結合虛擬實境與機器人之遠端操控系統

Abstract

遠端操控系統可應用於人類無法到達或是具有危險性的環境中，將虛擬實境技術引進於遠端操作中，則可方便操作者從不同視角來觀察操作環境。在這篇論文中，我們完成一結合虛擬實境與機器人的遠端操控系統，並提供一個智慧型介面來幫助操作者能更迅速且正確地完成任務，除了3D場景的視覺回饋與力回饋之外，此系統更利用勢能力場(Potential Field Approach)來規劃機器人的路徑，藉此能力提供操作者避障、引導到目的地、路徑追蹤等功能。我們也利用實驗來評估系統性能，證明不論機器人是處於自由空間或是正與環境接觸的狀態，操作者在此智慧型介面的輔助下，對完成任務所需的時間及接觸力這二方面都能有較佳的表現。

Nowadays, teleoperation has been used in the environment that is hazardous or inaccessible for human. The introduction of virtual reality technique to teleoperation provides the operators different viewing angles to look at the environment. In this thesis, we implement a VR-based telerobotic system, and develop an intelligent interface to help the operator for manipulation. Besides providing both 3D environmental visual and force feedbacks, the VR-based telerobotic system utilizes the potential field approach to plan robot path. Thus, this system can provide the operator the functions of obstacle avoidance, guidance toward the goal, and path tracking. We evaluated the its performance through several experiments to see whether better performance on task completion time and contact force can be achieved, when the operator manipulates the robot in the free space or contacting with the environment.