摘要 本文的研究目的在發展移動機器人之自動充電系統,包括:充電站的設計、電池電壓的量測與自動入塢系統,並整合於具大樓巡邏與避障功能之移動機器人上,擴充其工作能力。當此移動機器人在執行大樓巡邏任務時,乃依據內建地圖與馬達編碼器的資訊進行定位,因此機器人可以知道自身與充電站間之距離。於巡邏過程中,若前進路徑上出現障礙物時,藉由聲納感測器,機器人可以偵測障礙物的寬度與二側牆壁間的距離,依此規劃避障路徑閃避障礙物,再繼續巡邏。當感測到電池電壓不足時,機器人必須暫停巡邏工作,返回充電站充電。本文在充電站設置一個特定的定位標誌使機器人能夠修正因為馬達驅動系統或其他因素造成的位置誤差,因此,當機器人返回充電站附近區域時,可由機器人的視覺系統尋找定位標誌,再藉由立體視覺演算法估算距離,使機器人能準確移動至充電站前,最後,機器人執行自動入塢程序與充電座連接,進行充電。根據已完成的實驗結果顯示,本文設計的自動充電系統能成功的控制機器人自行返回充電站並入塢以進行充電。
Abstract The purpose of this dissertation is to develop an automatic recharging system for a mobile robot, including the design of recharging station, battery voltage measurement, and automatic docking system. Here, the automatic recharging system is integrated into a mobile robot with capabilities of building patrol and obstacle avoidance to expand its working ability. Based on the information of built-in maps and motor encoders, the robot can know its location and distance between itself and the recharging station when it executes the task of building portal. If an obstacle appears on the forward path during patrol, the robot can detect the width of the obstacle and distances between the obstacle and two side walls by sonar sensors. Therefore, a forward path will be re-planned and the robot can avoid the obstacle to continue patrol. However, when a low battery voltage condition is sensed, the robot must make a decision to suspend the patrol and returns to the recharging station for recharging. In this study, a specific location sign is installed at the recharging station to compensate the position error caused by the motor driver system and other factors. In other words, when the robot arrives near the recharging station, it can find the location sign by its vision system. Then the robot can accurately move to the front of the recharging station by using the stereo vision algorithm to estimate the distance. Finally, the robot performs the procedure of automatic docking system to connect into recharging socket for recharging. To confirm the proposed approach, some experiments have been done. The experimental results show that the proposed automatic recharging system can successfully control the robot to move back and dock in the recharging station itself for recharging.