Spiking feedforward network hippocampus function simulation system based on FPGA

Abstract

The invention provides a Spiking feedforward network hippocampus function simulation system based on an FPGA. The simulation system is characterized by comprising an FPGA development board and an upper computer, wherein the FPGA development board comprises a feedforward neural network, a NiosII soft core controller and a USB interface module, and the simulation system also comprises a three-layer feedforward network which is programmed, downloaded and compiled through a VHDL language and operates in the FPGA development board and simulates the hippocampus neuron function, an initial value signal transmitting module, a first sensing layer, a second hippocampus function layer and a third action output layer of the network, an LIF neuron assembly line model, a judgment control module and an STDP, namely a synaptic control updating matrix; the LIF neuron pipeline model, the judgment control module and the STDP synaptic update matrix in each layer of feedforward network are realized by adopting the VHDL language programming; the judgment control module is responsible for controlling the output of each layer of signals, and the control signals sent by the judgment control module are transmitted to each layer of neurons through the sensing layer signal transmission path, the hippocampus function layer signal transmission path and the action output layer signal transmission path.

Description

technical field [0001] The invention relates to biomedical engineering technology, in particular to an FPGA-based Spiking feedforward network hippocampus function simulation system. Background technique [0002] The hippocampus, present in all mammals and capable of retaining long-term memory, is one of the most important structures of the brain, responsible for episodic memory and task learning. The study found that when solving a current behavioral task, hippocampal neurons will be affected by the surrounding environment. According to the experiments done by the researchers on rats in a T-shaped maze, when the rats turned left or Two-thirds of hippocampal neurons fire differently when turning right; many studies have shown that hippocampal neurons encode memories in the brain both prospectively and retrospectively and are selective for specific events before firing . However, although there are many studies on hippocampal neurons, so far they have only analyzed the pheno...

AI Technical Summary

Problems solved by technology

[0005] The existing technology is still in the basic stage, so there are the following disadvantages: there is no FPGA-based simulation system with complete functions dedicated to exploring the function of hippocampal neurons; the hardware simulation neuron model implemented by FPGA is simple in structure and low in accuracy; The computer interface is not yet perfect, and it is not convenient for real-time control and data analysis. Therefore, it is difficult to simulate and analyze the dynamic characteristics of the hippocampal neuron network with FPGA

Images