基于改进遗传算法的FAST促动器油缸装配机器人运动学

促动器作为500 m口径球面射电望远镜(five-hundred-meter aperture spherical radio telescope,FAST)主动反射面变形的驱动装置,共计2225台,是世界罕见的野外台站大规模应用的设备群.针对促动器油缸组件的实际装配需求,采用D-H法建立油缸装配机器人的关节空间;采用解析法求出装配机器人逆运动学的部分解析解;采用基于自适应多种群的遗传算法通过建立多目标函数得到了其余数值解.结果证明,给出的逆运动学模型正确;相比单种群、定算子概率、定迭代次数的传统遗传算法,给出的改进自适应多种群遗传算法稳定性好、收敛快、求解精度高.研究结果为FAST液压促动器油缸装配机器人末端执行器的运动轨迹规划打下坚实基础,也为相似机器人的运动学研究提供思路.

Solution of Kinematics for Oil Cylinder Assembly Robot of FAST Actuators Based on Improved Genetic Algorithm

As the driving device of the active reflector deformation of the 500 m aperture spherical radio telescope (FAST), a total of 222 5 actuators are used in large-scale field stations. According to the actual assembly requirements of the actuator cylinder, the joint space of the cylinder assembly robot is established by D-H method. The partial analytical solution of the inverse kinematics of the assembly robot is obtained by using the analytic method. The remaining numerical solutions are obtained by establishing multi-objective function based on genetic algorithm of adaptive multi population. The results show that the inverse kinematics model is correct; compared with the traditional genetic algorithm with single population, fixed operator probability and fixed iteration times, the improved adaptive multi population genetic algorithm has better stability, faster convergence and higher precision. This conclusion lays a solid foundation for the motion planning of the end effector of the FAST hydraulic actuator cylinder assembly robot, and also provides ideas for the kinematics research of similar robots.