Artificial neuron circuit based on volatile threshold switching device

Abstract

The present application relates to the field of artificial neuromorphics, and provides an artificial neuron circuit based on a volatile threshold switching device. A membrane potential generation circuit is used for generating different membrane potentials under different input currents and different external reset voltages; a slow variable generation circuit is used for generating a slow variable; a membrane potential reset circuit comprises a volatile threshold switching device and an external reset voltage excitation source; the external reset voltage excitation source is used for providing different external reset voltages, so that the membrane potential and the slow variable are changed, and a spiking behavior of biological neurons is simulated; and the volatile threshold switching device is used for realizing a membrane potential reset function. Compared with traditional CMOS-circuit-based neurons, the neuron circuit provided by the present application requires a greatly reduced number of transistors and reduced hardware and power consumption overhead, has the characteristics of simple structure, high flexibility and rich functions, and is beneficial to large-scale integration in a hardware pulse neural network.

Description

An artificial neuron circuit based on a volatile threshold switching device TECHNICAL FIELD

[0001] The present application belongs to the field of artificial neural morphology, and more particularly, relates to an artificial neuron circuit based on a volatile threshold switching device. BACKGROUND

[0002] Bio-inspired artificial neural morphology computing architecture is considered to be an effective means to overcome the von Neumann bottleneck. Through the cross-connection of artificial neurons and synapses, high energy efficiency, low latency and high parallel brain-like computing functions can be achieved, which has great application prospects in the field of artificial intelligence.

[0003] Biological neurons have highly nonlinear and rich dynamic characteristics, and firing behavior is one of the main electrical activities of biological neurons, which plays a crucial role in information transmission, processing and coding. Known biological neuron firing behaviors include regular spiking (RS), intrinsically bursting (IB), chattering (CH) and fast spiking (FS).

[0004] In recent years, researchers have built various analog and digital circuits to simulate the high-order output characteristics of biological neurons. Among them, the artificial neuron represented by the Izhikevich model is based on CMOS circuit construction. This model is very suitable for large network simulation, but this kind of neuron circuit at least needs dozens of transistors, the circuit structure is complex, and the power consumption and area overhead are large, which is not conducive to large-scale integration.

[0005] Volatile threshold switching devices have been widely used in the construction of artificial neural morphology hardware due to their simple structure and rich dynamic characteristics. Currently, neuron circuits based on volatile threshold switching devices have been reported, but most of the circuits are based on a simplified model of leaky integrate-and-fire (LIF) neurons, with single firing mode and unable to simulate the multiple firing behaviors of biological neurons, making it difficult to construct a biomimetic high-order brain-like system.

[0006] Therefore, there is an urgent need for a simple artificial neuron circuit to realize multiple firing behaviors of biological neurons.

[0007] SUMMARY

[0008] In view of the defects of the prior art, the purpose of the present application is to provide an artificial neuron circuit based on a volatile threshold transition device, aiming to solve the problem that the current neuron circuit based on a volatile threshold transition device is based on a simplified model of a leaky integrate-and-fire neuron, the discharge mode is single, and the multiple discharge behaviors of a biological neuron cannot be simulated, thereby leading to the difficulty in constructing a biomimetic high-order brain-like system.

[0009] To achieve the above-mentioned purpose, in a first aspect, the present application provides an artificial neuron circuit based on a volatile threshold transition device, comprising a membrane potential generation circuit, a slow variable generation circuit and a membrane potential reset circuit.

[0010] The membrane potential generation circuit is used to generate different membrane potentials under different input currents and different external reset voltages; the membrane potential generation circuit is connected with the slow variable generation circuit; and the slow variable generation circuit is used to generate a slow variable.

[0011] The membrane potential reset circuit is connected with the membrane potential generation circuit and comprises a volatile threshold transition device and an external reset voltage excitation source; the volatile threshold transition device is connected with the external reset voltage excitation source; the external reset voltage excitation source is used to provide different external reset voltages to change the membrane potential and the slow variable and simulate the discharge behavior of a biological neuron; and the volatile threshold transition device is used to realize the membrane potential reset function.

[0012] The external reset voltage is less than the holding voltage of the volatile threshold transition device.

[0013] Further preferably, the membrane potential generation circuit comprises a first capacitor, a first NMOS transistor, a second PMOS transistor, a third PMOS transistor and a fourth NMOS transistor; one end of the first capacitor is connected with the gate of the first NMOS transistor, and the other end is connected with the source of the first NMOS transistor and grounded; the first capacitor is connected with the volatile threshold transition device; the first capacitor is used to drive the first NMOS transistor to generate a current by using a voltage; the second PMOS transistor and the third PMOS transistor constitute a current mirror circuit, which is used to mirror the current of the first NMOS transistor to the third PMOS transistor in proportion; the difference between the currents of the third PMOS transistor and the fourth NMOS transistor flows into the first capacitor, which is used to constitute a positive feedback loop to generate different membrane potentials under different input currents and different external reset voltages.

[0014] Further preferably, the slow variable generating circuit comprises a second capacitor, a fifth PMOS transistor and a sixth NMOS transistor; the fifth PMOS transistor and the second PMOS transistor constitute another set of current mirror circuit for mirroring the current of the first NMOS transistor to the fifth PMOS transistor in proportion; the second capacitor has one end connected to the gate of the fourth NMOS transistor and the other end connected to the source of the fourth NMOS transistor and grounded; the second capacitor is used to drive the fourth NMOS transistor to generate current by voltage; the gate and the source of the sixth NMOS transistor are short-circuited, so that the difference between the currents of the fifth PMOS transistor and the sixth NMOS transistor flows into the second capacitor to generate a slow variable.

[0015] Further preferably, when the difference between the membrane potential and the external reset voltage is greater than the threshold voltage of the volatile threshold switching device, the volatile threshold switching device is switched from a high resistance state to a low resistance state, and the first capacitor is used to discharge through the volatile threshold switching device; when the difference between the membrane potential and the external reset voltage is less than the holding voltage of the volatile threshold switching device, the volatile threshold switching device is switched from a low resistance state to a high resistance state, so as to realize the reset of the membrane potential.

[0016] Further preferably, the volatile threshold switching device is a metal insulator transition-based threshold switching device, a metal conductive filament diffusion-based threshold switching device and a chalcogenide-based threshold switching device.

[0017] The metal insulator transition-based threshold switching device adopts NbO x or VO x Mott insulator material; the metal conductive filament diffusion-based threshold switching device adopts Ag or Cu active metal material to form conductive filaments; the chalcogenide-based threshold switching device adopts single-element or multi-element doped tellurium-based, selenium-based or sulfur-based OTS material; wherein x>0.

[0018] Further preferably, the first capacitor and the second capacitor are fixed capacitors or variable capacitors, and the value range is 1 fF to 1 μF.

[0019] Further preferably, the biological neuron firing behavior includes regular firing, initial burst firing, burst firing and fast firing.

[0020] Further preferably, the state equation of the artificial neuron circuit is:

[0021] Wherein, V and U are the membrane potential and the slow variable respectively; I in is the input current; I3 and I5 are the currents of the third PMOS transistor and the fifth PMOS transistor respectively; I4 and I6 are the currents of the fourth NMOS transistor and the sixth NMOS transistor; C v is the capacitance value of the first capacitor; Cu a capacitance value of the second capacitor.

[0022] In a second aspect, the application provides a method for artificial neuron firing based on a volatile threshold transition device, which specifically comprises the following steps:

[0023] Step S1: Under the application of an external reset voltage and an input current, the membrane potential is rapidly increased by the positive feedback loop of the membrane potential generation circuit. When the difference between the membrane potential and the external reset voltage is greater than the threshold voltage of the volatile threshold transition device, the volatile threshold transition device is converted from a high resistance state to a low resistance state, providing a low resistance discharge path for the first capacitor, so that the membrane potential rapidly decreases. The current of the first NMOS transistor rapidly increases and then decreases, and through the current mirror circuit, the current corresponding to the fifth PMOS transistor also rapidly increases and then decreases, charging the second capacitor and increasing the slow variable;

[0024] Step S2: When the difference between the membrane voltage and the external reset voltage is less than the holding voltage of the volatile threshold transition device, the volatile threshold transition device is converted from a low resistance state to a high resistance state. The membrane potential is continuously charged and reset, and the slow variable is continuously increased, so that the neuron firing enters the initial stage;

[0025] Step S3: When the slow variable increases to the maximum value, the currents in the fourth NMOS transistor and the sixth NMOS transistor increase, so that the change rate of the membrane potential and the change rate of the slow variable are both less than 0, and the membrane potential and the slow variable decrease;

[0026] Step S4: During the decrease of the slow variable, the change rate of the membrane potential is greater than 0, and the change rate of the slow variable is less than 0. The membrane potential increases, and the slow variable continuously decreases. The neuron firing behavior enters the refractory period.

[0027] Step S5: The slow variable continuously decreases, and the change rate of the membrane potential and the change rate of the slow variable are both greater than 0, so that the membrane potential is continuously charged and reset. Go to step S1 until the artificial neuron firing is stopped.

[0028] Further preferably, the biological neuron firing behavior includes regular firing, initial burst firing, burst firing, and rapid firing, and the external reset voltage required for rapid firing is greater than that required for burst firing; the external reset voltage required for burst firing is greater than that required for initial burst firing; and the external reset voltage required for initial burst firing is greater than that required for regular firing.

[0029] It should be noted that the rapid firing mentioned in the application refers to a firing mode in which the increase of the slow variable cannot make the change rate of the membrane potential less than 0 when the external reset voltage exceeds a certain voltage size, and there is no refractory period in the entire neuron firing process.

[0030] Overall, compared with the prior art, the above technical solutions conceived by the present application have the following beneficial effects:

[0031] The present application provides an artificial neuron circuit based on a volatile threshold switching device, comprising a membrane potential generation circuit, a slow variable generation circuit and a membrane potential reset circuit; the unique electrical characteristics of the volatile threshold switching device are utilized to realize the membrane potential reset function, thereby reducing the number of transistors; by changing only one external reset voltage, the artificial neuron circuit can simulate the known four biological neuron firing behaviors in the same circuit. Compared with the conventional CMOS circuit-based neuron, the neuron circuit provided by the present application greatly reduces the number of transistors, reduces the hardware and power consumption, has the characteristics of simple structure, strong flexibility and rich functions, and is conducive to large-scale integration in a hardware pulse neural network. BRIEF DESCRIPTION OF DRAWINGS

[0032] Fig. 1 is a structural schematic diagram of an artificial neuron circuit based on a volatile threshold switching device provided by an embodiment of the present application;

[0033] Fig. 2 is a voltage-current scanning curve diagram of a SiTe chalcogenide threshold switching device provided by an embodiment of the present application;

[0034] Fig. 3(a) is a time-domain waveform diagram of regular firing of an artificial neuron circuit based on a chalcogenide threshold switching device provided by an embodiment of the present application;

[0035] Fig. 3(b) is a phase diagram of regular firing of an artificial neuron circuit based on a chalcogenide threshold switching device provided by an embodiment of the present application;

[0036] Fig. 4(a) is a time-domain waveform diagram of initial cluster firing of an artificial neuron circuit based on a chalcogenide threshold switching device provided by an embodiment of the present application;

[0037] Fig. 4(b) is a phase diagram of initial cluster firing of an artificial neuron circuit based on a chalcogenide threshold switching device provided by an embodiment of the present application;

[0038] Fig. 5(a) is a time-domain waveform diagram of cluster firing of an artificial neuron circuit based on a chalcogenide threshold switching device provided by an embodiment of the present application;

[0039] Fig. 5(b) is a phase diagram of cluster firing of an artificial neuron circuit based on a chalcogenide threshold switching device provided by an embodiment of the present application;

[0040] Fig. 6(a) is a time-domain waveform diagram of fast firing of an artificial neuron circuit based on a chalcogenide threshold switching device provided by an embodiment of the present application;

[0041] Fig. 6(b) is a phase diagram of fast firing of an artificial neuron circuit based on a chalcogenide threshold switching device provided by an embodiment of the present application;

[0042] Figure 7 is a one-dimensional summary diagram of four discharge behaviors of the artificial neuron circuit based on the sulfur threshold switching device provided in the embodiments of the present application.

DETAILED DESCRIPTION

[0043] The embodiments of the present application will be described below with reference to the accompanying drawings in the embodiments of the present application.

[0044] In order to meet the needs of developing artificial neuron circuits with simple structure and strong flexibility, the present application provides an artificial neuron circuit based on a volatile threshold switching device, comprising a membrane potential generation circuit, a slow variable generation circuit and a membrane potential reset circuit.

[0045] The membrane potential generation circuit comprises a first capacitor C v , a first NMOS transistor M1, a fourth NMOS transistor M4, a second PMOS transistor M2 and a third PMOS transistor M3; one end of the first capacitor C v is connected with the gate of the first NMOS transistor M1, and the other end is connected with the source of the first NMOS transistor M1 and grounded, and the voltage on the first capacitor C v drives the first NMOS transistor M1 to generate current; the second PMOS transistor M2 and the third PMOS transistor M3 constitute a current mirror circuit, which mirrors the current of the first NMOS transistor M1 to the third PMOS transistor M3 by a certain proportion; the difference between the currents of the third PMOS transistor M3 and the fourth NMOS transistor M4 flows into the first capacitor C v , thereby forming a positive feedback loop to generate the membrane potential V.

[0046] The slow variable generation circuit comprises a second capacitor C u , a fifth PMOS transistor M5 and a sixth NMOS transistor M6; the fifth PMOS transistor M5 and the second PMOS transistor M2 constitute another current mirror circuit; the current of the first NMOS transistor M1 is mirrored to the fifth PMOS transistor M5 by a certain proportion; one end of the second capacitor C u is connected with the gate of the fourth NMOS transistor M4, and the other end is connected with the source of the fourth NMOS transistor M4 and grounded, and the voltage on the second capacitor C u drives the fourth NMOS transistor M4 to generate current; the gate and source of the sixth NMOS transistor M6 are short-circuited, so that the difference between the currents of the fifth PMOS transistor M5 and the sixth NMOS transistor M6 flows into the second capacitor C u , thereby generating the slow variable U.

[0047] The membrane potential reset circuit comprises a volatile threshold switching device and an external reset voltage V c; the volatile threshold transition device is connected to the first capacitor C v at one end and to an external reset voltage V c at the other end.

[0048] In some embodiments, the volatile threshold transition device is capable of instant switching from a high resistance state to a low resistance state; specifically, when the voltage applied to the volatile threshold transition device increases from zero to a threshold voltage V th , the device is in a high resistance state; when the voltage applied to the volatile threshold transition device exceeds the threshold voltage V th , the device is in a low resistance state; when the voltage decreases and is less than a holding voltage V h , the device spontaneously returns to the high resistance state, wherein the threshold voltage V th is greater than the holding voltage V h .

[0049] In some embodiments, the membrane potential reset circuit utilizes the electrical characteristics of the volatile threshold transition device, when the difference between the membrane potential V and the external reset voltage V c is greater than the threshold voltage V th of the volatile threshold transition device, the volatile threshold transition device transitions from a high resistance state to a low resistance state, and the capacitor discharges through the volatile threshold transition device; when the difference between the membrane potential V and the external reset voltage V c is less than the holding voltage V h of the volatile threshold transition device, the volatile threshold transition device transitions from a low resistance state to a high resistance state, thereby achieving the membrane potential reset function.

[0050] In some embodiments, the external reset voltage V c is less than the holding voltage V h of the volatile threshold transition device; the capacitors C v and C u are fixed capacitors or variable capacitors, and the value of the capacitors ranges from 1 fF to 1 μF.

[0051] In some embodiments, the above-mentioned volatile threshold transition device can be a metal-insulator transition-based, metal-conductive filament diffusion-based, or chalcogenide-based threshold transition device; specifically, the metal-insulator transition-based threshold transition device uses NbO x or VO x Mott insulator material; the metal-conductive filament diffusion-based threshold transition device uses Ag or Cu active metal material to form conductive filaments; the chalcogenide-based threshold transition device uses single-element or multi-element doped tellurium-based, selenium-based, or sulfur-based OTS material; wherein x > 0.

[0052] In some embodiments, the artificial neuron circuit, under a certain current input, changes the external reset voltage V cThe membrane potential generating circuit outputs different membrane potentials V, so as to simulate four known biological neuron firing behaviors; specifically, the four biological neuron firing behaviors include regular firing, initial burst firing, burst firing and fast firing.

[0053] The above technical solutions are described in detail below in combination with specific embodiments.

[0054] The embodiment provides an artificial neuron circuit based on a volatile threshold switching device, and a structural schematic diagram is as shown in FIG. 1; wherein the volatile threshold switching device takes a chalcogenide threshold switching device as an example, and FIG. 2 provides a voltage-current scanning curve diagram of a SiTe chalcogenide threshold switching device, the chalcogenide threshold switching device has a threshold voltage V th of 1.14 V and a holding voltage V h of 0.9 V.

[0055] By using a general circuit analysis method, the state equation of the neuron circuit can be obtained from the topological structure of the artificial neuron circuit as follows:

[0056] Wherein V and U are the membrane potential and the slow variable respectively; I in is the input current; I3 and I5 are the currents of the third PMOS transistor M3 and the fifth PMOS transistor M5 respectively; I4 and I6 are the currents of the fourth NMOS transistor M4 and the sixth NMOS transistor M6 respectively; specifically, the equation (1) can be obtained from the membrane potential generating circuit, and the equation (2) can be obtained from the slow variable generating circuit.

[0057] Furthermore, the process of simulating the biological neuron firing behavior is described in detail; under the application of a certain external reset voltage V c and input current I in , the positive feedback loop of the membrane potential generating circuit makes the membrane potential V rapidly rise, when the difference between the membrane potential V and the external reset voltage V c is greater than the threshold voltage V th of the volatile threshold switching device, the volatile threshold switching device is switched from a high resistance state to a low resistance state, so as to provide a low resistance discharging path for the capacitor C v , so that the membrane potential V rapidly decreases to the set voltage c; because the membrane potential V rapidly rises and then decreases, the current of the first NMOS transistor M1 rapidly increases and then decreases, through the current mirror circuit, the current I5 also rapidly increases and then decreases, and a certain charge is charged into the capacitor C u , so that the slow variable U increases by an increment voltage d, and the reset mechanism is as follows:

[0058] If V>V reset , then

[0059] wherein, V reset = V th + V c ; c and d increase with the increase of the external reset voltage V c .

[0060] When the difference between the membrane potential V and the external reset voltage V c is less than the holding voltage V h of the volatile threshold transition device, the volatile threshold transition device is transitioned from the low resistance state to the high resistance state; in the following period of time, the membrane potential V is continuously charged and reset, and the slow variable U is continuously increased, and the neuron firing enters the initial stage.

[0061] As can be seen from the state equation of the neuron circuit, when the slow variable U increases to a certain extent, the increase of the currents I4 and I6 driven by the slow variable U makes the change rate dV / dt of the membrane potential V and the change rate dU / dt of the slow variable U both less than 0, resulting in the decrease of the membrane potential V and the slow variable U at a certain rate; when the slow variable U decreases to a certain extent, the change rate dV / dt of the membrane potential V is slightly greater than 0, but the change rate dU / dt of the slow variable U is still less than 0, making the membrane potential V slowly increase and the slow variable U continuously decrease; the above is the refractory period of the neuron firing behavior; generally, the greater the voltage amount d of each increase of the slow variable U, the longer the refractory period of the neuron firing; after the neuron firing enters the refractory period, the membrane potential V and the slow variable U change at a certain rate, and when the slow variable U decreases to a certain extent, the change rate dV / dt of the membrane potential V and the change rate dU / dt of the slow variable U are both greater than 0, making the membrane potential V continuously charged and reset and the slow variable U continuously increased; when the slow variable U increases to a certain extent, the neuron re-enters the refractory period stage, and then the neuron firing behavior is generated in a cycle, and the neuron firing enters the stable stage. The above two stages constitute the neuron firing behavior of the artificial neuron circuit.

[0062] When the external reset voltage V c changes, the voltages c and d in equation (3) also change, thereby changing the neuron firing behavior;

[0063] When the external reset voltage V c is 0 and the input current I in = 400 nA, because the external reset voltage V csmall, the voltage d in the reset mechanism is small, the slow variable U after the initial stage of neuron firing is also small, the rate of change of membrane potential V dV / dt is slightly greater than 0, but the rate of change of slow variable U dU / dt is less than 0, resulting in slow rise of membrane potential V, slight decrease of slow variable U; after the above refractory period, the neuron firing enters the stable stage. The stable stage has more links of slow variable U decrease and increase than the initial stage, so the firing frequency of the stable stage is lower than that of the initial stage, and the neuron circuit presents regular firing behavior as shown in Fig. 3(a), and the phase diagram is as shown in Fig. 3(b);

[0064] When the external reset voltage V c = 300 mV and the input current I in = 400 nA, since the external reset voltage V c is large, the voltage d in the reset mechanism is large, the slow variable U after the initial stage of neuron firing is also large, the neuron firing enters the refractory period, the slow variable U decreases, and due to the change of the rate of change of membrane potential V dV / dt, the membrane potential V first decreases and then increases; the slow variable U decreases to a certain extent, so that the rate of change of membrane potential V dV / dt and the rate of change of slow variable U dU / dt are both greater than 0, but the membrane potential V is charged and reset once, and the slow variable U increases the voltage d, which makes the neuron firing re-enter the refractory period, and the neuron circuit presents the initial cluster firing behavior as shown in Fig. 4(a), and the phase diagram is as shown in Fig. 4(b);

[0065] When the external reset voltage V c = 450 mV and the input current I in = 400 nA, the neuron firing behavior is similar to the initial cluster firing, but different from the initial cluster firing, since the external reset voltage V c is larger than that of the initial cluster firing, the slow variable U after the initial stage of neuron firing is larger, so that the slow variable U decreases for a longer time in the refractory period, and the neuron needs to be charged and reset multiple times after the refractory period to re-enter the refractory period, and the neuron circuit presents the cluster firing behavior as shown in Fig. 5(a), and the phase diagram is as shown in Fig. 5(b);

[0066] When the external reset voltage V c = 500 mV and the input current I in = 400 nA, since the external reset voltage V c is the largest compared with the first three firing behaviors, the voltage c after each reset of the membrane potential V and the voltage V resetThe difference between the current I3 and I4 is always greater than 0 during the discharging process, at this time, the increase of the slow variable U cannot make the change rate dV / dt of the membrane potential V less than 0, that is, dV / dt is always greater than 0 during the entire discharging process, the neuron discharges without a refractory period, and the neuron circuit presents the fast discharging behavior shown in Fig. 6(a), and the phase diagram is shown in Fig. 6(b);

[0067] Fig. 7 is a one-dimensional summary diagram of four discharging behaviors of an artificial neuron circuit based on a sulfur threshold switching device, when the external reset voltage V c is in the range of 0-180 mV, the neuron circuit presents a regular discharging behavior; when the external reset voltage V c is in the range of 180-320 mV, the neuron circuit presents an initial cluster discharging behavior; when the external reset voltage V c is in the range of 320-470 mV, the neuron circuit presents a cluster discharging behavior; when the external reset voltage V c is in the range of 470-500 mV, the neuron circuit presents a fast discharging behavior;

[0068] The application provides an artificial neuron circuit based on a volatile threshold switching device, which comprises a membrane potential generating circuit, a slow variable generating circuit and a membrane potential reset circuit. The reset function is realized by using the unique electrical characteristics of the volatile threshold switching device, and the number of transistors is reduced. The circuit can simulate the four known biological neuron discharging behaviors in the same circuit by changing only one external reset voltage V c Compared with a traditional neuron based on a CMOS circuit, the neuron circuit provided by the application has the characteristics of simple structure, strong flexibility and rich functions, and the number of transistors required is greatly reduced, the hardware and power consumption are reduced, and the neuron circuit is beneficial to large-scale integration in a hardware pulse neural network.

[0069] It should be understood that expressions such as "include" and "may include" used in the present application represent the existence of the disclosed functions, operations or constituent elements, and do not limit one or more additional functions, operations and constituent elements.

[0070] In the description of the embodiments of the present application, it should be noted that, unless otherwise explicitly specified and limited, the term "connection" should be understood in a broad sense, for example, "connection" can be detachable connection, or can be non-detachable connection; can be direct connection, or can be indirect connection through an intermediate medium.

[0071] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An artificial neuron circuit based on a volatile threshold transition device, characterized by, The application relates to a threshold voltage variable neuron model circuit. The circuit comprises a membrane potential generating circuit, a slow variable generating circuit and a membrane potential resetting circuit. The membrane potential generating circuit is used for generating different membrane potentials under different input currents and different external reset voltages. The slow variable generating circuit is connected with the membrane potential generating circuit, and the slow variable generating circuit is used for generating a slow variable. The membrane potential resetting circuit is connected with the membrane potential generating circuit, and the membrane potential resetting circuit comprises a volatile threshold transition device and an external reset voltage excitation source. The volatile threshold transition device is connected with the external reset voltage excitation source, the external reset voltage excitation source is used for providing different external reset voltages, the membrane potential and the slow variable are changed, and the discharge behavior of a biological neuron is simulated; and the volatile threshold transition device is used for realizing the membrane potential resetting function. The external reset voltage is smaller than the holding voltage of the volatile threshold transition device.

2. The artificial neuron circuit according to claim 1, wherein, The membrane potential generating circuit comprises a first capacitor, a first NMOS transistor, a second PMOS transistor, a third PMOS transistor and a fourth NMOS transistor; one end of the first capacitor is connected with the gate of the first NMOS transistor, the other end of the first capacitor is connected with the source of the first NMOS transistor and grounded, the first capacitor is connected with the volatile threshold transition device, the first capacitor is used for driving the first NMOS transistor to generate a current by using a voltage, the second PMOS transistor and the third PMOS transistor constitute a current mirror circuit, the current of the first NMOS transistor is mirrored to the third PMOS transistor in proportion, and the difference between the currents of the third PMOS transistor and the fourth NMOS transistor flows into the first capacitor, thereby forming a positive feedback loop, and different membrane potentials are generated under different input currents and different external reset voltages.

3. The artificial neuron circuit according to claim 2, wherein, The slow variable generating circuit comprises a second capacitor, a fifth PMOS transistor and a sixth NMOS transistor; the fifth PMOS transistor and the second PMOS transistor constitute another current mirror circuit, the current of the first NMOS transistor is mirrored to the fifth PMOS transistor in proportion, one end of the second capacitor is connected with the gate of the fourth NMOS transistor, the other end of the second capacitor is connected with the source of the fourth NMOS transistor and grounded, the second capacitor is used for driving the fourth NMOS transistor to generate a current by using a voltage, and the gate and the source of the sixth NMOS transistor are short-circuited, so that the difference between the currents of the fifth PMOS transistor and the sixth NMOS transistor flows into the second capacitor, thereby generating a slow variable.

4. The artificial neuron circuit according to any one of claims 1 to 3, wherein When the difference between the membrane potential and the external reset voltage is greater than the threshold voltage of the volatile threshold transition device, the volatile threshold transition device is changed from a high resistance state to a low resistance state, and the first capacitor is used for discharging through the volatile threshold transition device; when the difference between the membrane potential and the external reset voltage is smaller than the holding voltage of the volatile threshold transition device, the volatile threshold transition device is changed from the low resistance state to the high resistance state, thereby realizing the membrane potential resetting.

5. The artificial neuron circuit according to any one of claims 1 to 3, wherein The volatile threshold transition device is a threshold transition device based on a metal insulator transition, a metal conductive filament diffusion or a chalcogenide. Threshold switching devices based on metal-insulator transition employ NbO x or VO x Mott insulator materials; The threshold transition device based on metal conductive filament diffusion uses Ag or Cu active metal material to form conductive filament; the threshold transition device based on chalcogenide uses single-element or multi-element doped tellurium-based, selenium-based or sulfur-based OTS material; wherein x>0.

6. The artificial neuron circuit of claim 3, wherein, The first capacitor and the second capacitor are fixed capacitors or variable capacitors, and the capacitance value ranges from 1 fF to 1 μF.

7. The artificial neuron circuit of claim 1, wherein, The biological neuron discharge behavior includes regular discharge, initial burst discharge, burst discharge and fast discharge.

8. The artificial neuron circuit according to claim 1 or 7, wherein, The state equation of the artificial neuron circuit is: where V and U are the membrane potential and slow variable, respectively; I in is the input current; I3and I5are the third and fifth currents, respectively. the current of the third PMOS transistor and the fifth PMOS transistor; I4 and I6 are the currents of the fourth NMOS transistor and the sixth NMOS transistor; C v is the capacitance value of the first capacitor; C u is the capacitance value of the second capacitor.

9. An artificial neuron firing method based on the artificial neuron circuit according to claim 3, characterized by, The method comprises the following steps: Step S1: Under the application of an external reset voltage and an input current, the membrane potential is rapidly raised through the positive feedback loop of the membrane potential generation circuit, and when the difference between the membrane potential and the external reset voltage is greater than the threshold voltage of the volatile threshold transition device, the volatile threshold transition device is switched from a high resistance state to a low resistance state, providing a low resistance discharge path for the first capacitor, so that the membrane potential rapidly decreases; the current of the first NMOS transistor rapidly increases and then decreases, and through the current mirror circuit, the current corresponding to the fifth PMOS transistor also rapidly increases and then decreases, charging the second capacitor and increasing the slow variable; Step S2: When the difference between the membrane voltage and the external reset voltage is less than the holding voltage of the volatile threshold transition device, the volatile threshold transition device is switched from a low resistance state to a high resistance state; the membrane potential is continuously charged and reset, and the slow variable is continuously increased, so that the neuron discharge enters the initial stage; Step S3: When the slow variable increases to the maximum value, the currents in the fourth NMOS transistor and the sixth NMOS transistor increase, so that the change rate of the membrane potential and the change rate of the slow variable are both less than 0, and the membrane potential and the slow variable decrease; Step S4: During the decrease of the slow variable, the change rate of the membrane potential is greater than 0, and the change rate of the slow variable is less than 0, so that the membrane potential rises and the slow variable continuously decreases, and the neuron discharge behavior enters the refractory period; Step S5: The slow variable continuously decreases, and the change rate of the membrane potential and the change rate of the slow variable are both greater than 0, so that the membrane potential is continuously charged and reset, and the process returns to step S1 until the artificial neuron discharge is stopped.

10. The artificial neuron firing method of claim 9, wherein, The biological neuron discharge behavior includes regular discharge, initial burst discharge, burst discharge and fast discharge, and the external reset voltage required for fast discharge is greater than that required for burst discharge; the external reset voltage required for burst discharge is greater than that required for initial burst discharge; the external reset voltage required for initial burst discharge is greater than that required for regular discharge.

Images