二階多無人機系統於三維空間中之分散式編隊避碰控制

本論文旨在研究多代理人系統(MAS)的三維編隊調控，這些代理人以分散式之通訊方式，透過特定的通信連結進行控制，我們的控制器設計主要針對微型無人機（UAV），因此對於代理人之動態模型採用了二階積分器系統。近年來因無人機之性能提升與價格下降，使得其應用廣泛，如資料探勘、地圖建設及搜救任務等，而相關控制研究也備受關注。本論文的控制目標是實現編排幾何隊形與追蹤參考軌跡，同時確保在任何時刻無人機能避免碰撞。目前關於編隊控制文獻中，處理大量機群仍多以考慮全域通訊關係為主，在避碰演算法上，大多以兩兩互相避碰為主要策略，因此當運用於真實的多無人機群中會遇到相當大的障礙。在本研究中，我們所考慮之多無人機系統在編隊過程中只使用来自其鄰域且有通訊關係之無人機資訊，並在追蹤參考軌跡時會保持通訊連接以維持整體系統之通訊關係。而控制器之避碰演算法是基於鄰域機間的相對距離、相對速度與系統內的優先權，來達成避碰，可有效降低機載電腦之計算負擔，與解決死結(deadlock)問題。此外，透過我們所提出的隊形描述方法與適應性控制架構，本論文之控制器亦能使無人機群以給定之順序關係排成所需要之隊形。最後，我們提供一些模擬場景與結果以驗證理論推導。

關鍵字

編隊控制；避免碰撞；多代理人系統；無人機

並列摘要

This thesis considers the formation and maneuver control of multi-agent systems in the three-dimensional space. The agents are controlled in a distributed manner with given communication links. Our design is mainly for the micro unmanned aerial vehicles (UAVs) and thus a second-order integrator model is adopted to formulate the dynamics of each vehicle. In recent years, the improved performance and lower prices of UAVs have led to a wide applications, such as data exploration, map construction, and search/rescue missions, and the related control studies have also drawn significant attention. The control objectives of this thesis are to achieve desired formation and fulfill the reference trajectory tracking, meanwhile, guarantee collision avoidance at any time. To this end, we adopt a multi-agent framework to address our problem in this thesis. The agents only use information from their neighbors and keep the connectivity to maintain the communication while tracking the reference trajectory. Our collision avoidance algorithm is based on relative distance, relative velocity, and the priority in the system. By incorporating such algorithm, not only the computational burden on the on-board computer is reduced but also the deadlock problem is satisfactorily solved. In the current literature on formation control, handling large number of UAVs basically remains under centralized and reciprocal avoidance strategy, and hence encounters great hurdle when the work needs for real implementation. Besides this advantage of our approach, our formulation can even drive the agents to form a pattern of certain shape following a particular agent order. To verify the effectiveness of our approach several simulation experiments have been conducted, and highly promising results have been obtained.

並列關鍵字

maneuver control ； collision avoidance ； multi-agent system ； unmanned aerial vehicle

參考文獻

[1]Z. Lin, L. Wang, Z. Han, and M. Fu, “Distributed formation control of multi-agent systems using complex laplacian,”IEEE Transactions on Automatic Control, vol.59, no. 7, pp. 1765–1777, 2014.

Google Scholar

[2]S. Zhao, “Affine formation maneuver control of multiagent systems,”IEEE Transactions on Automatic Control, vol. 63, no. 12, pp. 4140–4155, 2018.

Google Scholar

[3]H. Ebel and P. Eberhard, “A comparative look at two formation control approaches based on optimization and algebraic graph theory,”Robotics and Autonomous Systems, vol. 136, p. 103686, 2021.

Google Scholar

[4]H.Rezaee and F.Abdollahi, “Pursuit formation control scheme for double-integrator multi-agent systems,” IFAC Proceedings Volumes, vol. 47, no. 3, pp. 10054–10059,2014.[5]M.H.Trinh, C.VanNguyen, Y.H.Lim, and H.S.Ahn, “Matrix-weighted consensus and its applications,” Automatica, vol. 89, pp. 415–419, 2018.

Google Scholar

[6]D. R. Foight, M. H. de Badyn, and M. Mesbahi, “Performance and design of consensus on matrix-weighted and time-scaled graphs,” IEEE Transactions on Control of Network Systems, vol. 7, no. 4, pp. 1812–1822, 2020

Google Scholar

國際替代計量

二階多無人機系統於三維空間中之分散式編隊避碰控制

未授權

主題瀏覽