摘要
针对欠自由度机器人在工作空间内运动学逆解的存在性问题 ,提出虚拟关节法 ,并利用神经网络的学习能力 ,区分出可解子空间和不可解子空间 。
For solving the solvability problem within the work space of underfreedom robot, a virtual joint method is proposed and BP neural network is used to distinguish the solvable and unsolvable posture space and to determine whether there exists an inverse kinematics solution. The result is proved ideal.
出处
《应用科学学报》
CAS
CSCD
2001年第1期44-47,共4页
Journal of Applied Sciences
