Abstract: The inverse kinematics problem of redundant manipulator can be conversed as equivalent minimization problem based on the positive kinematics equations, and the minimization problem can be solved using an self-adaptive particle swarm optimization algorithm（SAEPSO）. In order to keep the vitality of the particle swarm, the ejector operation is used in the algorithm. When the particle satisfies the self-adaptive discrimination function introduced in the algorithm, it will be ejected in probability to the faraway area from the current position. To cooperate with the ejector operation, a new method of judging the merits of particles is proposed, which makes the particle can be ejected out of the feasible region. It is revealed in numerical experiments that the algorithm possesses relatively strong global search ability and fast search speed, and is thus applicable to solve the inverse kinematics problem of redundant manipulator.

Key words: redundant manipulator, inverse kinematics, particle swarm optimization algorithm, self-adaptive discrimination function, ejector

摘要： 以正向运动学方程为基础，冗余机械臂逆运动学解问题转换为等效最小值问题，提出一种自适应粒子群算法求解该问题。为了保持粒子群的活力，在算法内引入弹射操作。如果粒子满足设定自适应判别函数，粒子将按概率被从当前位置发射到较远区域。为了配合弹射操作，提出一种新的粒子优劣的判断机制，使得粒子可以被弹射飞出可行域。数值实验表明，算法具有较强的全局搜索能力和较快的搜索速度，是求解冗余机械臂逆运动学解的一种有效方法。

关键词: 冗余机械臂, 逆运动学, 粒子群算法, 自适应判别函数, 弹射