基于自适应网络的动态双足机器人模糊控制

提出一种基于步态规划分级结构的自适应网络模糊推理系统控制策略,该方法不需要确定双足机器人运动学和动力学模型.以一种动态双足机器人为例,建立机器人的Sugeno模糊模型,对机器人系统的不确定上界进行自适应参数估计,采用自适应控制器逼近未知不确定界,解决了一类非线性系统的稳定控制问题.控制器的设计只要求不确定性满足匹配条件,而无需知道不确定界,能够处理不确定参数变化范围更广的情况,减少控制系统设计中的保守性.设计的分级控制系统可以学习试验的输入输出数据,从而在动态平衡下进行行走.同时,模糊控制器的进一步在线学习能力可以显著地改善步行机器人的动态性能.

Fuzzy Control for Dynamic Biped Robot Based on Adoptive Network

Proposed an adaptive network fuzzy inference system control strategy based on hierarchy structure of gait planning,which do not require detailed kinematics or dynamic biped models.The method is applied to construct Sugeno fuzzy model for dynamic biped robot and adaptively estimated uncertain upper bounds of robot system parameters,then approached unknown upper bounds based adaptive controller.Designed controller required uncertainty up to matching conditions but not actual uncertain bounds,which can deal with the fact that uncertain parameters have larger extension.Designed hierarchical control system can use the experimental input-output data pairs for the biped robot learning and walking with dynamic balance.It is also shown that the further on-line self-learning capability of the fuzzy controller can markedly improve the dynamic walking performance of the biped robot.