network that perfectly fits the prosthetic surface and can maintain stable signal acquisition performance under complex deformations. More importantly, its distributed neuromorphic computing architecture allows information to be processed on-site at the sensor end, avoiding the delay problem of round-trip data transmission in traditional systems and achieving a millisecond-level response speed. These breakthrough advantages make the present invention expected to bring a truly bionic, intelligent and low-power next-generation prosthetic solution to amputees.

[0067] In this embodiment, if Figure 6 As shown, the resistive flexible pressure sensor 11 preferably includes an upper PI flexible substrate 111, an upper interdigital electrode 112, a pressure-sensitive film 113 having a micro-cone array, a lower interdigital electrode 114, and a lower PI flexible substrate 115 stacked sequentially from top to bottom. The preparation method of the resistive flexible pressure sensor 11 may include, but is not limited to:

[0068] Preparation of a pressure-sensitive film 113 with a microcone array: 10 mg of graphene oxide (RGO) and 5 mg of multi-walled carbon nanotubes (CNTs) were poured into a glass beaker containing 10 ml of dimethylformamide and stirred for 2 h. The mixture was then ultrasonically treated at 100 Hz using an ultrasonic wall breaking machine. The resulting suspension was poured into a custom-made template with a microcone array and baked in a 70°C oven for 12 h to evaporate the organic solvent, thereby obtaining a pressure-sensitive film 113 with a microcone array.

[0069] Steps for preparing interdigitated electrodes: Use a microelectronic printer to print interdigitated electrodes on a polyimide (PI) substrate, and then perform hot-plate annealing in an atmospheric atmosphere at a temperature set to 120°C for 30 minutes. The printing material is silver paste, the printing pressure is 70 kPa, the printing speed is 1 mm / s, and the printing height is 50 μm.

[0070] Preparation of packaging material: polydimethoxysilane (PDMS) and curing agent are mixed at a mass ratio of 10:1 and stirred thoroughly, and then vacuumed in a vacuum dish for 20 minutes to remove bubbles to obtain a packaging material;

[0071] Packaging step: Use packaging material to package the two interdigitated electrodes and the pressure-sensitive film 113 with a micro-cone array into a resistive flexible pressure sensor 11 with a sandwich structure; wherein, the two interdigitated electrodes are distributed on both sides of the pressure-sensitive film 113 with a micro-cone array, namely the upper interdigitated electrode 112 and the lower interdigitated electrode 114, the polyimide (PI) substrate outside the upper interdigitated electrode 112 is the upper PI flexible substrate 111, and the polyimide (PI) substrate outside the lower interdigitated electrode 114 is the lower PI flexible substrate 115.

[0072] The resistance value of this resistive flexible pressure sensor 11 decreases as the pressure increases. When a linear pressure is applied to the pressure sensor, the current of the sensor increases linearly. As the pressure applied to the sensor increases, the current gradually reaches a saturated state. Figure 7 shown.

[0073] In this embodiment, if Figure 3 and Figure 4 As shown, the voltage-to-frequency circuit 22 preferably includes a common-mode amplifier circuit 21, an integral operation circuit 22, a hysteresis comparator circuit 23 and a zero-crossing comparator circuit 24; wherein, the input end of the common-mode amplifier circuit 21 is connected to the Wheatstone bridge, which is the total input port of the signal; the output end of the common-mode amplifier circuit 21 is connected to the input end of the integral operation circuit 22; the output end of the integral operation circuit 22 is connected to the input end of the hysteresis comparator 23; the output end of the hysteresis comparator 23 is connected to the input end of the integral operation circuit 22 in series through a diode and a resistor, and is also directly connected to the input end of the zero-crossing comparator circuit 24; the output port of the zero-crossing comparator is the total output port of the voltage-to-frequency circuit 2.

[0074] Resistors R4, R5 and op amp U1 constitute a non-inverting amplifier circuit 21. The amplification ratio of the non-inverting amplifier circuit 21 is determined by the ratio of R4 and R5, and the amplification factor is (1+R4 / R5); resistors R7, R8, capacitor C1 and op amp U2 constitute an integration operation circuit 22; resistors R9, R10, R11 and op amp U3 constitute a hysteresis comparator circuit 23; resistor R12 and op amp U4 constitute a zero-crossing comparator circuit 24.

[0075] Figure 5 It is the output curve of the integral operation circuit 22, the hysteresis comparator circuit 23 and the zero-crossing comparator circuit 24. When the output voltage of the integral operation circuit 22 is lower than the lower threshold of the hysteresis comparator circuit 23, the hysteresis comparator circuit 23 outputs a negative voltage, and the charge of the capacitor C1 is released through the diode and the resistor R6. At this time, the output of the integral operation circuit 22 begins to rise. When the output voltage of the integral operation circuit 22 is higher than the upper threshold of the hysteresis comparator circuit 23, the hysteresis comparator circuit 23 outputs a positive voltage. Return to the charging process of the capacitor C1. The zero-crossing comparator circuit 24 converts the positive voltage into zero volts and the negative voltage into a high voltage, forming a periodically changing pulse signal. Figure 8 That is, it is a relationship diagram between the pressure exerted on the resistive flexible pressure sensor and the output frequency of the voltage-to-frequency circuit 22 (specifically, the pressure gradient increases with time, and the pulse signal output by the voltage-to-frequency circuit 22).

[0076] In this embodiment, the synaptic transistor 3 is preferably a flexible structure formed by inkjet printing. Figure 9 As shown, the synaptic transistor 3 may include, but is not limited to, a polyimide film substrate 36, a zirconium oxide gate dielectric layer 34, an amorphous indium gallium zinc oxide metal oxide channel layer 33, and a metal silver (Ag) conductive layer. The entire transistor adopts a vertical structure design. The preparation method of the synaptic transistor 3 may include, but is not limited to:

[0077] Step A: inkjet print the bottom gate electrode with silver ink on the PI substrate (i.e., polyimide film substrate 36), and then perform gradient temperature post-treatment: first heat at 90°C until the liquid film turns silver, then heat at 130°C for 10 minutes, and finally heat at 150°C for more than 30 minutes.

[0078] The specific materials of the silver ink include silver acetate, ammonium hydroxide and formic acid. Silver acetate (2g) is dissolved in 4ml of a mixed solution of ammonium hydroxide and formic acid (the ratio of the two is 3:1), and a small amount of 2,3-butanediol (10% by volume) is added as a wetting agent and viscosity enhancer.

[0079] The silver ink has a viscosity of 2-10 cP and a particle size of less than 100 nm.

[0080] The microelectronic printer parameters were set as follows: printing voltage 37 V, substrate temperature 30 °C, nozzle aperture 20 μm, nozzle temperature 30 °C, and the printed dot pitch was set to 35 μm;

[0081] Step B: Based on step A, continue to inkjet print the zirconium oxide gate dielectric layer 34 with zirconium oxide ink, then cure the sample on a hot stage at 250°C for 5 minutes. This process needs to be repeated two to four times to achieve the required thickness (200-400nm) and good insulation properties. Then, the zirconium oxide gate dielectric layer 34 (i.e., ZrOx film) is hot-stage annealed at 300°C under normal pressure for 2 hours.

[0082] The specific materials of zirconium oxide ink include zirconium oxychloride (ZrOCl4), acetonitrile and ethylene glycol. Zirconium oxychloride is dissolved in acetonitrile and ethylene glycol in a ratio of 3.5:6.5 to prepare a precursor ink with a concentration of 0.4 mol / L. It is then vigorously stirred in air for 2 hours and filtered with a 0.45 μm polytetrafluoroethylene (PTFE) filter.

[0083] The viscosity of the zirconium oxide ink is 2-10 cP and the particle size is less than 100 nm.

[0084] The printer parameters were set as follows: printing voltage 37 V, substrate temperature 30 °C, nozzle aperture 20 μm, nozzle temperature 30 °C, and dot pitch 30 μm.

[0085] Step C: Based on step B, continue to inkjet print the amorphous indium gallium zinc oxide metal oxide channel layer 33 with IGZO ink, and then perform hot stage annealing in an atmospheric atmosphere at a temperature of 300°C for 1 hour.

[0086] The IGZO ink consists of indium nitrate hydrate (In(NO3)3·xH2O), gallium nitrate hydrate (Ga(NO3)3·xH2O), and zinc nitrate hydrate (Zn(NO3)2·xH2O) in a molar ratio of 9:1:2. The ink has a viscosity of 2-10 cP and a particle size of less than 100 nm. After preparation, the ink is stirred vigorously in air for 2 hours and filtered through a 0.45 μm polytetrafluoroethylene (PTFE) filter.

[0087] The printer parameters were set as follows: printing voltage 37 V, substrate temperature 30 ° C, nozzle aperture 20 μm, nozzle temperature 30 ° C. The printed dot pitch was set to 30 μm.

[0088] In step D, based on step C, the top source and drain electrodes are further inkjet-printed using silver ink. The ink formulation, annealing method, and printer printing parameters used are consistent with those in step A.

[0089] Figure 10 Here is the corresponding curve for 7V, 380ms interval pulses for this synaptic transistor. Figure 11These are the characteristic curves of the synaptic transistor under 3V, 20ms interval pulse, 3V, 380ms interval pulse and 3V, 850ms interval pulse (corresponding pressures are 50, 200 and 2000Pa respectively). Figure 12 This is the postsynaptic current output by the synaptic transistor 3 under different pressures.

[0090] Specifically, the synaptic transistor 3 uses metal oxide materials and full-printing preparation technology to directly deposit multiple layers of functional materials under normal temperature and pressure conditions, which significantly simplifies the large-scale preparation process, reduces production costs, and can meet the needs of large-scale integration.

[0091] In this embodiment, if Figure 13 As shown, the current-to-voltage circuit 4 includes an operational amplifier and a capacitor C2, which plays a filtering role. Considering the cost requirements of large-scale production, the operational amplifier is preferably a high-precision OPA140.

[0092] In this embodiment, preferably, the palm-like electromagnetic actuator 5 includes a laminated fabric layer and a composite film. The composite film is formed by hot-pressing a hot-pressed film material and a copper mesh. The fabric material can be cotton, nylon, polyester, or silk. The hot-pressed film material can be, but is not limited to, TPU (thermoplastic polyurethane), which has an operating temperature of 130°C and an instantaneous operating temperature of up to 160°C. The copper mesh specifications can be, but are not limited to, 300 mesh and 50μm thick. The preparation method of the palm-like electromagnetic actuator 5 of this structure can be, but is not limited to:

[0093] Step a, using a hot stamping machine to press the copper mesh and the TPU film at 170° C. for 10 seconds to form a copper mesh / TPU composite film;

[0094] Step b, cutting the composite film into a design pattern (i.e., a pattern similar to a palm) by a die-cutting machine and removing waste;

[0095] Step c, leading out the central electrode through a printing process using a microelectronic printer;

[0096] In step d, the composite film was again pressed onto the fabric substrate using a hot stamping machine at 170° C. for 10 seconds, and the release paper was removed to complete the preparation.

[0097] Among them, the hot stamping machine (model CM1001) used in the preparation process was from Cooldiy Digital Graphic Co., Ltd., and the die-cutting machine (model ZK500) was from Hefei Zhike CNC Equipment Co., Ltd.

[0098] like Figure 14As shown, the prepared palm-like electromagnetic actuator 5 can have a middle area and several blade areas arranged along the circumference of the middle area. The middle area is used to simulate the palm, and a resistive flexible pressure sensor 11 is set. The blade area is used to simulate the fingers. The copper mesh is only located in the blade area. The base of the copper mesh is energized. Due to the principle of electromagnetic induction, the copper mesh will roll up adaptively according to the voltage size.

[0099] This palm-like electromagnetic actuator 5 can adaptively open and close upon receiving the voltage signal after synaptic processing. Figure 15 It is the opening and closing degree of the palm-like electromagnetic actuator 5 under different voltages. Figure 16 The relationship between the voltage received by the palm-like electromagnetic actuator 5 and its closing angle is shown in Figure 1. Within a voltage range of 0 to 5V, the voltage and closing angle are essentially linear, with a maximum closing degree of 80 degrees. The palm-like electromagnetic actuator 5 is equivalent to a prosthetic hand. The resistive flexible pressure sensor 11 is mounted on the palm or fingertips of the hand. With the palm facing upward, an object is held. As the resistive flexible pressure sensor 11 senses increasing pressure, it causes the synaptic transistor 3 to generate a higher postsynaptic current. After passing through the current-to-voltage circuit 4, a linearly increasing voltage is generated, which increases the blade angle of the palm-like electromagnetic actuator 5. The greater the voltage applied to the palm-like electromagnetic actuator 5, the tighter the palm-like electromagnetic actuator 5 closes. In other words, the heavier the object, the tighter the grip, resulting in a tighter grip. This prevents the object from falling and causing injury to the prosthetic user.

[0100] Specifically, the synaptic transistor 3, the resistive flexible pressure sensor 11, and the palm-like electromagnetic actuator of this embodiment are all made of flexible materials, which are more compatible with the soft tissue of the human body, greatly increasing the comfort of the wearable device and the diversity of application scenarios, and are less likely to cause signal distortion due to deformation during dynamic use.

[0101] In the above embodiment, the intelligent prosthesis-based neural pressure sensing system based on synaptic transistors combines the synaptic transistor 3 with the resistive flexible pressure sensor 11. By simulating the dynamic sensitivity adjustment of biological touch, it can adaptively adjust the grasping force of the palm-like electromagnetic actuator 5. When this system is installed on an intelligent prosthesis, it can enable the prosthesis user to sense subtle pressure changes (such as holding an egg) and strong pressure (such as lifting a heavy object). When the object being grasped is heavier, the intelligent prosthesis-based neural pressure sensing system can adaptively increase the grasping force, thereby preventing the heavy object from falling and causing harm to the prosthesis user. The adaptive grasping force can also prevent the grasping force from being too strong and damaging the grasped object. Importantly, this process does not require complex CPU calculations to complete. Traditional systems usually issue commands from the CPU and require pre-programming. This process will cause delays and cause action jams. In addition, traditional systems require continuous power supply to the CPU, while the pressure sensing system of the above embodiment is controlled by synaptic transistors and has extremely low power consumption.

[0102] Energy consumption comparison between this embodiment and the traditional system:

[0103]

[0104] With the above-described preferred embodiments of the present invention as a guide, and with reference to the above description, relevant personnel are fully capable of making various changes and modifications without departing from the technical scope of this invention. The technical scope of this invention is not limited to the contents of the specification and must be determined according to the scope of the claims.

Claims

1. An intelligent prosthetic limb mimicking neural pressure sensing system based on synaptic transistors, characterized in that: include: A sensor (1) detects pressure through its resistive flexible pressure sensor (11) and is used to convert the pressure signal into a voltage signal; A voltage-to-frequency circuit (2) is connected to the sensor (1) and is used to convert the voltage signal into an electrical pulse of a corresponding frequency; a synaptic transistor (3), connected to the voltage-to-frequency circuit (2), for converting the electrical pulse into a postsynaptic current; a current-to-voltage circuit (4), connected to the synaptic transistor (3), for converting the post-synaptic current into a voltage signal after synaptic processing; A palm-like electromagnetic actuator (5) is provided with the resistive flexible pressure sensor (11) installed at the palm or fingertips of the hand, connected to the current-to-voltage circuit (4), and configured to open and close in response to the voltage signal after synaptic processing, and the degree of opening and closing is positively correlated with the voltage signal after synaptic processing.

2. The intelligent prosthetic limb mimicking neural pressure sensing system based on synaptic transistor according to claim 1 is characterized in that: The sensor (1) is a Wheatstone bridge, and the resistive flexible pressure sensor (11) is a bridge arm of the Wheatstone bridge.

3. The intelligent prosthetic limb mimicking neural pressure sensing system based on synaptic transistor according to claim 2 is characterized in that: The resistive flexible pressure sensor (11) comprises an upper PI flexible substrate (111), an upper interdigital electrode (112), a pressure-sensitive film (113) having a micro-cone array, a lower interdigital electrode (114), and a lower PI flexible substrate (115), which are stacked sequentially from top to bottom.

4. The intelligent prosthetic limb mimicking neural pressure sensing system based on synaptic transistor according to claim 3 is characterized in that: The preparation method of the resistive flexible pressure sensor (11) comprises: The steps of preparing a pressure-sensitive film (113) having a micro-cone array are as follows: graphene oxide and multi-walled carbon nanotubes are poured into a glass beaker of dimethylformamide and stirred, and then ultrasonically treated with an ultrasonic wall breaking machine, and the obtained suspension is poured into a customized template having a micro-cone array, and baked in an oven to evaporate the organic solvent, thereby obtaining a pressure-sensitive film (113) having a micro-cone array; Preparation of interdigital electrodes: Printing interdigital electrodes on a polyimide substrate using a microelectronic printer, followed by hot-plate annealing in an atmospheric atmosphere, wherein the printing material is silver paste; The steps of preparing the packaging material are as follows: polydimethoxysilane and a curing agent are mixed and stirred thoroughly, and then vacuumed in a vacuum dish to remove bubbles to obtain a packaging material; Packaging step: using packaging material to package two interdigitated electrodes and a pressure-sensitive film (113) with a micro-cone array into a resistive flexible pressure sensor (11) with a sandwich structure; wherein the two interdigitated electrodes are distributed on both sides of the pressure-sensitive film (113) with a micro-cone array.

5. The intelligent prosthetic limb mimicking neural pressure sensing system based on synaptic transistor according to claim 1 is characterized in that: The synaptic transistor (3) is a flexible structure formed by inkjet printing.

6. The intelligent prosthetic limb mimicking neural pressure sensing system based on synaptic transistor according to claim 5 is characterized in that: The preparation method of the synaptic transistor (3) comprises: Step A: inkjet printing a bottom gate electrode on a PI substrate using silver ink, followed by a gradient temperature post-treatment; wherein the silver ink is prepared by dissolving silver acetate in a mixed solution of ammonium hydroxide and formic acid and adding 2,3-butanediol; Step B: Continue to inkjet print the zirconium oxide gate dielectric layer with zirconium oxide ink, then cure it on a hot stage, repeat this process two to four times to reach the desired thickness, and then hot stage anneal the zirconium oxide gate dielectric layer; wherein the zirconium oxide ink is prepared by dissolving zirconium oxychloride in acetonitrile and ethylene glycol in a precursor ink and filtering it with a filter; Step C, continuing to inkjet print the indium gallium zinc oxide channel layer with IGZO ink and performing hot stage annealing; wherein the IGZO ink is obtained by mixing indium nitrate hydrate, gallium nitrate hydrate, and zinc nitrate hydrate in a molar ratio of 9:1:2 and filtering; Step D: Continue to use the silver ink in step A to inkjet print the top source and drain electrodes.

7. The intelligent prosthetic limb mimicking neural pressure sensing system based on synaptic transistor according to claim 1 is characterized in that: The palm-like electromagnetic actuator (5) comprises a stacked fabric layer and a composite film, wherein the composite film is formed by hot-pressing a hot-pressed film material and a copper mesh.

Images