本研究以人工蜂群演算法(ABC)結合快速搜索隨機樹(RRT)來創新路徑規劃演算法應用在移動式機器人當中，路徑規劃對於移動機器人是非常重要的研究之一，本文目的是如何在有障礙物的環境中，規劃出一條適合行走且無危險和有效率的路徑，讓機器人從起始點移動到目標點是安全且正確的。 與傳統的演算法有所不同，本文是先用RRT演算法的方式來尋找延伸點，在數個延伸點經過我們比較之後，選擇最佳的延伸點來使蜜蜂移動，因為人工蜂群演算法擁有結構簡單、容易操作且收斂速度較快的性能，他改善了以往用於路徑規劃的演算法在收斂速度慢和容易陷入區域最佳點的問題，雖然RRT在搜索未知區域方面有優良的特性，但它在每次規劃路徑上是不穩定的，所以本文結合人工蜂群的特性加入到RRT的演算法上，並處理障礙物的問題來模擬機器人的路徑。 總而言之，本文提出改良的演算法較以往單獨的RRT或單獨的人工蜂群演算法更具有效率與穩定且路徑最短。 ;In this study, we use the Artificial Bee Colony algorithm incorporating the fast search Random Tree (RRT) for innovation path planning, and is implemented in a mobile robot. Path planning for mobile robot is one of the very important researches. The aim is how to plan a path in the environment which have obstacles, and make the robot walk from start point to target safely and correctly. Different from conventional algorithms, we use the RRT algorithm to find several extending points, and after comparison we choose the best extending point to make the extension of bees move. Because Artificial Bee Colony algorithm is simple structure, easy operation and fast convergence, it improved the problem of path planning which in the slow convergence and easy to local optimal solution in previous path planning algorithms. While the RRT has excellent characteristics in seeking unknown area, it is unstable for each planning. We herein combine the characteristics of artificial bee colony with the RRT algorithms to deal with the problem of obstacles and make practical simulation in a mobile robot. In summary, it is shown that the data of the new algorithm on path planning are much more effective and stable than those of either single ABC or single RRT, and the path is the shortest.
