查詢結果分析
來源資料
頁籤選單縮合
題名 | 田間自走式載具避障路徑規劃演算法之配置與實現=The Implementation and Realization of the Obstacle Avoidance Path Planning Algorithms for Field Autonomous Vehicle |
---|---|
作者姓名(中文) | 彭祥瑋; 羅茂榮; 陳澤民; | 書刊名 | 農林學報 |
卷期 | 62:4 2013.12[民102.12] |
頁次 | 頁291-304 |
分類號 | 448.94 |
關鍵詞 | 路徑規劃; 避障; A*演算法; 自主機器人; E-PUCK機器人; Path planning; Collision-free; A* algorithms; Autonomous vehicles; E-PUCK; |
語文 | 中文(Chinese) |
中文摘要 | 田間自走式載具在作業環境中,常常要考量節能與工作效率等問題,所以載具所行進的路徑除必須是安全無碰撞外,更須為距離行進最短或時間最短的路徑。本論文主要目的即在於實現以A*演算法來進行全域避障路徑之規劃。在本論文所建置的硬體架構下,針對Diagonal、Euclidean、Dijkstra 和Manhattan等不同的啟發函數(heuristic function),運用不同的障礙分佈環境場景進行比較。最後則使用本研究所撰寫的演算程式,藉由手動環境設定或應用網路攝影機(Webcam)偵測障礙物二種方式,於程式介面中建立環境資訊並進行載具路徑規劃與模擬。最後則於本研究所建立之測試平台以E-puck機器人來驗證演算法與路徑指令傳遞之成效。驗證結果顯示,本研究以A*演算法來進行全域避障路徑規劃,已可規劃出全無碰撞且距離最短之路徑,導引無人載具從起始點安全行走抵到終點完成自主避障之行走任務。 |
英文摘要 | The paths by which autonomous vehicles are guided have to be collision-free and the shortest distance moving or less time required since the powers demanded and work efficiencies of autonomous vehicles are always take into evaluate. The main purpose of the paper is employing A* algorithms to implement global avoidance path planning. Various heuristic functions such as Diagonal, Euclidean, Dijkstra and Manhattan for different obstacle distributions will be compared in this paper. After start point and destination are confirmed on the user interface, path planning and simulation can be accomplished on the same interface on which obstacle distributions can be established by manual setting or Webcam detecting. Finally, the E-puck autonomous vehicle is guided by the planned path and moves on the testing platform to validate the efficiency and realization of the algorithms to autonomous vehicles. The realization and guidance-testing results show that the developed path planning system dominated by A* algorithms is capable to provide a collision-free and the shortest path for the navigation of autonomous vehicles from start to destination. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。