作　　者： ;

机构地区： 佛山科学技术学院

出　　处： 《机械与电子》 2020年第6期74-80,共7页

摘　　要： 为了解决传统的解析法用于机械臂逆运动学求解过程中存在操作烦琐和奇异点处无法逆运算等问题,提出了基于长短期记忆网络的机械臂逆运动学求解算法。通过对机械臂的正运动学和工作空间的分析,获取大量有效的机械臂关节空间数据和相对应的末端姿态数据,并作为训练样本。经过训练后,获得逆运动学解的高精度结果。对比传统解析法求解,基于长短期记忆网络方法的求解速度较快。此外,机器人仿真结果显示,相关轨迹无明显的抖动偏移。 In order to solve the problems of complex operation and inevitable singularity in the traditional analytical method of solving the inverse kinematics of the manipulator,an algorithm based on long short-term memory(LSTM)network was proposed to solve the inverse kinematics of the manipulator.A large amount of joint data of the manipulator and the corresponding attitude data were obtained as training data by the analysis of forward kinematics and working space of the manipulator.the result calculated by the trained LSTM network achieves high accuracy.Compared with the traditional analytical method,the proposed long and short-term memory network is faster.The simulation results show that there is no obvious displacement shift in the trajectory.

关 键 词： 机械臂逆运动学 运动学模型 长短期记忆网络 机器人仿真

领　　域： []