大部分传统的机器人控制方法依赖于一精确模型 .随着机器人的日趋复杂 ,此类方法已受到种种限制 .本文在借鉴国际最新研究成果的基础上 ,提出了一种新的操作臂控制算法 .算法继承了多 Agent的思想 ,易于实现并行运算 ;吸收了增量定向运动学思想的优点 ,使运算相对简单 .算法采用使操作臂末端点与目标点间距离单调减小的控制策略 ,并推导出了相应的非迭代公式 .针对基于多Agent的操作臂控制算法中存在的问题 ,本算法消除了初始阶段运动轨迹出现的振荡现象 ,从根本上避免了迭代算法带来的收敛速度问题 . Most of the traditional robot control methods rely on an accurate model.With the increasing complexity of robots, these methods have been limited.This paper based on the latest international research results, proposed a new operating arm control algorithm. The algorithm inherits the idea of ​​multi-agent and easy to implement parallel computing, and absorbs the advantages of incremental directional kinematics and makes the operation relatively simple.The algorithm uses a control strategy that monotonically decreases the distance between the end point of the operating arm and the target point A corresponding non-iterative formula is presented.According to the problems existing in the manipulator control algorithm based on multi-agent, this algorithm eliminates the oscillation phenomenon of the initial trajectory and basically avoids the convergence rate caused by the iterative algorithm.