作　　者： (李娟）; (蔡校蔚）; (李伟达）; (张宵）; (朱其欢）;

机构地区： 苏州大学机电工程学院,江苏苏州215021

出　　处： 《机电工程》 2017年第9期1075-1080,共6页

摘　　要： 针对步态规划中仅用几个参数确定机器人关键位置、再对关键位置进行插值或多阶曲线拟合而导致的实际重心轨迹与目标轨迹相差较大的问题,对下肢外骨骼机器人步态规划方法进行了研究,提出了一种逆向的利用遗传算法直接根据重心求解关节位置的方法。采用D-H法对下肢外骨骼机器人进行了运动学建模,得到了重心与机器人关节位置之间的关系。基于稳定裕度最大规划了机器人的重心轨迹。在几何关系约束下采用遗传算法求解了各关节位置,并根据逆运动学模型得到各关节转角值,使实际重心轨迹与目标轨迹高度吻合。最后,进行了机器人带动穿戴者行走实验。实验结果证明:该规划方法具有可行性和有效性。 In order to solve the problem of large error between the actual and objective COG（ center of gravity） trajectory,which caused by using only a few parameters to determine the key positions of the robot and then interpolating the key positions,a gait planning method of the lower limb exoskeleton robot was investigated. The genetic algorithm was used to solve the joints position directly based on the COG in this method. The kinematics model of the robot using D-H method was established,and the relationship between the COG and the joints position was obtained. The COG trajectory of the robot was planned based on the maximum stability margin. Under geometric constraint,joints position was obtained using the genetic algorithm and each joint angle was calculated according to the inverse kinematics model. The method made the actual and objective COG trajectories be highly consistent. Finally,the robot walking experiment was carried out to verify the reasonableness of the method. The experimental result shows the feasibility and effectiveness of the proposed method.

关 键 词： 下肢外骨骼 步态规划 重心 遗传算法