A predictive circuit based on memristive neural network

Abstract

The invention provides a prediction circuit based on a memristor neural network, and solves the problem that an existing artificial neural network depends on a computer and needs to consume a large amount of calculation force. The memristor neural network comprises a memristor neural network module, an iteration device, a signal input module and a signal output module, an input terminal of the signal input module is connected with an input signal; an output terminal of the signal input module is connected with a first input terminal of the iteration device; A second input terminal of the iteration device is connected with an output terminal of the memristor neural network module, an output terminal of the iteration device is connected with an input terminal of the memristor neural networkmodule and an input terminal of the signal output module, and an output terminal of the signal output module outputs an output signal of the circuit. The prediction result output by the method is intelligently judged on the input information according to the information which is trained and stored in the memristor neural network, and the memristor neural network circuit with the prediction function has very important practical significance.

Description

technical field [0001] The invention relates to the technical field of digital-analog circuits, in particular to a prediction circuit based on a memristive neural network. Background technique [0002] In recent years, artificial neural network technology has developed rapidly. Artificial neural networks have been used to solve a variety of problems. They perform well in various engineering fields such as pattern recognition, automatic control, prediction and estimation, biology, medicine, and economy, and solve many problems that cannot be solved by traditional computing methods. question. However, most of the current artificial neural networks are implemented based on computer programming, and they still run on computers based on the traditional von Neumann architecture, consuming a lot of computing power. [0003] In 2008, Hewlett-Packard prepared a resistor with memory properties in the laboratory, and published an article in Nature magazine saying that this is the fif...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the technical problem that the existing artificial neural network consumes a large amount of computing power, the present invention proposes a prediction circuit based on the memristive neural network, which can make intelligent judgments on the information input into the circuit, thereby outputting a signal representing the prediction result

Images