請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/92617
標題: | 應用於多機器人平滑搬運瘦長物體之規劃方法 A Smooth Planning Algorithm for Transporting Slender Objects using Multi-Robot Systems |
作者: | 李冠賢 Kuan-Hsien Lee |
指導教授: | 詹魁元 Kuei-Yuan Chan |
關鍵字: | 路徑規劃,路徑連續性,瘦長物體搬運,多機器人系統, Path Planning,Path Continuity,Slender Object Transportation,Multi-Robot System, |
出版年 : | 2024 |
學位: | 碩士 |
摘要: | 隨著工業技術的進步,移動式機器人在各個領域中的應用日益廣泛。其中由於搬運為移動式機器人的基本能力之一,在現今許多情境中搬運任務已經實現高度自動化。然而,當搬運物體的尺寸增加時,單一機器人面臨載重等方面的能力限制,這促使多機器人系統成為能 夠靈活適應需求的理想替代方案。然而,目前對多機器人搬運系統的運動規劃通常採用特定的運動模式來簡化問題,這往往限制了系統的運動彈性,特別是在處理瘦長物體時,這可能導致更複雜的避障問題。
為了克服這些限制,本研究分析了多差速式機器人參與搬運系統後耦合的運動限制。我們確定了使用差速式機器人進行搬運時,被搬運物必須滿足路徑為 G2 連續以及朝向對路徑長參數為 C2 連續的條件。基於這些限制,本研究提出了一種新的運動規劃方法,將路徑和朝向 的規劃分為兩個階段,以最大程度發揮系統的運動能力。具體而言,我們引入了自行建立的「朝向佔據地圖」和「沿著路徑的朝向網格地圖」來連接這兩個階段。接著,通過修改過的 A* 演算法在各階段中進行路徑規劃,獲得最佳結果。這種方法生成的路徑在滿足各機器人各輪速度連續的條件下,實現了等速率搬運物體,從而節省了以往切換運動模式所需的時間。同時,通過減少不必要的朝向變化,我們獲得了相對較短的總路徑長,進一步節省了搬運所需的能源。這項研究的方法不僅提高了搬運系統的靈活性和效能,同時也為解決目前搬運系統中存在的運動模式限制問題提供了新的洞察。 Mobile robots are now widely used in various applications, especially in automation of objects transportation. As the size of the object increases, transporting with a single robot may become infeasible due to the limitations such as their weight-bearing capacity. In such cases, employing multi-robot systems that can be flexibly configured becomes a suitable alternative. However, existing planning methods for multi-robot systems often rely on predefined motion modes to simplify planning, thereby limiting the overall motion flexibility of the system. Additionally, when dealing with slender objects, the use of motion modes can lead to more complex challenges related to obstacle avoidance. To address these issues, this research examines the coupled motion constraints arising from multiple differential-drived robots in transportation systems. It is established that the transported object must follow paths which satisfy G2 continuity and the orientation change must be C2 continuous with respect to the arc-length parameter of the path. A two-stage motion planning method is proposed, comprising path planning and orientation planning, with the goal of maximizing the utilization of system''s motion capabilities. In particular, A* algorithm is modified and utilized in each stage to achieve optimal results. The two stages are integrated using the Orientation Occupancy Map (OOM) and Orientation Grid Map along path (OGMap), constructed from static information. With continuous wheel velocities guaranteed for each robot, the generated paths using this approach enable the constant-speed transportation of objects, eliminating unnecessary time associated with motion mode switching. Furthermore, the method minimizes unnecessary changes in orientation, resulting in a relatively shorter total path length that conserves energy and enhances the overall efficiency of the transportation process. This research not only improves the flexibility and efficiency of transportation systems but also introduces a novel perspective on addressing challenges associated with motion modes. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/92617 |
DOI: | 10.6342/NTU202400757 |
全文授權: | 同意授權(全球公開) |
顯示於系所單位: | 機械工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-112-2.pdf | 6.29 MB | Adobe PDF | 檢視/開啟 |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。