双足爬壁机器人壁面凹过渡步态规划研究

针对腿足式爬壁机器人在壁面过渡时的步态规划问题,以一种真空吸附式双足爬壁机器人为研究对象,在步态分析的基础上,基于有限状态机建立了机器人的步态模型,进而提出了基于加权插值和BP神经网络的双足爬壁机器人壁面凹过渡在线步态规划算法,为提高机器人壁面过渡的自主控制能力奠定了基础.仿真分析和实验结果表明,该步态规划算法对于实际的机器人系统是有效的和可行的.

In order to solve the gait planning issue in wall transitions for bipedal wall-climbing robots fitted with vacuum suction cups, a gait model based on finite state machine was established after deep analysis of the robot’s gait. On this basis, an on-line gait planning algorithm for concave transitions was proposed based on weighted interpolation and BP neural networks. This lays a foundation for improving the ability of robots to transit complex walls under automatic control. Simulations and experiments show that the on-line gait planning algorithm can successfully help wall-climbing robots complete a concave transition.

国家高技术研究发展计划资助项目(2004AA4201102)