基于中枢神经模式产生器机器人的控制网络

Control Network for Humanoid Robots Based on Central Pattern Generator

分析了基于中枢神经模式产生器(Central Pattern Generator,CPG)的仿人机器人控制网络系统结构的特点,介绍了振荡器的数学模型。研究了CPG网络中各神经元的刺激方式,采用Hopf非线性振荡器构造神经元,模仿人类的行走步态,设计一种6关节控制网络。计算仿真中该网络输出信号稳定,运动节奏符合设计要求。最后,应用一仿人机器人完成了实验,提高了其行走的速度和稳定性,验证了该网络的有效性。

This paper analyzes the characters of control network framework based on Central Pattern generator (CPG) for humanoid robots, introducesmathematical model of oscillators. It studies the methods of stimulation between neurons in the CPG network, applies Hopf non-linear oscillators torepresent neurons and designs a six-joint control network according to humans' walking gaits. In the simulation, the signal of this network is stable and adesired dynamic pattern is obtained. Finally, a experiment with a humanoid demonstrates that this network is useful and increases the speed and stabilityof the robot.