基于多传感器信息的双足机器人步行系统的研究

提出了一种基于多传感器信息和实时修正算法的双足机器人分级递阶行走控制结构。在系统设计中，充分考虑零力矩点（ZMP）稳定性规则以及机器人行走时所受的其它物理性约束条件。为提高机器人对不平整路面的鲁棒性和克服系统建模误差，行走控制系统中采用了鲁棒地面反力控制器和基于地面反力最优规划的姿态控制策略。

Sensor-based Locomotion Control System of Biped Robot

The hierarchical control system integrated with a direct and real time modified algorithms for biped robot based on information obtained from some sensors was described. The proposed approach uses the previously established ZMP(Zero Moment Point) theory for realizing stably dynamic walk. Meanwhile, other physical conditions are also satisfied during the locomotion. Since the stability of biped robot is intrinsically low, for improving the walking robustness to the irregular terrain and modeling error, a robust reactive force control and trunk's position/attitude control with optimal force distribution are also considered in the course of system design.