外骨骼机器人高实时高能效控制器研究

doi:10.3901/JME.2025.15.209

摘要/Abstract

摘要： 外骨骼机器人在多个实验室环境中取得了成功，但实际部署中仍面临着一系列与控制器硬件相关的挑战，包括控制器体积过大，实时性不足，计算资源瓶颈等问题。为了解决这些问题，开发了一套用于外骨骼机器人的具有可扩展性的硬件电路系统，同时兼顾了实时性和高算力特性。通过对控制器硬件信号流向进行建模，优化通信线路和流量分配，降低了系统通信延迟。根据外骨骼机器人在空间和时间上对算力需求的差异，设计了基于步态的实时任务调度算法，优化了算力分配。通过并行流水和模块折叠复用的方式来设计运算加速器，有效减少了硬件资源的消耗，提高了系统算力。开展了控制器实时性测试、算力测试和助力外骨骼机器人负重行走试验。经测试，所设计的控制器主从通信延迟小于20μs，算力功耗比15 GOPS/W，相比于主流控制器架构具有低延迟高算力特性。在平地，上下坡负重行走体现出良好的跟踪效果，为外骨骼机器人更广泛的应用打下了硬件基础。

关键词: 外骨骼机器人, 控制器硬件, 实时性, 任务调度, 算力优化

Abstract: The exoskeleton robot has achieved success in multiple laboratory environments, but there are still a series of challenges related to the controller hardware in real-world deployment, including large controller size, insufficient real-time performance, and computational resource bottlenecks. To address these issues, a scalable hardware circuit system for the exoskeleton robot, balancing both real-time performance and high computational power characteristics is developel. By modeling the signal flow of the controller hardware, optimizing communication lines and traffic distribution, the system’s communication delay is reduced. Based on the varying computational power demands of the exoskeleton robot in space and time, a gait-based real-time task scheduling algorithm is designed to optimize the allocation of computational resources. The computational accelerator is designed using parallel pipelining and module folding to effectively reduce hardware resource consumption and improve system performance. Real-time performance tests, computational power tests, and load-walking experiments with the exoskeleton robot are conducted. The tests showed that the designed controller’s master-slave communication delay is less than 20 μs, and the computational power efficiency is 15 GOPS/W, demonstrating low latency and high computational power compared to mainstream controller architectures. The exoskeleton robot demonstrated excellent tracking performance during weighted walking on flat ground and uphill, laying the hardware foundation for its broader application.

Key words: exoskeleton robots, controller hardware, real-time performance, task scheduling, computational optimization

中图分类号:

TG242

王泽正, 包颖炜, 李慧莱, 李驰, 孙茂文, 欧阳小平. 外骨骼机器人高实时高能效控制器研究[J]. 机械工程学报, 2025, 61(15): 209-220.

WANG Zezheng, BAO Yingwei, LI Huilai, LI Chi, SUN Maowen, OUYANG Xiaoping. Real-time and High-efficiency Hardware Circuits Design for Assistive Exoskeleton Robots[J]. Journal of Mechanical Engineering, 2025, 61(15): 209-220.

参考文献

[1] 黄思翰,王柏村,张美迪,等.面向人本智造的新一代操作工:参考架构、使能技术与典型场景[J].机械工程学报,2022,58(18):251-264.HUANG Sihan,WANG Baicun,ZHANG Meidi,et al.Operator 4.0 towards human-centric smart manufacturing:Framework,enabling technologies and typical scenarios[J]. Journal of Mechanical Engineering,2022,58(18):251-264.
[2] KAGIROV I,KAPUSTIN A,KIPYATKOVA I,et al.Medical exoskeleton "Remotion" with an intelligent control system:Modeling,implementation,and testing[J].Simulation Modelling Practice and Theory,2021,107:102200.
[3] ZHANG X,CHEN X,HUO B,et al. An integrated evaluation approach of wearable lower limb exoskeletons for human performance augmentation[J]. Scientific Reports,2023,13(1):4251.
[4] YAO Y M, SHAO D Q, TARABINI M, et al.Advancements in sensor technologies and control strategies for lower-limb rehabilitation exoskeletons:A comprehensive review[J]. Micromachines,2024,15(4):489.
[5] MEHR J K,SHARIFI M,MUSHAHWAR V K,et al.Intelligent locomotion planning with enhanced postural stability for lower-limb exoskeletons[J]. IEEE Robotics and Automation Letters,2021,6(4):7588-7595.
[6] LIAO H P,CHAN H H,LIU G Y,et al. Design,control and validation of a novel cabledriven series elastic actuation system for a flexible and portable back-support exoskeleton[J]. IEEE Transactions on Robotics,2024,40:2769-2790.
[7] XING X Y,ZHANG S N,HUANG T H,et al. Spatial iterative learning torque control of robotic exoskeletons for high accuracy and rapid convergence assistance[J].IEEE/ASME Transactions on Mechatronics,2024,29(6):4215-4227.
[8] WANG C,GUO Z M,DUAN S C,et al. A real-time stability control method through sEMG interface for lower extremity rehabilitation exoskeletons[J]. Frontiers in Neuroscience,2021,15:645374.
[9] WANG J Q,WU D M,GAO Y Z,et al. Integral real-time locomotion mode recognition based on GA-CNN for lower limb exoskeleton[J]. Journal of Bionic Engineering,2022,19(5):1359-1373.
[10] ZHEN T, YAN L. Real-time control strategy of exoskeleton locomotion trajectory based on multi-modal fusion[J]. Journal of Bionic Engineering,2023,20(6):2670-2682.
[11] SALAMUN K,PAVI I,DAPO H,et al. Weakly hard real-time model for control systems:A survey[J].Sensors (Basel,Switzerland),2023,23(10):4652.
[12] HE L, FAN W, DAI Z, et al. A human-machineenvironment interactive measurement system for non-anthropomorphic exoskeletons[J]. IEEE Transactions on Instrumentation and Measurement,2023,73:1-11.
[13] WU X Y,LI J K,LIU L,et al. The visual footsteps planning system for exoskeleton robots under complex terrain[J]. IEEE Transactions on Systems, Man, and Cybernetics:Systems,2023,53(8):5149-5160.
[14] WANG J,LIU J H,ZHANG G W,et al. Periodic event-triggered sliding mode control for lower limb exoskeleton based on human-robot cooperation[J]. ISA Transactions,2022,123:87-97.
[15] KAWAMOTO H,LEE S,KANBE S,et al. Power assist method for HAL-3 using EMG-based feedback controller[C]//SMC'03 Conference Proceedings. 2003IEEE International Conference on Systems,Man and Cybernetics. Conference Theme-System Security and Assurance (Cat. No. 03CH37483). IEEE,2003,2:1648-1653.
[16] SU B B, SMITH C, FAREWIK E G. Gait phase recognition using deep convolutional neural network with inertial measurement units[J]. Biosensors,2020,10(9):109.
[17] PARK K W,CHOI J,KONG K. Data-driven modeling for gait phase recognition in a wearable exoskeleton using estimated forces[J]. IEEE Transactions on Robotics,2023,39(4):3072-3086.
[18] WU X Y,YUAN Y,ZHANG X K,et al. Gait phase classification for a lower limb exoskeleton system based on a graph convolutional network model[J]. IEEE Transactions on Industrial Electronics,2021,69(5):4999-5008.
[19] WEI H C,TONG R K,WANG M Y,et al. Gait phase detection based on LSTM-CRF for stair ambulation[J]. IEEE Robotics and Automation Letters,2023,8(9):6029-6035.
[20] KANG I,MOLINARO D D,DUGGAL S,et al. Real-time gait phase estimation for robotic hip exoskeleton control during multimodal locomotion[J]. IEEE Robotics and Automation Letters,2021,6(2):3491-3497.
[21] ZHANG Z W,CAI X F,ZHANG M B,et al. Real-time gait intention recognition for active control of unilateral knee exoskeleton[J]. Applied Bionics and Biomechanics,2024,2024(1):9426782.
[22] CAO W J,MA Y,CHEN C J,et al. Hardware circuits design and performance evaluation of a soft lower limb exoskeleton[J]. IEEE Transactions on Biomedical Circuits and Systems,2022,16(3):384-394.
[23] LI W,SHI P,LI S J,et al. Current status and clinical perspectives of extended reality for myoelectric prostheses[J]. Frontiers in Bioengineering and Biotechnology,2024,11:1334771.
[24] JUN K,AKIRA K,SHOGO A,et al. Adaptive task scheduling for an assembly task coworker robot based on incremental learning of human's motion patterns[J]. IEEE Robotics and Automation Letters,2017,2(2):856-863.
[25] LEE W K, JUNG S. FPGA design for controlling humanoid robot arms by exoskeleton motion capture system[C]//2006 IEEE International Conference on Robotics and Biomimetics. IEEE,2006:1378-1383.
[26] SUN M W,OUYANG X P,MATTILA J,et al. One novel hydraulic actuating system for the lower-body exoskeleton[J]. Chinese Journal of Mechanical Engineering,2021,34(1):31.
[27] SHETTY P R, MENEZES J A, SONG S. Ankle exoskeleton control via data-driven gait estimation for walking,running,and inclines[J]. IEEE Robotics and Automation Letters,2025,10(6):5855-5862.
[28] LUO R M,SUN S Q,ZHANG X F,et al. A low-cost end-to-end s EMG-based gait sub-phase recognition system[J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering,2019,28(1):267-276.
[29] LIU Jia,BJORKMAN A,ANTFOLK C,et al. The Impact of stimulation parameters on reaction times following transcutaneous electrical stimulation in the lower leg[J].IEEE Transactions on Haptics,2025:1-11.
[30] KIM J,LEE G,HEIMGARTNER R,et al. Reducing the metabolic rate of walking and running with a versatile,portable exosuit[J]. Science,2019,365(6454):668-672.
[31] HAUFE F L,DUROYON E G,WOLF P,et al. Outside testing of wearable robots for gait assistance shows a higher metabolic benefit than testing on treadmills[J].Scientific Reports,2021,11(1):14833.