Human stochastic closed-loop behavior for master-slave teleoperation using multi-leap-motion sensor

• 来源期刊: 中国科学:技术科学（英文版）
• 年，卷（期）: 2017,（3）
• 页码范围: 374-384
• 页数: 11页
• 语种: 英文
• DOI: 10.1007/s11431-016-0434-x

摘要

Teleoperation has a wide range of applications that have been under development over the past two decades.Previous researches have focused on the control design of teleoperation machine systems to deal with obstacles such as time-delayed stability and transparency.Recent researches have shown that the inclusion of human closed-loop dynamics in control design can improve the performance of robot telemanipulation.The complexity of human behavior arises from the uncertainty of both human physiology and psychology;hence,the investigation can benefit from empirical studies.This study develops a type of statistical learning method to model and evaluate human stochastic closed-loop behavior,which is considered as a hand motion during the direct incremental control process of master-slave teleoperation.The hand trajectory is empirically considered as having a binary linear regression relationship with the error and error rate between the demanded and simulated teleoperator trajectories,while random movements with zero error and error rate are discovered.Hand movement tracking is achieved using a multi-leap-motion sensor (MLM),which is a markerless and natural infrared vision-based manner for motion capture.The established behavior model and statistical learning results reveal certain human properties of operational activities including visual perception,decision making,and robot telemanipulation.The properties indicate some probable system enhancements for future work.