基于多Agent的轨迹约束操作臂控制

随着机器人的日趋复杂，逆运动学算法受到了许多限制。90年代以来，基于多Agent的操作臂控制算法被提出，突破了传统的机器人控制依赖于精确模型的限制。但这些基于多Agent的算法中，操作臂末端点的运行轨迹是自由的。该文基于实际操作的需要，提出了基于多Agent的操作臂轨迹约束的控制策略。即依次使操作臂末端点到达轨迹上的各个采样点，达到操作臂末端点沿任一给定轨迹运行的目的，这将为避障操作打下必要的基础。该文提出拟合度的概念来控制末点轨迹的精度，仿真结果证明了此策略的可行性。

Manipulator Control with Trajectory-bounded Based on Multi-Agent

With the complexity of robots, the Inverse Kinematics Algorithm has fallen under many restricts. Since 1990s some control algorithms have been presented in manipulator based on multi agent. Although the algorithms broke the limitation of traditional robot control ones depended on the accurate model, the trajectory of manipulator end was free in the algorithms for a manipulator based on multi agent. On the basis of the need for real operation, the control strategy of planned trajectory based on multi agent was presented, the manipulator end got to each sampling point of the trajectory in turn, so that the manipulator end could run along the expected trajectory. It will provide the necessary foundation for the manipulator in operation of obstacle avoiding. The concept of fitness was presented in order to control the precision of the end and the feasibility of strategy was demonstrated by the simulation result.