• CN: 11-2187/TH
  • ISSN: 0577-6686

Journal of Mechanical Engineering ›› 2023, Vol. 59 ›› Issue (5): 29-40.doi: 10.3901/JME.2023.05.029

Previous Articles     Next Articles

Research on Multimodal Interaction System of Lower Limb Exoskeleton Oriented to Competitive Training

QI Wenqian1,2, SUN Shouqian1,2, CHEN Chao1,2, ZHANG Yebingqing1,2, ZHAO Dongwei1,2   

  1. 1. Modern Industrial Design Institute, Zhejiang University, Hangzhou 310027;
    2. Zhejiang Key Laboratory of Design Intelligence and Digital Creativity Research, Zhejiang University, Hangzhou 310000
  • Received:2022-03-01 Revised:2022-11-29 Online:2023-03-05 Published:2023-04-20

Abstract: Currently, there is a lack of research on wearable devices that focus on assisting lower limbs to improve physical fitness training. The synergistic research of exoskeleton and virtual reality is an emerging direction in recent years. The exoskeleton system with multimodal information fusion can optimize the training effect and experience. Therefore, a multimodal interaction system of the lower limb exoskeleton for athletic training is developed. Which not only provides a wide variety of standardized training but also realizes a multisensory immersion experience. Firstly, a multi-degree-of-freedom step-by-step action planning method is designed to reproduce training actions based on lower limb movement action points. Second, a multimodal interaction system for the lower limb exoskeleton is proposed to realize the presentation and indication of normative movements in information space through a virtual reality athletic training simulator. The action data set and the lower limb exoskeleton robot are constructed to realize the assistance and correction of basic training in physical space. The system is based on a multimodal interaction execution strategy to give trainers visual, auditory, and tactile multisensory feedback in real-time to achieve an immersive experience that meets safety norms. Finally, the system is evaluated through two experiments on system functionality and user experience. The functional experiment proves that the system can improve the action accuracy rate by 27.62% on average compared with the original traditional training, and has certain generality. The experimental results of user experience show that the system’s function meets the design expectations, and the comfort compared to other indicators needs to be improved.

Key words: lower limb exoskeleton robot, virtual reality, action data set, multimodal interaction

CLC Number: