Self-adaptive regulation and control system and method simulating neuromorphic perception and action reflex

Abstract

The invention discloses a self-adaptive regulation and control system and method simulating neuromorphic perception and action reflex. The self-adaptive regulation and control system comprises a dual synaptic transistor, a signal processing circuit and an actuator terminal, external mechanical stimulation is sensed through the double synaptic transistors, excitatory / suppressive post-synaptic current is generated and connected to the signal processing circuit, synaptic current signals are converted and amplified into voltage signals capable of driving the actuator terminal through the signal processing circuit, and the actuator terminal executes response actions after receiving the voltage signals. A closed-loop feedback process of sensing-processing-action feedback is realized; the dual synaptic transistor can continuously sense the dynamic change of external stimulation, generates the action of an excitatory / inhibitory post-synaptic current self-adaptive regulation actuator in real time, and realizes the simulation of rapid reflex and self-adaptive adjustment capability of a biological nervous system.

Description

Technical Field

[0001] This invention belongs to the field of biomimetic sensing technology, specifically relating to an adaptive control system and method for neuromorphic perception and motor reflexes. Background Technology

[0002] With the rapid development of fields such as intelligent robots, medical rehabilitation, and intelligent manufacturing, intelligent sensing and feedback systems have become an important research direction in human-machine integration technology. As the core component for intelligent devices to achieve environmental interaction and autonomous decision-making, the performance of the sensing and feedback system directly determines the system's adaptability and response accuracy in dynamic and unstructured scenarios, thus possessing irreplaceable importance.

[0003] In existing technologies, most sensing systems are based on the von Neumann architecture and rely on complex central processing units for signal conversion and computation. This serial processing mode is prone to signal delay when faced with a large amount of sensing data, and the system has high power consumption, complex circuit structure, and difficulty in achieving high-density integration. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide an adaptive control system and method for neuromorphic perception and action reflex, which solves the problems of high power consumption, high latency and complex circuits in the traditional sensing system based on the von Neumann architecture, which relies on a complex central processing unit for signal conversion and operation and cannot achieve parallel sensing and processing of signals.

[0005] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0006] An adaptive control system that mimics neuromorphic perception and motor reflexes includes a dual-synaptic transistor, a signal processing circuit, and an actuator terminal.

[0007] The actuator terminal is a bionic body, with the dual synaptic transistor deployed in its key sensing parts and the signal processing circuit set in its non-key sensing parts.

[0008] The dual-synaptic transistor includes a flexible material substrate, an electrode layer, a semiconductor layer, an ion gel layer, and a triboelectric layer, and its output terminal is connected to the input interface of the signal processing circuit.

[0009] The signal processing circuit includes a signal conversion circuit and a driving circuit connected in sequence. The input terminal of the signal conversion circuit is connected to the output terminal of the dual synaptic transistor, and the output terminal of the driving circuit is connected to the actuator terminal.

[0010] The substrate is a flexible material.

[0011] The ionogel layer is composed of piezoelectric polymer, ion electrolyte and additives.

[0012] The signal conversion circuit includes three parts: a high-pass filter, a transimpedance amplifier, and a low-pass filter. The transimpedance amplifier is a low-bias transimpedance amplifier that converts microampere-level synaptic current signals into voltage signals.

[0013] The drive circuit includes an input circuit, a power amplifier circuit, and an output circuit, which converts the weak electrical signal from the signal conversion circuit into a strong electrical drive signal to provide stable power to the actuator.

[0014] Both the input circuit and the output circuit include an input impedance matching circuit and a coupling circuit.

[0015] An adaptive modulation method mimicking neuromorphic perception and motor reflexes includes the following steps:

[0016] Step 1: The dual-synaptic transistor senses external mechanical stimulation, adaptively generates excitatory / inhibitory postsynaptic currents, and inputs them to the signal conversion module of the signal processing circuit.

[0017] Step 2: The signal conversion module converts the current signal into a voltage signal and connects it to the drive module of the signal processing circuit;

[0018] Step 3: The drive module amplifies the voltage signal into a strong voltage signal that can directly drive the actuator terminal, and inputs it into the drive port of the actuator terminal.

[0019] Step 4: The actuator terminal, as a bionic body, receives the drive signal and performs the corresponding action. The dual synaptic transistors located in its key sensing parts monitor the feedback signals generated by the action in real time, including reaction force and displacement changes, and adaptively adjust the action amplitude, response speed and stability of the actuator terminal.

[0020] In step 1, the dual-synaptic transistor senses external mechanical stimulation and converts the mechanical signal into an electrical signal, including:

[0021] When an external object only contacts the dual-synaptic transistor, an excitatory postsynaptic current signal is generated based on the synergistic mechanism of triboelectric charging and electrostatic induction.

[0022] When an external object applies a pressing action to the dual-synaptic transistor, an inhibitory postsynaptic current signal is generated based on the piezoelectric polarization effect.

[0023] A computer-readable storage medium storing computer instructions that, when invoked, are used to perform the steps of the method.

[0024] The application of an adaptive control system based on neuromorphic perception and motor reflexes is applied in fields including bionic prostheses, industrial flexible grasping, service robots, and medical assistive devices.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] 1. Integrating mechanical stimulus perception with neural-like processing eliminates the need for complex microprocessor circuits for analog-to-digital conversion, thus solving the problems of high power consumption, transmission delay, and circuit complexity inherent in the von Neumann architecture.

[0027] 2. The piezoelectric ion gel used can achieve dynamic switching between excitatory and inhibitory synaptic responses, and can actively respond to and accurately adapt to external stimuli of different intensities and modes, which is more adaptable than traditional single-response sensing systems.

[0028] 3. Construct a closed-loop reflex mechanism of "perception-processing-action feedback" to simulate the rapid reflexes and adaptive adjustment capabilities of the biological nervous system, thereby enhancing the naturalness of the interaction of bionic devices.

[0029] 4. Through the above closed-loop mechanism, the system can complete the adaptive control of action behavior without additional algorithms and microprocessors, and achieve rapid intrinsic regulation similar to spinal reflexes without complex calculations. Attached Figure Description

[0030] Figure 1 This is a structural block diagram of the adaptive control system for neuromorphic perception and motor reflexes of the present invention.

[0031] Figure 2 This is a schematic diagram of the dual-synaptic transistor structure of the present invention.

[0032] Figure 3 This is a flowchart of the adaptive control method for neuromorphic perception and motor reflexes of the present invention. Detailed Implementation

[0033] The structure and working process of the present invention will be further described below with reference to the accompanying drawings.

[0034] Existing sensing systems generally rely on the traditional von Neumann architecture, where sensing, computing, and control units are separated. Information reading, transmission, and processing must be performed through a large number of microprocessors and peripheral circuits, resulting in inherent defects such as high latency, high power consumption, and complex structures. On the other hand, most sensing units only have passive response capabilities and it is difficult to achieve adaptive control of external stimuli at the device level.

[0035] The purpose of this invention is to leverage the significant advantages of the parallel sensing and feedback mechanism of the biological reflex arc system, enabling adaptive sensing and active feedback to external stimuli, and providing a solution for adaptive regulation of morphological perception and motor reflexes at the device level. This invention provides a biomimetic adaptive regulation system and method for neural morphological perception and motor reflexes, constructing an integrated biomimetic regulation mechanism of "sensing-processing-motor feedback" to simulate the synaptic plasticity and rapid reflex function of the biological nervous system.

[0036] An adaptive control system for neuromorphic perception and motor reflexes includes a dual-synaptic transistor, a signal processing circuit, and an actuator terminal. The actuator terminal is a biomimetic body, with the dual-synaptic transistor deployed in key sensing areas and the signal processing circuit located in non-key sensing areas. The dual-synaptic transistor comprises a flexible material substrate, an electrode layer, a semiconductor layer, an ion gel layer, and a triboelectric layer, and its output terminal is connected to the input interface of the signal processing circuit. The signal processing circuit includes a signal conversion circuit and a driving circuit connected in sequence. The input terminal of the signal conversion circuit is connected to the output terminal of the dual-synaptic transistor, and the output terminal of the driving circuit is connected to the actuator terminal.

[0037] Specific embodiments, such as Figures 1 to 3 As shown:

[0038] An adaptive control system mimicking neuromorphic perception and motor reflexes includes the following modules: a dual-synaptic transistor, a signal processing circuit, and an actuator terminal; such as Figure 1 As shown, the actuator terminal, which serves as the bionic body, is equipped with a dual synaptic transistor located at the key sensing part of the actuator, and a signal processing circuit connecting the synaptic transistor and the actuator terminal. The signal processing circuit includes a signal conversion circuit and a driving circuit. The dual synaptic transistor, as the core sensing unit, senses external mechanical stimuli in real time and converts them into synaptic-like currents. After being converted and amplified by the signal processing circuit, it drives the actuator terminal to complete adaptive action feedback, ultimately forming a closed-loop feedback link of "sensing-processing-action feedback".

[0039] like Figure 2 As shown, the dual-synaptic transistor is a flexible device, comprising a substrate, an electrode, a semiconductor layer, a piezoelectric ionogel layer, and a dielectric triboelectric layer arranged sequentially from bottom to top. It is disposed at the key sensing part of the actuator terminal. The device achieves differentiated responses to external light touch and pressure stimuli through the dual regulation of triboelectric-electrostatic induction synergistic mechanism and piezoelectric polarization effect, generating excitatory postsynaptic current (EPSC) and inhibitory postsynaptic current (IPSC) respectively, thus completing neuromorphic tactile perception and processing.

[0040] The substrate is made of a flexible material, and patterned electrodes are deposited on its surface by magnetron sputtering. Then, a semiconductor solution is uniformly coated on the electrode channel region by spin coating to form a semiconductor layer. Piezoelectric ion gel is then attached to the surface of the semiconductor layer, and hot pressing is combined to enhance the interfacial bonding effect between the two. A triboelectric dielectric layer is composited on the piezoelectric ion gel layer, and the surface is provided with micro-nano structures to enhance the triboelectric effect, thereby completing the fabrication of the dual synaptic transistor.

[0041] In this embodiment, when the dual synaptic transistor located at the actuator terminal comes into slight contact with an external object, its piezoelectric ion gel generates a triboelectric effect at the contact interface, causing equal amounts of opposite charges to be generated on both surfaces. Due to electrostatic induction, the electrostatic charge at the interface forms a strong electrostatic field inside the piezoelectric ion gel layer, driving ions of opposite polarity to the charge carriers in the ion gel layer to migrate directionally to the ion gel-semiconductor interface, ultimately constructing a positive double layer at the interface. The formation of the double layer significantly reduces the interface barrier of the semiconductor layer, increases the hole accumulation in the channel region, and instantaneously enhances the device's conductivity characteristics, causing the source and drain output currents to rise synchronously, generating an EPSC similar to the "excited" state of biological synapses. This process is analogous to the nerve ending excitation phenomenon caused by light touch stimulation in biological touch. The entire signal conversion process is completed by the device's own physical effects, without additional computing power consumption.

[0042] When the contact process is accompanied by a pressing action, the mechanical pressure applied from the outside causes the piezoelectric ionogel layer to deform. At this time, the piezoelectric polarization effect of the ionogel layer is significantly enhanced, and the polarization direction is reversed. The resulting strong reverse electric field drives ions with the same polarity as the charge carriers in the semiconductor layer to accumulate at the ionogel-semiconductor interface. The accumulation of ions with the same polarity will build a charge depletion layer in the semiconductor channel region, which will strongly repel the hole carriers in the channel. This will lead to a decrease in hole mobility and a significant reduction in channel conductivity. The source-drain output current will drop to a trough, generating an IPSC similar to the "inhibition" state of biological synapses. This process is analogous to the inhibition of neural response caused by strong pressure stimulation in biological touch, and the signal strength is dynamically related to the pressing pressure, thereby achieving adaptive control of the actuator action.

[0043] The signal processing circuit includes a signal conversion circuit and a drive circuit, which are integrated into the non-sensing area of ​​the actuator terminal and adopt a miniaturized packaging design to meet portability requirements.

[0044] In this embodiment, the signal conversion circuit mainly includes three parts: high-pass filter, transimpedance amplifier (TIA), and low-pass filter. Its core component is the TIA, which can realize current-to-voltage conversion. Chips such as TLC2652, OPA189, ADA4528, and ADA4522, which have low input bias current, zero drift, low noise, low offset, and high stability, can be selected.

[0045] The current signal generated by the dual synaptic transistor is fed into a high-pass filter circuit to prevent noise from entering the TIA and causing instability or loop oscillation. After being filtered, the current signal is input into the TIA and converted into a voltage signal, which is then fed into a low-pass filter to suppress high-frequency noise, spikes and aliasing, thereby improving the signal-to-noise ratio.

[0046] In this embodiment, the driving circuit mainly includes an input circuit, a power amplifier, and an output circuit. The power amplifier can be selected from chips such as OPA548, AD8555, and ENS002, which are adapted to miniaturized, low-power bionic devices.

[0047] The input circuit includes an input impedance matching circuit and a coupling circuit, which transmits the weak current control signal output from the signal conversion circuit to the power amplifier, ensuring efficient transmission and fidelity of the input signal. The output circuit includes an output impedance matching circuit and a coupling circuit, which transmits the strong current drive signal amplified by the power amplifier to the actuator terminal, an external load, ensuring that the amplified signal can stably and effectively drive the actuator and provide stable power to the actuator.

[0048] In this embodiment, the actuator terminal is a bionic body. Further, the actuator is a bionic body, and the dual synaptic transistors are distributed in the key sensing area of ​​the bionic body. They can be humanoid, animaloid, or organ-shaped. The core components can be ion polymer-metal composites, shape memory alloys, or micro servo motors.

[0049] The actuator terminal receives the drive signal from the drive circuit, completes adaptive motion feedback, and simulates the reflexive motion regulation of a living organism.

[0050] like Figure 3 The flowchart shown is the adaptive control system and method of the present invention that mimics neuromorphic perception and motor reflex. The sequence is as follows: a dual synaptic transistor senses mechanical stimulation to generate a current signal, a signal conversion circuit converts the current signal into a voltage signal, a driving circuit amplifies the voltage signal, and a driver terminal responds to the driving signal to realize motor feedback.

[0051] In the initial contact stage of the actuator's grasping of the object, the dielectric triboelectric layer of the dual-synaptic transistor makes slight contact with the object's surface. Based on the triboelectric effect, equal amounts of opposite electrostatic charges are formed at the interface, thereby constructing a strong electrostatic field inside the piezoelectric ion gel layer. This drives ions with opposite polarity to the hole carriers in the semiconductor layer to migrate directionally to the ion gel-semiconductor interface, forming a positive double layer and significantly reducing the interface barrier. This causes a continuous increase in the accumulation of holes in the channel, and the source-drain output current increases synchronously, generating an EPSC similar to the "excited" state of biological synapses. The signal conversion circuit captures the continuously changing EPSC signal in real time, converts and amplifies it, and then connects it to the drive circuit. The drive circuit converts the weak electrical command into a strong electrical drive signal through a power amplifier, controlling the gripping force of the actuator to increase synchronously. At the same time, the amplitude of the EPSC also increases continuously with the deepening of the contact, continuously regulating the gripping action of the actuator and forming a positive feedback process.

[0052] The actuator successfully grasps the object and continues to apply clamping force. The object generates a reaction force on the dual synaptic transistor on the actuator, causing the piezoelectric ionogel layer to deform. The internal piezoelectric polarization effect is significantly enhanced and the polarization direction is reversed, forming a strong reverse polarization electric field. This drives ions with the same polarity as hole carriers to gather at the ionogel-semiconductor interface, building a charge depletion layer and repelling hole carriers in the channel. This results in a significant reduction in the device conductivity, generating an IPSC signal similar to a biological "inhibition" state. The drive signal output by the signal processing circuit is reduced accordingly, causing the gripping force to gradually weaken.

[0053] During this process, the dual synaptic transistors continuously sense the dynamic changes in clamping pressure and adaptively adjust the gripping force through real-time switching between EPSC and IPSC until the gripping force is adapted to the weight and hardness of the object, ultimately maintaining the optimal state of stable clamping without damaging the object.

[0054] The adaptive control system and method for mimicking neuromorphic perception and motor reflexes disclosed in this invention can be applied to the following situations:

[0055] 1. Bionic prosthesis field: Providing upper limb amputees with bionic hands that have tactile perception and adaptive grasping functions. By simulating the sensory ability of skin tactile nerves through dual synaptic transistors, combined with closed-loop control, it can safely grasp objects of different weights and materials, and improve the natural interactivity and practicality of prosthesis use;

[0056] 2. Industrial flexible gripping field: Applied to automated assembly lines for fragile / vulnerable workpieces such as electronic components and precision instruments. The actuator end can adaptively adjust the gripping force according to the weight and hardness of the workpiece, avoiding workpiece damage caused by mechanical gripping and improving production yield.

[0057] 3. Service Robots: The robotic arms or end effectors integrated into home service robots enable them to have human-like tactile feedback and motion adjustment capabilities when performing household tasks, thereby improving the safety and intelligence of human-computer interaction;

[0058] 4. Medical assistance field: End-effectors used in minimally invasive surgical robots, which can perform gentle manipulation of human soft tissues by accurately sensing the contact force of tissues, thereby reducing surgical trauma and improving surgical safety.

[0059] It should be understood that this solution is not limited to the specific embodiments described above. Devices and structures not described in detail herein should be understood as being implemented in a manner common to the art. Any person skilled in the art can make many possible variations and modifications to this solution, or modify it into equivalent embodiments, without departing from the scope of this solution, using the methods and techniques disclosed above. This does not affect the substantive content of this solution. Therefore, any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of this solution, without departing from its scope, still fall within the protection scope of this solution.

Claims

1. An adaptive control system for neuromorphic perception and motor reflexes, characterized in that: Includes dual-synaptic transistors, signal processing circuits, and actuator terminals; The actuator terminal is a bionic body, with the dual synaptic transistor deployed in its key sensing parts and the signal processing circuit set in its non-key sensing parts. The dual-synaptic transistor includes a flexible material substrate, an electrode layer, a semiconductor layer, an ion gel layer, and a triboelectric layer, and its output terminal is connected to the input interface of the signal processing circuit. The signal processing circuit includes a signal conversion circuit and a driving circuit connected in sequence. The input terminal of the signal conversion circuit is connected to the output terminal of the dual synaptic transistor, and the output terminal of the driving circuit is connected to the actuator terminal.

2. The adaptive control system for neuromorphic perception and motor reflexes according to claim 1, characterized in that: The substrate is a flexible material.

3. The adaptive control system for neuromorphic perception and motor reflexes according to claim 1, characterized in that: The ionogel layer is composed of piezoelectric polymer, ion electrolyte and additives.

4. The adaptive control system for neuromorphic perception and motor reflexes according to claim 1, characterized in that: The signal conversion circuit includes three parts: a high-pass filter, a transimpedance amplifier, and a low-pass filter. The transimpedance amplifier is a low-bias transimpedance amplifier that converts microampere-level synaptic current signals into voltage signals.

5. The adaptive control system for neuromorphic perception and motor reflexes according to claim 4, characterized in that: The drive circuit includes an input circuit, a power amplifier circuit, and an output circuit, which converts the weak electrical signal from the signal conversion circuit into a strong electrical drive signal to provide stable power to the actuator.

6. The adaptive control system for neuromorphic perception and motor reflexes according to claim 5, characterized in that: Both the input circuit and the output circuit include an input impedance matching circuit and a coupling circuit.

7. An adaptive control method for neuromorphic perception and motor reflexes based on the system described in any one of claims 1 to 6, characterized in that, Includes the following steps: Step 1: The dual-synaptic transistor senses external mechanical stimulation, adaptively generates excitatory / inhibitory postsynaptic currents, and inputs them to the signal conversion module of the signal processing circuit. Step 2: The signal conversion module converts the current signal into a voltage signal and connects it to the drive module of the signal processing circuit; Step 3: The drive module amplifies the voltage signal into a strong voltage signal that can directly drive the actuator terminal, and inputs it into the drive port of the actuator terminal. Step 4: The actuator terminal, as a bionic body, receives the drive signal and performs the corresponding action. The dual synaptic transistors located in its key sensing parts monitor the feedback signals generated by the action in real time, including reaction force and displacement changes, and adaptively adjust the action amplitude, response speed and stability of the actuator terminal.

8. The adaptive control method for neuromorphic perception and motor reflexes according to claim 7, characterized in that: In step 1, the dual-synaptic transistor senses external mechanical stimulation and converts the mechanical signal into an electrical signal, including: When an external object only contacts the dual-synaptic transistor, an excitatory postsynaptic current signal is generated based on the synergistic mechanism of triboelectric charging and electrostatic induction. When an external object applies a pressing action to the dual-synaptic transistor, an inhibitory postsynaptic current signal is generated based on the piezoelectric polarization effect.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions that, when invoked, are used to perform the steps of the method of claim 7 or 8.

10. The application of an adaptive control system based on neuromorphic perception and motor reflexes, characterized by: This system has applications in fields including bionic prostheses, industrial flexible grippers, service robots, and medical assistive devices.

Images