臺灣博碩士論文加值系統

隨著物聯網 (Internet of Things, IoT) 的興起、無線感測網路 (WSN) 技 術成熟,與智慧製造 (Wisdom Manufacturing) 的需求下,增加了判斷事件 狀態,甚至是預測事件狀態之需求。然而判斷或是預測事件之狀態,最重 要的環節在於資料,因此資料的蒐集與資料的正確性為最關鍵之因素,所 以資料的蒐集定義與處理可謂預測精準度之關鍵。一般常見的資料蒐集方 式都是藉由實體感測器採集實體環境的原始資料[1],再依照不同的應用做不同的資料轉換。但隨著應用的複雜度提升，以及在事件預測中不可或缺之領域知識，使得有些事件的判斷不能僅僅只依靠某種實體環境原始資料來做判斷。
本研究針對蒐集情境資料過程中所遇到之問題進行探討，將相關解決方法塑模而成資料蒐集模型 (Data Collection Model)，以資料流程圖 (Data Flow Diagram, DFD) 及相關處理工作演算法呈現資料蒐集模型。方法中應用複合式感測器概念 (Composite Sensor Concept, CSC) 以及排位得分函式 (Rank-Score Function, RSF) 資料融合 (Data Fusion) 技術,構成資料蒐集模型之主要兩個處理工作 (Process)，並提出領域知識 (Domain Knowledge) 規則格式。
本研究藉由複合式感測器概念 (Composite Sensor Concept)[1]將各種感測器間之關係、操作也考慮進來，以及排位得分函式 (Rank-Score Function, RSF) 之資料融合技術[2]對不正確之資料予以排除，並額外可對事件進行偵測、抑或對領域知識進行修正，提出資料蒐集模型 (Data Collection Model)，以達到能夠動態調整、定義資料蒐集來源與領域知識，並應用資料融合技術以提升事件偵測之準確性 。

With the rise of the Internet of Things (IoT), wireless sensor network (WSN) technology is mature, and the demand of wisdom manufacturing to
increase the demand for event state judging, or even for event state predicting. However, judging or predicting event state , the most important part is the information, so the data collection and data accuracy is the most critical factor, so the definition of data collection and processing is the key of the accuracy for predicting. General common data collection methods are collecting raw data in environment by the physical sensor, and then in accordance with different applications do different data conversion. However, with the complexity of the application upgrade, as well as in the event state predicting is very important domain knowledge, making the judgment of some events can not only rely on raw data in physical environment.
In this study, the problems encountered in the context data collection were discussed, and the data collection model was modeled by the relevant solutions. The data collection model was presented with the data flow diagram (DFD) and related processing algorithms. The method uses Composite Sensor Concept (CSC) and Rank-Score Function (RSF) of Data fusion technology, which constitutes the main two data collection model process, and proposed domain knowledge rules format.
In this study, consider the relationship between the various sensors, operation by the Composite Sensor Concept (CSC), and exclude the incorrect information, and additional to detect the event, or fix domain knowledge by Rank-Score Function (RSF) of Data fusion technology. The Data Collection Model is proposed to dynamically adjust and define the source of data collection and domain knowledge and apply data fusion technology to improve the accuracy of event detection.

摘要......i
Abstract......iii
誌謝......v
目錄......vi
表目錄......vii
圖目錄......viii
第一章 緒論......1
1.1 研究背景與動機......
1 1.2 研究目的......1
1.3 論文架構......1
第二章 文獻探討......3
2.1 情境的定義......3
2.2 資料蒐集......3
2.3 資料融合......5
第三章 研究方法......8
3.1 資料蒐集模型之環境架構......8
3.2 資料蒐集模型之運作機制......10
3.3 資料蒐集模型之運作流程......10
第四章 實例分析應用......6
4.1 實例環境......26
4.2 實例設計......27
4.3 實例分析應用......29
第五章 結論......31
參考文獻......32
Extended Abstract......33
簡歷 (CV) ......37

[1] Yusheng, X. U., Zhixin, M. A., Xiaoyun, C., & Lian, L. I. (2008). A composite sensor- based context modeling method for context- aware pervasive computing. In IMECS, March 2008.
[2] D. F. Hsu and I. Taksa, “Comparing rank and score combination methods for data fusion in information retrieval,” Inf. Retr. Boston., vol. 8, no. 3, pp. 449–480, 2005.
[3] A. K. Dey, Providing Architectural Support for Building Context-Aware Applications, Ph. D thesis, Georgia Institute of Technology, Atlanta, U.S.A., 2000.
[4] 吳瑞豐, “最佳權重法應用於交通資料融合, ” 國立交通大學碩士論文, 2006.
[5] Z. Yuan, R. Tan, G. Xing, C. Lu, Y. Chen and J. Wang, “Fast sensor placement algorithms for fusion-based target detection,” IEEE Real-Time Systems Symposium, 2008, pp. 103-112.
[6] 張佳雯, “資料融合於異質性資料推估路段行駛速率之研究,” 國立臺灣大 學碩士論文, 2007.
[7] R.J. Linn and D.L. Hall, “A Survey of Multi-sensor Data Fusion Systems” Proceedings of the SPIE-The International Society for Optical Engineering, Vol 1470, pp. 13-29, 1991.
[8] H. Wu, M. Siegel, R. Stiefelhagen, and J. Yang, “Sensor Fusion Using Dempster- Shafer Theory”, IEEE Instrum. Meas. Technol. Conf., vol. 12, no. May, pp. 21–23, 2002.
[9] H. Wu, M. Siegel, and S. Ablay, “Sensor Fusion Using Dempster-Shafer Theory II: Static Weighting and Kalman Filter-like Dynamic Weighting” Instrum. Meas. Technol. Conf. 2003. IMTC ’03. Proc. 20th IEEE, vol. 2, no. May, pp. 907–912 vol.2, 2003.
[10] Li, Y., Hsu, D.F., and Chung, S.M. "Combining Multiple Feature Selection Methods for Text Categorization by Using Rank-Score Characteristics". 21st IEEE International Conference on Tools with Artificial Intelligence, (2009), 508-517.
[11] Thomas, J, Cashman, Harry, J, Rosenblatt, Gary, B, Shelly. Systems Analysis and Design By Shelly, Cashman, & Rosenblatt (7th Edition). Boston:CourseTechnology, 2007.
[12] 謝萬澔, “應用類神經網路與振動訊號之微銑刀具狀態偵測系統開發,” 國 立中興大學碩士論文, 2009.
[13] 黃耀賢, “主軸振動與聲射訊號於微銑刀具磨耗監測之應用研究,” 國立 中興大學碩士論文, 2010.