The Study of Using Interpreted Petri Net (IPN) and Dynamic Bayesian Network (DBN) Approach to Construct an Internet of Things (IoT) Based Intelligent Agent under the Fog Computing Environment

瀏覽統計 Email 通知 RSS Feed

簡歷

基本資料

Project title

Code/計畫編號

MOST104-2633-E019-001

Translated Name/計畫中文名

以動態貝式網路與詮釋性裴氏網路方法建構霧運算環境下之物聯網智慧型代理人

Project Coordinator/計畫主持人

Meng-Ru Tu

Funding Organization/主管機關

National Science and Technology Council

Department/Unit

Department of Transportation Science

Website

https://www.grb.gov.tw/search/planDetail?id=11568674

Year

2015

Start date/計畫起

01-08-2015

Expected Completion/計畫迄

31-07-2016

Bugetid/研究經費

533千元

ResearchField/研究領域

工業工程
資訊科學--軟體
資訊工程--硬體工程

隨著越來越多的物體變得智能，或者能夠感應所在環境，接入互聯網，甚至接受遠端指令，物聯網的發展讓更多的物體加入網路，例如從飛機引擎到冰箱的一切物體都可加入無線網路，組成了物聯網。隨著物聯網和移動互聯網的高速發展，人們越來越依賴雲計算，但隨著物聯網節點數與物聯網資料量不斷增加，使現有的雲端運算架構不堪負荷。為因應物聯網的分散式網路環境與資料量過大等挑戰，最近學界與業界相繼提出一種分散式霧運算架構觀念，讓貼近物聯網的環境的裝置能具備資料預處理能力與較多的智慧以減輕雲端系統的負擔。因此本計晝研究目標為發展霧運算環境下之物聯網智慧型代理人系統架構，讓該架構在最大程度上可滿足於工業與民生等不同的物聯網應用需求，並且希望這樣的軟體系統架構能在最大幅度上實作到不同種類的物聯網裝置上以提昇物聯網裝置之自我控制能力並降低系統開發的複雜度。本研究以BDI (信念-慾望-意圖) 推理模式結合態貝式網路來設計代理人的認知模型，以詮釋性裴氏網路來建構代理人的控制模型。最後透過系統實作與實驗來評估此架構的可行性與效益。"As more and more objects become intelligent or capable of sensing their environment, accessing the internet, and even accepting remote commands, the development of internet of things (IoT) are making more objects join the internet, from aircraft engines to refrigerators, they all can join the wireless network to form a big IoT network. With the rapid development of the IoT and mobile internet, people are increasingly rely on cloud computing ； but with the number of device nodes in IoT and increasing amount of IoT data, the cloud system will be overloaded and the current cloud computing architecture will not sufficient to handle these challenges. Academia and industry have recently proposed a distributed fog computing architecture concepts to cope with the aforementioned challenges bring up by IoT. Fog computing endow devices which close to the IoT environment with more wisdom and data preprocessing capabilities in order to reduce and share the burden on the cloud system. The research objective for this proposal is to study and develop IoT based intelligent agent system architecture under the fog computing environment, allowing the proposed architecture, to the maximum extent, to meet the demand from both industrial and consumer applications. We also hope that this proposed IoT software architecture can be implemented different types of IoT devices to enhance the self-control capabilities, and reduce the complexity of IoT system development. In this study, we combine BDI (Beliefs-Desire-Intention) reasoning model with Dynamic Bayesian Network to develop agent’s cognitive model. We use Interpreted Petri net modeling scheme to develop agent’s control model. Finally, we will evaluate the feasibility and benefits of our proposed system architecture through systems implementing and experiments."

Keyword(s)

物聯網
霧運算
智慧型代理人
動態貝式網路
詮釋性裴氏網路
Internet of Things (IoT)
Fog Computing
Intelligent Agent
Dynamic Bayesian Network
Interpreted Petri net

DSpace CRIS

The Study of Using Interpreted Petri Net (IPN) and Dynamic Bayesian Network (DBN) Approach to Construct an Internet of Things (IoT) Based Intelligent Agent under the Fog Computing Environment

基本資料

Description