Safe Human Robot Inter action using Model Matching Control
Other Title
人機安全互動利用模型匹配控制
Journal
Journal of Chinese Society of Mechanical Engineers
Journal Volume
37
Journal Issue
6
Start Page
587
End Page
596
ISSN
0257-9731
Date Issued
2016-12
Author(s)
Abstract
Most of industrial robot arms in factory will reduce their speeds for guaranteeing human safety when human goes into the robots' working areas. But decreasing the robot speed will reduce production efficiency. To deal with this problem, this paper proposes an approach to satisfy safe human robot interaction while keeping the robot operation speed. The proposed model matching control (MMC) with danger index is used to not only change the robot stiffness but also avoid human injury to achieve safe human robot interaction (sHRI). The danger index (DI) based on the relative velocity and distance between human and the robot is developed to generate stiffness command so that human safety can be ensured even there is collision with the robot. Bond graph is adopted to model the active-passive variable stiffness elastic actuator (APVSEA) and design MMC controller. Experimental results show the proposed approach can achieve safe human robot interaction even under collision.
大部分工廠端的機械手臂，當有人員進入機械手臂的操作空間時，會降低手臂的速度以確保人員安全。但降低機械手臂的工作速度，也會影響產品的生產效率。為了解決這個問題，本論文提出一個方法，可以滿足人機安全互動同時又不用降低機械手臂的工作速度。本論文採用模型匹配控制來改變機械手臂的剛性，使得人機互動過程中，不會對人類產生傷害。我們利用Kinect偵測人與機械手臂的相對速度與位置，並且計算危險因子。危險因子用來產生剛性控制的命令，使得即使人類與機器人產生碰撞，也能夠確保人類安全。本論文使用鍵合圖來建構主動與被動變剛性彈性機構的模型，並且設計模型匹配控制器。本文的最後同時提供實驗結果。
Subjects
Model Matching Control
Safe Human Robot Interaction
Bond Graph
SDGs
Other Subjects
Accident prevention; Graph theory; Man machine systems; Multivariable systems; Robots; Stiffness; Bond graph; Elastic actuators; Industrial robot arms; Model matching control; Production efficiency; Relative velocity; Safe human-robot interaction; Variable stiffness; Human robot interaction
Type
journal article