An artificial neural network circuit based on memristive Pavlovian associative memory

Abstract

The invention discloses a memristor based artificial neural network circuit for Pavlov associative memory. The memristor based artificial neural network circuit comprises a synaptic circuit, a nerve cell circuit and a synaptic weight control circuit, wherein the synaptic circuit comprises n input ends, n output ends and (2x+1) control ends; the n input ends are used for separately receiving n input voltages; the nerve cell unit comprises n input ends and one output end, wherein the n input ends are in one-to-one correspondence connection with the n output ends of the synaptic circuit, and the output end is used for outputting excitation voltage Vout; the synaptic weight control circuit comprises n input ends, one feedback end and (2x+1) output control ends, wherein the n input ends are in one-to-one correspondence connection with the n input ends of the synaptic circuit, the feedback end is connected to the output end of the nerve cell unit, and the (2x+1) output ends are separately in one-to-one correspondence connection with the (2x+1) control ends of the synaptic circuit; and the synaptic weight control circuit is used for outputting a corresponding control signal according to the input state and the feedback state to regulate the resistance value of a memristor.

Description

Technical field [0001] The invention belongs to the field of analog-to-digital circuits, and more specifically, relates to an artificial neural network circuit based on Pavlov's associative memory based on memristive. Background technique [0002] In 1971, Professor Cai Shaotang of the University of California at Berkeley predicted from the perspective of symmetry that in addition to resistance, capacitance, and inductance, there should be a fourth basic element in electronic circuits—memristor. Cai Shaotang pointed out that voltage v, current i, charge q and magnetic flux There should be six mathematical relationships between these four basic circuit variables: current is defined as the rate of change of charge with respect to time i=dq / dt; voltage is defined as the rate of change of magnetic flux with respect to time Resistance is defined as the rate of change of voltage with current R=dv / di; capacitance is defined as the rate of change of charge with voltage C=dq / dv; inductan...

AI Technical Summary

Problems solved by technology

[0008] For the defects of the prior art, the purpose of the present invention is to provide a kind of artificial neural network circuit based on memristive Pavlovian associative memory, aiming at solving the problem that there is no forgetting function in the prior art

Images