一种用于多足步行机器人步态控制的CPG模型  

A CPG model of multi-legged walking robot gait controlling

作　　者：袁鹏[1] 

机构地区：[1]中国海洋大学工程学院,青岛266100

出　　处：《制造业自动化》2007年第10期34-39,共6页Manufacturing Automation

摘　　要：基于对生物节律性周期运动的CPG控制的概念,采用Lakshmanan和Murali的双变量简化Hodgkin-Huxley振荡神经元模型作为CPG的振子。对该模型进行了不动点及稳定性分析和分岔分析,指出了其二次Hopf分岔特性和振荡条件。在此基础上,建立了一种用于八足步行机器人步态控制的CPG模型。通过分阶段遗传算法,分别针对几种步态进行了参数优化,通过仿真验证了所提出的CPG模型的可行性。Based on the anatomy fact that the rhythmic motor pattern is controlled by CPG function modular, the conception of CPG control in multi-legged gait is presented. Neuraxon＇ s nonlinear oscillation model given by Lakshmanan and Murali is analyzed, Hopf bifurcation phenomenon and effects of parameters on its oscillation behavior are discussed. A CPG model for multi-legged walking robot＇ s gait control is made by adopting the neuron oscillator, parameters of the model is optimized by using multi-stage genetic algorithm and some gaits are generated using this model by simulation. Simulation result proved the constructed CPG model is feasible.

关 键 词：CPG 步态 多足步行机器人 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]