基于神经反射的双足机器人步态控制的研究

Biped Robot Based on Neural Reflex Control of Gait

双足步行机器人因为其高度的非线性系统且无固定支撑底面,使得传统的控制策略或算法产生的步态控制运动十分的僵硬,缓慢,缺乏灵活的自组织能力,与真正生物步态存在很大差异;而生物能很好利用环境的反射自激行为产生有节律的协调运动从而适应多种复杂环境,文中基于神经反射原理设计了反射网络控制器,从而克服常规控制策略的步态规划的不足,提高了系统性能,消除了关节的抖动,能够很好地适应其外界环境干扰因素的影响,通过MATLAB和ADAMS联合动态仿真,证明了该控制方案的优越性,并得到很好的控制效果。

Biped walking robot because of its high degree of nonlinear systems with no fixed bottom support,the traditional control of gait motion is generated by the rigid,slow,lack of flexible self-organizing capacity,and there is a big difference in the real biological gait.Can make good use of the environment and biological self-excited behavior reflected the coordination of rhythmic movements to accommodate a variety of complex environments,the paper design principles based on the reflection reflex network controller,to overcome the conventional control strategy is not lack of planning to improve system performance,eliminating the joint jitter,can well adapt to its external environment interference factors,through the joint MATLAB and ADAMS dynamic simulation to prove the superiority of the control scheme and get a good result.