CMAC在三关节机械臂逆运动学求解中的应用研究

The application research of CMAC in the inverse kinematics solution of three joint mechanical arms

传统的方法用于解决机械臂逆运动学问题不仅计算量大,而且还有不唯一解问题,很难用于机械臂的实时控制.系统地研究了CMAC神经网络在三关节机械臂逆运动学求解中的应用.简单介绍了CMAC网络,形象描述了三关节机械臂在二维平面的运动;接着,详细解释提出的三关节机械臂逆运动学问题解决方案,并且分析了基于2个CMAC网络的正、逆模型;最后,通过仿真得到验证:在求解逆运动学方面,CMAC网络学习算法比较简单,收敛速度快,且不存在局部极小问题.

The traditional method used for solving the problem of the inverse kinematics of manip- ulator not only needs a large amount of calculation, but also has problem of many answers to the inverse kinematics question, it is very difficult to be used for the real time control of mechanical arm. This paper systematically studies the application of CMAC neural network in the inverse ki- nematics solution of the three joint mechanical arms. This paper simply introduced the CMAC network, vividly described motion of the three joint mechanical arms on a two-dimensional surface, explained the inverse kinematics solution of the three joint mechanical arms in detailed, and then analyzed plus and inverse models based on the two CMAC networks finally, to be verified through the simulation： in solution of inverse kinematics learning algorithm of the CMAC network is relatively simple, has no local minimum problem, and its convergence rate is relatively fast.