[1]YU Haoyang,LI Xingsen,YAN Yiqing.Extension gating network-driven intelligent control method for lower-limb exoskeletons[J].CAAI transactions on intelligent systems,2026,21(5):1323-1334.[doi:10.11992/tis.202510022]
Copy

Extension gating network-driven intelligent control method for lower-limb exoskeletons

CAAI transactions on intelligent systems[ISSN 1673-4785/CN 23-1538/TP] Volume: 21 Number of periods: 2026 5 Page number: 1323-1334 Column: 学术论文—智能系统 Public date: 2026-09-05
Title:
Extension gating network-driven intelligent control method for lower-limb exoskeletons
Author(s):
YU Haoyang1; 2;  LI Xingsen1; 2;  YAN Yiqing1; 2
1. Institute of Extenics and Innovation Methods, Guangdong University of Technology, Guangzhou 510006, China;
2. School of Electromechanical Engineering, Guangdong University of Technology, Guangzhou 510006, China
Keywords:
extension neural network; temporal convolutional network; mixture-of-experts; extension intelligence; extension distance; lower-limb exoskeleton; joint moment estimation; intelligent control system
CLC:
TP18
DOI:
10.11992/tis.202510022
Abstract:
Lower-limb exoskeleton technology holds substantial promise for applications in rehabilitation medicine and daily mobility assistance. However, achieving adaptive control across complex terrains remains a critical challenge. To address the limited transferability of control strategies and inadequate dynamic adaptability in current systems, this study introduces a mixture-of-experts model that integrates extension gating network with temporal convolutional experts. The model utilizes an extension distance function to dynamically quantify the compatibility between terrain features and expert modules, thereby enabling intelligent routing and adaptive decision-making based on multi-source gait information. By leveraging the temporal modeling capabilities of temporal convolutional experts, the system generates precise assistive torques aligned with human motion patterns. In experiments involving multiple terrains including flat ground and slopes, the proposed model reduced the average hip joint torque prediction error by 18.71% compared to conventional LSTM, with an average inference latency of merely 37.09 ms. In scene transition experiments, the model demonstrates a 17.75% reduction in average error compared to LSTM. This model effectively enhances the cross-terrain adaptability of exoskeleton systems, and its extension gating network routing mechanism and information fusion architecture provide valuable insights for multimodal robotic decision-making and dynamic neural network routing optimization.
References:
[1] KAPSALYAMOV A, JAMWAL P K, HUSSAIN S, et al. State of the art lower limb robotic exoskeletons for elderly assistance[J]. IEEE access, 2019, 7: 95075-95086
[2] 胡湛, 彭希哲. 应对中国人口老龄化的治理选择[J]. 中国社会科学, 2018(12): 134-155, 202 HU Zhan, PENG Xizhe. Governance choices in dealing with China’s aging population[J]. Social sciences in China, 2018(12): 134-155, 202
[3] KIM M S, SHIN I, PARK D G, et al. Post-stroke movement disorders disappearance: a report of disappearance of tardive dyskinesia after stroke and a literature review[J]. Acta neurologica belgica, 2023, 123(5): 2005-2007
[4] 郭毅锋, 邓凯, 黄丽敏, 等. 主动助力下肢康复外骨骼机器人研究现状[J]. 机械传动, 2025, 49(5): 161-176 GUO Yifeng, DENG Kai, HUANG Limin, et al. Current research status of active-assist lower limb rehabilitation exoskeleton robots[J]. Journal of mechanical transmission, 2025, 49(5): 161-176
[5] 杨洁, 王旭鹏, 唐欣尧, 等. 无动力髋关节助力外骨骼设计与分析[J]. 机械传动, 2024, 48(8): 70-81 YANG Jie, WANG Xupeng, TANG Xinyao, et al. Design and analysis of non-dynamic hip joint assisted exoskeletons[J]. Journal of mechanical transmission, 2024, 48(8): 70-81
[6] JEZERNIK S, COLOMBO G, KELLER T, et al. Robotic orthosis lokomat: a rehabilitation and research tool[J]. Neuromodulation: technology at the neural interface, 2003, 6(2): 108-115
[7] FRANKS P W, BRYAN G M, REYES R, et al. The effects of incline level on optimized lower-limb exoskeleton assistance: a case series[J]. IEEE transactions on neural systems and rehabilitation engineering, 2022, 30: 2494-2505
[8] FRANKS P W, BRYAN G M, MARTIN R M, et al. Comparing optimized exoskeleton assistance of the hip, knee, and ankle in single and multi-joint configurations[J]. Wearable technologies, 2021, 2: e16
[9] ZHANG Juanjuan, FIERS P, WITTE K A, et al. Human-in-the-loop optimization of exoskeleton assistance during walking[J]. Science, 2017, 356(6344): 1280-1284
[10] 牛苗赫, 雷飞. 面向下肢外骨骼的运动意图识别算法研究[J]. 智能系统学报, 2025, 20(2): 407-415 NIU Miaohe, LEI Fei. Motion intention recognition algorithms for lower limb exoskeleton[J]. CAAI transactions on intelligent systems, 2025, 20(2): 407-415
[11] 孙国安, 赵明, 张廷丰, 等. 基于肌电预测模型的下肢外骨骼自适应控制方法[J]. 郑州大学学报(工学版), 2025, 46(3): 34-41 SUN Guoan, ZHAO Ming, ZHANG Tingfeng, et al. Adaptive control method of lower limb exoskeleton based on myoelectric prediction model[J]. Journal of Zhengzhou University (engineering science), 2025, 46(3): 34-41
[12] 衣淳植, 贾翊丞, 姜峰, 等. 基于惯性测量单元的人体运动意图识别方法: 现状与挑战[J]. 智能系统学报, 2025, 20(4): 763-775 YI Chunzhi, JIA Yicheng, JIANG Feng, et al. Human motion intention recognition method based on inertial measurement unit: current situation, and challenges[J]. CAAI transactions on intelligent systems, 2025, 20(4): 763-775
[13] 颜兵兵, 宋佳宝, 单琳娜, 等. 基于足压与姿态信息融合的步态相位识别方法[J]. 兵器装备工程学报, 2025, 46(5): 177-184 YAN Bingbing, SONG Jiabao, SHAN Linna, et al. Gait phase recognition method based on fusion of foot pressure and posture information[J]. Journal of ordnance equipment engineering, 2025, 46(5): 177-184
[14] 刘洋, 彭世国, 马宏志, 等. 下肢外骨骼机器人动力学参数辨识与步态跟踪[J]. 广东工业大学学报, 2022, 39(6): 44-52 LIU Yang, PENG Shiguo, MA Hongzhi, et al. Dynamic parameter identification and gait tracking of lower limb exoskeleton robot[J]. Journal of Guangdong University of Technology, 2022, 39(6): 44-52
[15] 崔占贺, 艾莉莎, 马欣雨, 等. 基于1D-CNN-SVM的下肢外骨骼步态信息识别研究[J]. 电子测量技术, 2025, 48(12): 71-78 CUI Zhanhe, AI Lisha, MA Xinyu, et al. Gait information recognition study of lower limb exoskeleton based on 1D-CNN-SVM[J]. Electronic measurement technology, 2025, 48(12): 71-78
[16] 李鑫, 徐赵洁, 柳金康, 等. 基于双源传感与LSTM的运动步态相位识别研究[J]. 传感器与微系统, 2025, 44(4): 13-16, 20 LI Xin, XU Zhaojie, LIU Jinkang, et al. Research on phase recognition of motion gait based on dual source sensing and LSTM[J]. Transducer and microsystem technologies, 2025, 44(4): 13-16, 20
[17] 汤璐, 杨玺霖, 王祥瑞, 等. 基于注意力机制的CNN-LSTM网络下肢膝关节角度预测[J]. 数据采集与处理, 2024, 39(4): 996-1008 TANG Lu, YANG Xilin, WANG Xiangrui, et al. An attention mechanism-based CNN-LSTM framework for lower limb knee joint angle prediction[J]. Journal of data acquisition & processing, 2024, 39(4): 996-1008
[18] REN Jiale, WANG Aihui, LI Hengyi, et al. A transformer-based neural network for gait prediction in lower limb exoskeleton robots using plantar force[J]. Sensors, 2023, 23(14): 6547
[19] ZHANG Yuanwen, XIONG Jingfeng, XIAN Haolan, et al. Joint moment estimation for hip exoskeleton control: a generalized moment feature generation method[J]. Biomimetic intelligence and robotics, 2025, 5(4): 100246
[20] MOLINARO D D, SCHERPEREEL K L, SCHONHAUT E B, et al. Task-agnostic exoskeleton control via biological joint moment estimation[J]. Nature, 2024, 635(8038): 337-344
[21] MOLINARO D D, KANG I, CAMARGO J, et al. Subject-independent, biological hip moment estimation during multimodal overground ambulation using deep learning[J]. IEEE transactions on medical robotics and bionics, 2022, 4(1): 219-229
[22] DARYAKENARI F H, FARIZEH T. A novel Transformer-based method for full lower-limb joint angles and moments prediction in gait using sEMG and IMU data[EB/OL]. (2025-06-05)[2025-10-22]. https://arxiv.org/abs/2506.04577.
[23] ZHU Yeqing, XIA Di, ZHANG Heng. Using wearable sensors to estimate vertical ground reaction force based on a transformer[J]. Applied sciences, 2023, 13(4): 2136
[24] BAUD R, MANZOORI A R, IJSPEERT A, et al. Review of control strategies for lower-limb exoskeletons to assist gait[J]. Journal of NeuroEngineering and rehabilitation, 2021, 18(1): 119
[25] 杨春燕, 蔡文, 汤龙. 可拓学[M]. 2版. 北京: 科学出版社, 2024.
[26] 杨春燕, 李兴森. 可拓学40年发展历程及研究进展[J]. 广东工业大学学报, 2023, 40(6): 1-11 YANG Chunyan, LI Xingsen. Development and research progress of extenics over the past 40 years[J]. Journal of Guangdong University of Technology, 2023, 40(6): 1-11
[27] 赵燕伟, 朱芬, 桂方志, 等. 基于可拓距的改进k-means聚类算法[J]. 智能系统学报, 2020, 15(2): 344-351 ZHAO Yanwei, ZHU Fen, GUI Fangzhi, et al. Improved k-means algorithm based on extension distance[J]. CAAI transactions on intelligent systems, 2020, 15(2): 344-351
[28] 陈晓华, 刘大莲, 田英杰, 等. 可拓支持向量分类机[J]. 智能系统学报, 2018, 13(1): 147-151 CHEN Xiaohua, LIU Dalian, TIAN Yingjie, et al. Extension support vector classification machine[J]. CAAI transactions on intelligent systems, 2018, 13(1): 147-151
[29] 张丽芳, 李兴森. 基元潜部特征元挖掘的智能方法研究[J]. 智能系统学报, 2025, 20(2): 457-464 ZHANG Lifang, LI Xingsen. Research on intelligent methods for latent features mining of basic element[J]. CAAI transactions on intelligent systems, 2025, 20(2): 457-464
[30] 田英杰, 刘大莲, 李兴森. 面向标签可变性的可拓分类方法[J]. 广东工业大学学报, 2025, 42(4): 1-7 TIAN Yingjie, LIU Dalian, LI Xingsen. Extension classification method for label variability[J]. Journal of Guangdong University of Technology, 2025, 42(4): 1-7
[31] 周玉, 钱旭, 张俊彩, 等. 可拓神经网络研究综述[J]. 计算机应用研究, 2010, 27(1): 1-5 ZHOU Yu, QIAN Xu, ZHANG Juncai, et al. Survey and research of extension neural network[J]. Application research of computers, 2010, 27(1): 1-5
[32] 王体春, 夏天, 费叶琦. 融合卷积深度置信网络与可拓神经网络的齿轮故障诊断方法[J]. 计算机集成制造系统, 2025, 31(6): 2178-2193 WANG Tichun, XIA Tian, FEI Yeqi. Gear fault diagnosis method integrating convolutional deep belief network and extensible neural network[J]. Computer integrated manufacturing systems, 2025, 31(6): 2178-2193
[33] 陈鑫. 可拓神经网络及其在汽车智能驾驶路径规划中的应用[D]. 广州: 广东工业大学, 2024. CHEN Xin. Extension neural network and its applications in path planning method for smart driving cars[D]. Guangzhou: Guangdong University of Technology, 2024.
[34] 夏崇坤, 苏成利, 曹江涛, 等. 基于MDP-ISAP的ENN分类算法及其应用[J]. 控制工程, 2017, 24(9): 1958-1964 XIA Chongkun, SU Chengli, CAO Jiangtao, et al. MDP-ISAP with ENN-based classification algorithm and its application[J]. Control engineering of China, 2017, 24(9): 1958-1964
[35] JACOBS R A, JORDAN M I, NOWLAN S J, et al. Adaptive mixtures of local experts[J]. Neural computation, 1991, 3(1): 79-87
[36] MOLINARO D D, KANG I, YOUNG A J. Estimating human joint moments unifies exoskeleton control, reducing user effort[J]. Science robotics, 2024, 9(88): eadi8852
[37] CAMARGO J, RAMANATHAN A, FLANAGAN W, et al. A comprehensive, open-source dataset of lower limb biomechanics in multiple conditions of stairs, ramps, and level-ground ambulation and transitions[J]. Journal of biomechanics, 2021, 119: 110320
Similar References:

Memo

-

Last Update: 2026-09-05

Copyright © CAAI Transactions on Intelligent Systems