|本期目录/Table of Contents|

[1]孙国瑞,吴震,汤腾飞.基于深度残差网络的超冗余度蛇形臂机器人实时逆运动学求解方法[J].浙江理工大学学报,2026,55-56(自科四):474-486.
　SUN Guorui,WUZhen,TANGTengfei.A real-time inverse kinematics solution for hyper-redundant snake-like robots based on deep residual network[J].Journal of Zhejiang Sci-Tech University,2026,55-56(自科四):474-486.
点击复制

基于深度残差网络的超冗余度蛇形臂机器人实时逆运动学求解方法()

浙江理工大学学报[ISSN:1673-3851/CN:33-1338/TS]

卷:
55-56
期数:
2026年自科第四期
页码:
474-486
栏目:
出版日期:
2026-07-10

文章信息/Info

Title:
A real-time inverse kinematics solution for hyper-redundant snake-like robots based on deep residual network
文章编号:
1673-3851(2026) 07-0499-11
作者:
孙国瑞 ; 吴震 ; 汤腾飞
1. 浙江理工大学机械工程学院 ,杭州 310018;2. 浙江交通职业技术学院海运学院 ,杭州 311112
Author(s):
SUN Guorui ; WUZhen ; TANGTengfei 
1. School of Mechanical Engineering, Zhejiang Sci-Tech University, Hangzhou 310018, China; 2. Maritime Department, Zhejiang Institute of Communications, Hangzhou 311112, China
关键词:
超冗余度蛇形臂机器人; 逆运动学 ;  实时性 ; 深度残差网络; 多层感知机
分类号:
TH112
文献标志码:
A
摘要:
针对超冗余度蛇形臂机器人在复杂受限环境作业中 ,逆运动学求解存在的多解性、强非线性及实时性不足等问题 ,提出一种基于深度残差网络(Deepresidualnetwork, ResNet)的逆运动学求解方法 ,设计了基于 ResNet的深度残差多层感知机(Deep residualmulti-layerperceptron, Res-MLP)模型 。采用“编码-深度残差主体-解码”三级架构 ,通过残差连接机制缓解深层网络梯度消失现象 ;在此基础上 ,设计融合正运动学重投影误差与最小范数正则项的复合损失函数 ,并引入 自监督训练闭环对网络输出进行物理约束 , 以保障解的合理性 。 以 12 自 由度超冗余度蛇形臂机器人为研究对象 ,开展单点逆运动学求解与深腔受限环境轨迹跟踪递进式仿真实验 ,并与经典雅可比伪逆法进行对比分析 。结果表明 :Res-MLP模型单点逆运动学求解平均耗时仅 1.87ms,为雅可比伪逆法的 1/3;在轨迹跟踪任务中求解总效率提升近 2倍 ,且生成构型更平滑紧凑 ,可有效适配狭窄环境作业 。该研究提出的方法在保障高精度定位的同时显著提升计算效率 ,为超冗余度蛇形臂机器人在航空发动机探伤、核电管道检测等场景的实时运动控制提供了高效求解方案。

参考文献/References:

[1] 吕炜枫, 周静, 冉文王, 等. 典型事故工况下压水堆核电厂内工作人员辐射风险分析方法研究[J]. 核科学与工程, 2023, 43(4):835‑841.
[2] 李鹏涛, 左洪福, 肖文, 等. 航空发动机叶片损伤及其修复技术研究与展望[J]. 航空学报, 2024, 45(15):140‑167.
[3] Mu Z G, Zhang L Y, Yan L, et al. Hyper-redundant manipulators for operations in confined space: typical applications, key technologies, and grand challenges[J]. IEEE Transactions on Aerospace and Electronic Systems, 2022, 58(6):4928‑4937.
[4] Ma X, Wang X C, Zhang Z H, et al. Design and experimental validation of a novel hybrid continuum robot with enhanced dexterity and manipulability in confined space[J]. IEEE/ASME Transactions on Mechatronics, 2023, 28(4):1826‑1835.
[5] Dimitropoulos N, Tzirtzilaki N, Michalos G, et al. Bio-inspired robots for confined space industrial applications: a novel cable-driven snake robot design[J]. Procedia CIRP, 2024, 125:107‑112.
[6] Chen Q H, Li M T, Wu H X, et al. Design, self-calibration and compliance control of modular cable-driven snake-like manipulators[J]. Mechanism and Machine Theory, 2024, 193:105562.
[7] Manara F, Poka A, Ludovico D, et al. Tendon-driven hyper‑redundant manipulators: a review on design solutions[J]. IEEE/ASME Transactions on Mechatronics, 2025, 30(5):3852‑3867.
[8] Chen Z, Zhang X B, Liu M L, et al. Review of cable-driven hyper-redundant robotic arms[C]//2024 3rd International Conference on Robotics, Artificial Intelligence and Intelligent Control(RAIIC). July 5-7, 2024, Mianyang, China. IEEE, 2024:110‑119.
[9] 闫庆梓. 超冗余机器人运动学标定及轨迹规划方法研究[D]. 大连: 大连理工大学, 2022:1‑21.
[10] Lakshmi Narayanan V, Narayan J, Gritli H, et al. A decade of inverse kinematics methods for serial manipulators: a systematic review[J]. Journal of Field Robotics, 2026, 43(1):184‑229.
[11] Colomé A, Torras C. Closed-loop inverse kinematics for redundant robots: comparative assessment and two enhancements[J]. IEEE/ASME Transactions on Mechatronics, 2015, 20(2):944‑955.
[12] Peng J Q, Xu W F, Liu T L, et al. End-effector pose and arm‑shape synchronous planning methods of a hyper-redundant manipulator for spacecraft repairing[J]. Mechanism and Machine Theory, 2021, 155:104062.
[13] Liu T L, Xu W F, Yang T W, et al. A hybrid active and passive cable-driven segmented redundant manipulator: design, kinematics, and planning[J]. IEEE/ASME Transactions on Mechatronics, 2021, 26(2):930‑942.
[14] 刘世平, 曹俊峰, 孙涛, 等. 基于BP神经网络的冗余机械臂逆运动学分析[J]. 中国机械工程, 2019, 30(24):2974‑2977,2985.
[15] Calzada-Garcia A, Victores J G, Naranjo-Campos F J, et al. Inverse kinematics for robotic manipulators via deep neural networks: experiments and results[J]. Applied Sciences, 2025, 15(13):7226.
[16] Ho C K, Chan L W, King C T, et al. A deep learning approach to navigating the joint solution space of redundant inverse kinematics and its applications to numerical IK computations[J]. IEEE Access, 2023, 11:2274‑2290.
[17] Malik A, Lischuk Y, Henderson T, et al. A deep reinforcement-learning approach for inverse kinematics solution of a high degree of freedom robotic manipulator[J]. Robotics, 2022, 11(2):44.
[18] Wang X Q, Liu X, Chen L R, et al. Deep-learning damped least squares method for inverse kinematics of redundant robots[J]. Measurement, 2021, 171:108821.
[19] 李艳辉, 霍琦, 李昂, 等. 模块化超冗余度空间机械臂的设计与实验[J]. 机器人, 2022, 44(1):55‑65.
[20] 朱晓俊, 王学谦, 马云萱, 等. 绳驱超冗余空间柔性机械臂遥操作系统设计与实验研究[J]. 机器人, 2022, 44(1):9‑18.

备注/Memo

备注/Memo:
基金项目 : 国家自然科学基金项目(52405040) ;浙江省自然科学基金项目(LQN25E050012,LQN26E050041)收稿日期 : 2026-01-13 网络出版日期 : 2026-03-30
作者简介 : 孙国瑞(1998— ) ,男 ,江苏连云港人 ,硕士研究生 ,主要从事机械电子工程方面的研究。通信作者: 汤腾飞,E-mail:tengfei@zstu. edu. cn
更新日期/Last Update: 2026-07-05