基于吸盘负压吸附的六足爬墙机器人关节转矩优化分配  

Joint Torque Distribution of a Six-legged Wall-climbing Robot with Negative Pressure Adsorption Sucker

作　　者：魏武[1] 李金龙[1] 任回兴 

机构地区：[1]华南理工大学,广州510640 [2]中交公路养护工程技术有限公司,北京100084

出　　处：《中国机械工程》2013年第10期1289-1295,共7页China Mechanical Engineering

基　　金：国家自然科学基金资助重点项目(60736024);交通运输部西部交通建设科技项目(B1110210)

摘　　要：针对采用吸盘负压吸附的六足爬墙机器人在多种运动模式下的关节转矩分配问题,提出基于总能耗目标优化的转矩分配方法。在确定吸盘翻转力矩约束、实现关节转矩与足底接触力转换的基础上,推导机器人动力学模型,进而用Moor-Penrose逆矩阵法和简体爬山算法求解总能耗最优目标下的机器人总功率与机器人各运动模式下各位姿(表现为各关节转角)之间的关系。该方法能有效适应爬墙机器人的多种运动模式,提高了机器人的环境适应能力。实验样机仿真结果表明,所提出的算法简单有效、实现方便,并且能有效确定站立模式最小总能耗位姿,以及在设定的侧立安全指标范围内的侧立最安全位姿。This paper focused on the joint torque distribution problem of a six-legged wall- climbing robot with negative pressure adsorption sucker. A torque distribution method was proposed based on energy consumption optimization. The robot＇s dynamics models were derived on the basis of establishing the sucker overturning moment constraints, and the conversion between joint torque and the foot contact force, was completed, employing the Moor-Penrose inverse matrix method and the simple hill-climbing algorithm. These algorithms were implemented to calculate the relationship be- tween the total energy consumption and the pose of the robot during various movement patterns.

关 键 词：吸盘负压吸附 六足爬墙机器人 关节转矩分配 Moor-Penrose逆矩阵法 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]