本研究以差速驅動無人車，在Gazebo模擬環境中建立3台自主移動的無人車，並探討其在分散式架構下的路徑規劃與調度策略。系統採用A*演算法在靜態環境下進行路徑規劃，確保車輛以最小代價尋找最佳路徑抵達目標，各車輛之間透過TCP/IP通訊傳遞路徑點與車輛狀態，並依據任務接收時間作為主要優先級判斷，若多台車同時收到任務，則根據車輛編號進行優先級排序。
    當兩台無人車的路徑點發生重疊時，系統會依照優先級動態調整無人車行駛順序，優先級高的車輛可以先移動，優先級低的車輛則需要等待優先級高的車輛離開所有重疊的路徑點後再繼續行駛，本研究有效降低車輛間碰撞的風險，相比於中央調度系統具備更好的擴充性，以及更低的系統計算時間，使企業在面臨產能擴充需求時，能以低成本彈性增減車隊規模，無需受限於中央伺服器的運算瓶頸，降低自動化物流的導入門檻與維護成本。

This study develops a distributed scheduling system for differential-drive AMR using the Gazebo simulation environment. Three AMR employ the A* algorithm for path planning in a static map to minimize travel cost. Vehicles communicate via TCP/IP to share path points and status, while task reception time determines priority. When multiple vehicles receive tasks simultaneously, priority is further assigned by vehicle ID.
    In case of path conflicts, the system dynamically adjusts the movement order so that higher-priority vehicles proceed first. This decentralized approach reduces collision risk and improves scalability while maintaining lower computational load compared with centralized scheduling methods.

目錄
目錄	I
圖目錄	IV
表目錄	V
第一章 緒論	1
1.1前言	1
1.2實習機構簡介	2
1.3實習內容概述	4
1.4實習心得及自我期許	5
1.5研究動機與背景	6
1.6文獻探討	8
1.7名詞解釋	10
第二章 研究範圍與研究方法	13
2.1研究範圍	13
2.2研究方法	15
第三章 系統架構	27
3.1虛擬環境建立	27
3.2實驗設定	30
3.3綜合運作流程	31
3.4 A*(A-star)演算法路徑規劃	32
3.5無人車路徑點衝突決策機制	37
3.5.1路徑接收與任務判斷	37
3.5.2他車資訊整合	37
3.5.3衝突檢查與優先權判定	38
3.5.4協調命令發布與路徑廣播	39
3.5.5路徑點通過與衝突解除	39
3.6無人車控制策略	39
3.6.1系統層級架構	40
3.6.2運動控制策略	41
3.6.3通行權協調策略	41
第四章 實驗結果	43
4.1實驗環境與評估指標	43
4.1.1衝突協調機制之時序演示	43
4.1.2效能評估指標定義	47
4.2不同障礙物密度下之任務統計	52
4.3.1移動效率與單位網格耗時	56
4.3.2平均等待時間變化偏差	57
4.4衝突檢測與調度分析	58
4.4.1總路徑長度與生存者偏差驗證	58
4.4.2衝突點數量與協調負載分佈	59
4.4.3等待時間與衝突點之相關性分析	59
第五章 結論與未來展望	62
5.1結論	62
5.2未來展望	63
第六章 參考文獻	65
附錄	67
 
圖目錄
圖 1有利康股份有限公司	2
圖 2深度相機於RVIZ警示範圍	4
圖 3 GAZEBO模擬環境中之差速驅動無人車模型	13
圖 4研究方法流程圖	16
圖 5 GAZEBO之顯示畫面	19
圖 6  3台無人車路徑規劃示意圖	22
圖 7里程計運動模型示意圖	22
圖 8 差速驅動移動示意圖	23
圖 9里程計運動控制模組流程圖	26
圖 10 整體系統流程圖	31
圖 11 A* 路徑規劃流程圖	32
圖 12 A* 演算法流程示意圖1	34
圖 13 A* 演算法流程示意圖2	35
圖 14 A* 演算法流程示意圖3	36
圖 15 衝突檢查決策圖	38
圖 16 時序分解圖一	43
圖 17 時序分解圖二	44
圖 18 時序分解圖三	45
圖 19 時序分解圖四	46
圖 20 平均網格耗時與障礙物數量關係圖	53
圖 21實驗批次成功率與任務完成率與障礙物數量關係圖	54
圖 22 總路徑長與總衝突點與障礙物數量與平均等待秒數關係圖	55
圖 23 等待時間與衝突點關係圖	59

 
表目錄
表 1 國內外無人車調度與導航策略文獻比較彙整表	9
表 2 工廠環境設計參數表	27
表 3機器人採用差分驅動結構表	28
表 4 物理引擎參數表	29
表 5 任務狀態轉換與系統行為關係表	42
表 6 四種不同障礙物數量數據統計	52
表 7 十種隨機障礙物之任務統計表	67
表 8 二十種隨機障礙物之任務統計表	71
表 9 三十種隨機障礙物之任務統計表	75
表 10 四十種隨機障礙物之任務統計表	79

[1] D. Weyns and T. Holvoet, "Decentralized Control of Automatic Guided Vehicles: Applying Multi-Agent Systems in Practice," in OOPSLA Companion '08, pp. 663–674, 2008. DOI: 10.1145/1449814.1449819.
[2] M. De Ryck, M. Versteyhe, and F. Debrouwere, "Automated guided vehicle systems, state-of-the-art control algorithms and techniques," Journal of Manufacturing Systems, vol. 54, pp. 152–173, 2020. DOI: 10.1016/j.jmsy.2019.12.002.
[3] T. Schmidt, K.-B. Reith, N. Klein, and M. Däumler, "Research on Decentralized Control Strategies for Automated Vehicle-Based In-House Transport Systems: A Survey," Logistics Research, vol. 13, no. 10, 2020. DOI: 10.23773/2020_10.
[4] Q. Guo, H. Yao, Y. Liu, Z. Tang, X. Zhang, and N. Li, "A Distributed Conflict-Free Task Allocation Method for Multi-AGV Systems," Electronics, vol. 12, no. 18, Art. no. 3877, 2023. DOI: 10.3390/electronics12183877.
[5] J. Santos, P. M. Rebelo, L. F. Rocha, P. Costa, and G. Veiga, "A* Based Routing and Scheduling Modules for Multiple AGVs in an Industrial Scenario," Robotics, vol. 10, no. 2, Art. no. 72, 2021. DOI: 10.3390/robotics10020072.
[6] Z. B. Rivera, "Unmanned Ground Vehicle Modelling in Gazebo/ROS," Machines, vol. 7, no. 2, Art. no. 42, 2019. DOI: 10.3390/machines7020042.
[7] R. Walenta, C. Schiffer, M. Rambow, and A. Verl, "A Decentralised System Approach for Controlling AGVs with ROS," in 2017 IEEE AFRICON, Cape Town, South Africa, 2017, pp. 1199–1204. DOI: 10.1109/AFRCON.2017.8095653. [8] S. Krishnan, G. A. Rajagopalan, S. Kandhasamy, and M. Shanmugavel, "Towards Scalable Continuous-Time Trajectory Optimization for Multi-Robot Navigation," arXiv preprint, 2019.
[9] I. Draganjac, D. Miklić, Z. Kovačić, G. Vasiljević, and S. Bogdan, "Decentralized Control of Multi-AGV Systems in Autonomous Warehousing Applications," IEEE Transactions on Automation Science and Engineering, vol. 13, no. 4, pp. 1433–1447, 2016. DOI: 10.1109/TASE.2016.2603781.
[10] H. Martínez-Barberá and D. Herrero-Pérez, "Autonomous navigation of an automated guided vehicle in industrial environments," Robotics and Computer-Integrated Manufacturing, vol. 26, no. 4, pp. 296–311, 2010.
[11] V. Digani, L. Sabattini, C. Secchi, and C. Fantuzzi, "Hierarchical traffic management of multi-AGV systems with deadlock prevention applied to industrial environments," in 2015 IEEE International Conference on Robotics and Automation (ICRA), Seattle, WA, USA, pp. 6072–6077, 2015.
[12] S. Warita and K. Fujita, "Online planning for autonomous mobile robots with different objectives in warehouse commissioning task," Information, vol. 15, no. 3, Art. no. 130, 2024. doi: 10.3390/info15030130.
[13] T. Raamets, J. Majak, K. Karjust, K. Mahmood, and A. Hermaste, "Autonomous mobile robots for production logistics: A process optimization model modification," Proceedings of the Estonian Academy of Sciences, vol. 73, no. 2, pp. 134–141, 2024. doi: 10.3176/proc.2024.2.06.
[14] A. C. Jiménez, V. García-Díaz, and S. Bolaños, "A decentralized framework for multi-agent robotic systems," Sensors, vol. 18, no. 2, Art. no. 417, 2018.
[15] H. Kim and J. Kim, "Web-based real-time alarm and teleoperation system for autonomous navigation failures using ROS 1 and ROS 2," Sensors, vol. 23, no. 19, Art. no. 8089, 2023. doi: 10.3390/s23198089.