安全巡邏自動車之路徑規劃與遠端呈現

Abstract

由於網路與微處理器的快速進步，促使遠端操控的性能大幅提高，遠端操控機器人也得以廣泛地應用在各種領域中，包括安全巡邏任務。為了更有效率的操控安全巡邏自動車，根據近代研究中指出將路徑規劃技術引進至遠端搖控自走車可更有效率的導航，本論文實現一結合路徑規劃技術之智慧型人機介面系統，同時也設計一虛擬實境遠端呈現系統，藉由智慧型人機介面系統及遠端呈現系統，使用者可輕鬆的操控自走車到達目的地，並可經由多視角的輔助工具，來觀看虛擬場景，在此系統中也運用一力回饋策略，幫助使用者在操控自走車時避免和障礙物發生碰撞，而使用者可以透過此呈現系統以及力回饋搖桿，同時擁有視覺和觸覺的回饋資訊。

Rapid development in networks and microprocessors much enhances the teleoperation systems, which have been applied to different fields, such as security patrolling mission. Recent research results show that path planning techniques can be incorporated into the teleoperation of mobile robot for efficient navigation. In this thesis, we implement a path planning algorithm in the proposed intelligent human-machine interface for mobile robot navigation. In addition, we design a VR-Based telepresence system. Through this intelligent human-machine interface and telepresence system, the user can not only guide the mobile robot to destination easily, but also view the VR scene via the supporting tool of multi-view. We also develop a force-reflecting strategy to assist the user to manipulate the robot for obstacle avoidance. Consequently, the user may receive visual and haptic information simultaneously through the VR system and the force reflection joystick.