基于BP神经网络算法的3-CRS/SP并联机构正解分析  

Forward Kinematics Solution Analysis of 3-CRS/SP Parallel Mechanism based on BP Neural Network Algorithm

作　　者：赵耀虹[1] 夏昊 王自谦 李瑞琴[1] 夭银银 王勇军 Zhao Yaohong;Xia Hao;Wang Ziqian;Li Ruiqin;Yao Yinyin;Wang Yongjun(School of Mechanical Engineering,North University of China,Taiyuan 030051,China)

机构地区：[1]中北大学机械工程学院,山西太原030051

出　　处：《机械传动》2018年第11期58-61,共4页Journal of Mechanical Transmission

基　　金：国家自然科学基金(51275486)

摘　　要：提出一种通过BP神经网络求解并联机构位置正解的方法。根据BP神经网络的算法,采用了3层前向神经网络。对3-CRS/SP并联机构建立旋转矩阵求得该机构的位置逆解。将位置逆解所得数据作为神经网络的输入,将求解位置逆解的动平台位姿包含4个参数值作为输出,进行神经网络的训练。仿真和实验结果表明,该方法对于求解并联机构位置正解快速有效;相对于解析法,该方法直观简洁,能够为并联机构的控制过程提供数值参考。A method for solving the forward kinematics analysis of parallel mechanism based on BP neural network is presented.According to the algorithm of BP neural network,three layers forward neural network is adopted.The inverse kinematics solution of the mechanism can be obtained by establishing the rotation matrix of 3-CRS/SP parallel mechanism.It can be used as input of neural network and the dynamic platform position including four parameters can be regard as the output,then neural network can be trained.The simulation and experimental results show that the method is fast and effective for solving the parallel mechanism forward kinematics,and it is intuitive and simple compared with the analytical method.It can provide numerical reference for the control process of parallel mechanism.

关 键 词：BP神经网络 并联机构 位置正解 

分 类 号：TH112[机械工程—机械设计及理论]