摘要
针对双足机器人逆运动学的数值解法中存在的雅可比矩阵奇异性和调节参数固定问题,提出了一种改进的求解方法.运用微分运动方程的近似解避开雅可比矩阵求逆,利用能够减小跟踪误差的自适应模糊控制法,调节自适应参数以使近似解任意逼近精确解,从而得到了精确性极高和强鲁棒性的模糊自适应算法.通过双足机器人运动学的仿真分析,验证了该算法的有效性.而且整套算法的计算时间约为0.35ms,可以用于实际双足机器人的实时控制.
An improved inverse kinematics control scheme for biped robot is presented to solve the problems of the Jacobian matrix singularity and fixed-parameters in the numerical solution.The Jacobian inversion is replaced with the approximate solution of differential motion equation.Combined with the adaptive fuzzy control to reduce the tracking error,it can regulate adaptive parameters in order to arbitrarily approach the exact solution by approximate solution.An extremely accurate and strong robust fuzzy adaptive algorithm can be obtained.Results of simulation on a biped robot show the effectiveness of the scheme.The total computation time of the presented algorithm is about 0.35 ms,so it can satisfy the requirements of real-time control of the actual biped robot.
出处
《机器人》
EI
CSCD
北大核心
2010年第4期534-539,546,共7页
Robot
基金
国家自然科学基金资助项目(60274022
60574076)
教育部博士点基金资助项目(20070611018)
