一种新型人工神经元振荡器的设计与应用

神经元振荡器在动物神经系统感知、运动、记忆等方面发挥重要作用，设计人工神经元振荡器不但可用研究生物神经元振荡器机理，而且可用于设计仿生机器人的仿生神经网络控制系统。本文首先提出了一种新型人工神经元振荡器，该振荡器由两个神经元构成，神经元自身存在非线性反馈联接，两者之间为线性突触联接。然后证明了其存在稳定的、近似圆形的极限环，振荡的收敛速度、幅度和频率分别由动力学方程中的三个参数独立控制。最后介绍了采用这种振荡器设计的鱼形机器人新型仿生神经网络控制系统，该控制系统可以实现对包括启停、稳定游动等动作的良好控制。

Design and Application of a Novel Artificial Neural Oscillator

This paper presents and implements a novel artificial neural oscillator which consists of two neurons and simple connections.The dynamic analysis on the oscillator indicates that the oscillatory convergence speed,amplitude and frequency are individually controlled by three parameters of the dynamic formula.A bionic neural network control system based on the oscillator is designed to control a fish-robot.Experiments are carried out to validate that the neural system can effectively control several motions such as start,stop and cruise,and it promotes the robotic imitation of the living being to progress from "similar in shape" to "alike in spirit".