以多代理人技術設計智慧型搬運車系統

Design of A Smart Automated Guided Vehicle System with Mutil-Agent Technology

王彥凱(Yen-Kai, Wang)

指導教授 ： 項衛中

中原大學/電機資訊學院/工業與系統工程研究所/碩士(2018年)

https://doi.org/10.6840/cycu201800086

摘要

隨著物聯網的概念被提出，已經有越來越多的應用，使得作業模式開始漸漸地轉變。在未來的智慧工廠生產模式中，製造流程將更具彈性並能夠快速轉換。整合無人搬運車與其他物流設備的運作，將能使企業快速且準確地配合產線需求。
本研究運用多代理人技術與物聯網三層式架構的概念，設計一個智慧型搬運車系統，並反轉無人搬運車過去的主從式關係，變成以無人搬運車為主，能夠自行相互溝通協調，找出最適合的決策。本研究運用統一塑模語言呈現多代理人間系統化、結構化以及模組化的設計，並展現各代理人間的合作運行流程，並設計相關實驗以驗證此系統之可行性。本研究以Arduino 與相關模組建構無人搬運車，並且導入物聯網概念進行搬運任務，以情境實驗模擬真實情境下的執行結果。本研究已經完成了有無使用多代理人技術的實驗，並驗證物聯網的三層式架構，結果顯示使用多代理人技術的系統所花費時間低於未使用多代理人技術，而且設備的穩定性對於實驗結果有很大的影響，此結果可供導入無人搬運車技術的企業當作參考。

關鍵字

物聯網 ； 多代理人 ； 統一塑模語言

並列摘要

As the concept of the internet of things (IOT) was proposed, many IOT applications have been realized and this trend began to gradually change the manufacturing operations. In the future smart factory, the manufacturing process will be more flexible and can be quickly modified. The automated guided vehicle (AGV) and other logistics equipments will enable plants to quickly and accurately align with their customer needs.
This research used a three-layer structure of IOT with multi-agent technology to design a smart AGV system, and AGVs communicate and coordinate with each other to find the most suitable solution to fulfill the requirements in this system. This research applied unified modeling language (UML) to present the systematic, structured and modular design of this multi-agent system, and this research also developed how agents cooperated in performing different tasks, Experiments were proposed and conducted to verify this AGV system. The experiments used Arduino board and related modules to build simple AGV models. The three-layer structure of IOT was applied in the experiments, and control schemes with and without multi-agent technology were implemented to perform the material handling tasks. The experiments simulated the real world situations and AGV models operated to finish the tasks. The results show that the system with multi-agent technology spent less time than the one without multi-agent technology. The stability of the electronic devices has a great impact on the experimental results, and this result can be used for developing the AGV technology as a reference.

並列關鍵字

Internet of Things ； Multi Agent ； Unified Modeling Language

參考文獻

19. 尹者賢. (2010). 共用儲位自動倉儲系統作業模式之研究. 中原大學工業工程學系碩士學位論文, 1-65.

22. 沈傑銘. (2015). 運用物聯網技術於智慧型票卡系統設計之規劃. 中原大學工業與系統工程學系碩士學位論文, 1-62.

23. 岳希洋. (2016). 運用物聯網技術於無人搬運車系統之設計. 中原大學工業與系統工程學系碩士學位論文, 1-120.

26. 徐婉瑄. (2003). 以UML構建之整合性物流派車資訊系統. 中原大學工業工程學系碩士學位論文, 1-116.

29. 曾溫祥. (2005). 多代理人智慧家庭系統之設計與實作. 國立臺北科技大學自動化科技研究所碩士學位論文, 1-121.