【作者】 周坤玲；

【导师】 张秀丽；

【作者基本信息】 北京交通大学 ， 机械工程（专业学位）， 2013， 硕士

【摘要】 从仿生学的角度出发,通过调研足式动物行走和奔跑的运动模式及特点,定性得出可应用于四足机器人运动规划的仿生学依据。基于此提出一种非匀速足端轨迹的生成方法,并设计了动力学仿真对比实验。研究结果表明,基于仿生学的足端轨迹规划可有效的减小四足机器人足端打滑程度,提高行走速度；且非匀速的轨迹规划可有效的降低四足机器人着地时足底受到的地面冲击力,对改善机器人trot步态下的行走效果,提高机器人行走的稳定性有重要作用。论文完成的工作如下：1)进行了仿生学调研,总结出足式动物行走和奔跑的运动特点,得出在规划四足机器人的足端轨迹时应考虑的因素及基本原则。2)基于仿生学的研究结论,提出了一种非匀速足端轨迹的生成方法,即,给定轨迹上2个基本点和其他n-2个点的时间和相位,求解φ角的n-1次多项式表达式,生成用极坐标表示的速度连续变化的足端轨迹表达式。3)采用运动学逆解和CPG相结合的控制方法规划了四足机器人的关节运动轨迹,建立了单腿轨迹模型和四足机器人CPG网络模型。4)设计了四足机器人运动的动力学仿真对比实验,分别针对足端轨迹形状、速度变化对四足机器人行走效果的影响进行了分析。5)编写了“四足仿生机器人轨迹规划计算软件”,具有计算足端工作空间、足端轨迹形状、各关节轨迹拟合曲线等功能。更多还原

【Abstract】 Walking movement of several typical quadrupeds is investigated. The results are used as references for gait planning of a quadruped robot, aiming to increase its walking speed. Based on the conclusions, a biologically-inspired non-uniform trajectory planning algorithm is presented for a16-DOF robot. Dynamic walking simulations are conducted. The simulation results show that the biologically-inspired trajectory planning algorithm can reduce foot sliding, and improve walking speed of quadruped robot effectively. Moreover, the non-uniform trajectory planning algorithm can reduce the impact force when the feet touch down the ground.Our work includes:1) Investigate the motion of several typical quadrupeds, obtain the factors and basic principles that is important for a successful walking.2) Propose a biologically-inspired trajectory planning algorithm represented in polar coordinate, which can generate non-uniform velocity for the feet of a16-DOF quadruped robot.3) Build a joint trajectory generation model using inverse kinematics and central pattern generator for the quadruped robot. The CPG compose of single leg oscillator and CPG network.4) Conduct the dynamic simulation experiments and analyze the simulation results by relating the robot walking speed with two factors:the foot trajectory shape and velocity distribution.5) Compile a trajectory planning software for quadruped robots.更多还原