駕駛人操控介入意圖推論之研究｜國立臺灣科技大學博碩士論文系統

簡易檢索 / 詳目顯示

回結果列表

研究生：	蘇威丞 Wei-Cheng Su
論文名稱：	駕駛人操控介入意圖推論之研究 The Investigation of a Vehicle Driver Override Inference System
指導教授：	陳亮光 Liang-kuang Chen
口試委員:	徐繼聖 Gee-Sern Jison Hsu 洪博雄 Boe-Shong Hong
學位類別：	碩士 Master
系所名稱：	工程學院 - 機械工程系 Department of Mechanical Engineering
論文出版年：	2010
畢業學年度：	98
語文別：	中文
論文頁數：	89
中文關鍵詞：	隱馬可夫模型、動態貝氏網路、駕駛意圖
外文關鍵詞：	Hidden Markov Model, Dynamic Bayesian Network, Driver Intention
相關次數：	點閱：165 下載：3
分享至:	分享至facebook 分享至twitter

查詢本校圖書館目錄查詢臺灣博碩士論文知識加值系統勘誤回報

本研究透過駕駛模擬器進行實驗收集所需要的數據，並將數據整理成訓練以及驗證所需要的資料庫，然後利用此資料庫進行隱馬可夫模型與動態貝氏網路的模型訓練，並探討不同資訊源以及有無疲勞資訊對識別駕駛人是否意圖介入主動轉向輔助系統之影響與優劣性，最後將訓練好的機率模型實現於線上駕駛模擬器中，使系統能夠識別出正確的駕駛人意圖發生點，以及可適應駕駛狀態的切換時間。最後結合控制權重調節器，當駕駛意圖改變時，調節器可藉由識別出的駕駛行為資訊，調節駕駛人與安全系統間的控制權重來避免衝突行為，以確保駕駛人的行車安全

In this research a vehicle driver override inference algorithm is developed. The data is collected from a driving simulator several subject human drivers, and divided into training data set and validation data set. Two classification models, namely the hidden Markov model and the dynamic Bayesian network, are trained. Several different choices of information sources are investigated to compare their effects on the classification performance. The trained classification model is implemented on the driving simulator to illustrate the on-line inference capability, and the possible integration with a decision making rule for the control authority of a lane keeping assist controller. The driving simulator experiments indicate that the inferred driver override intention can effectively assist the control authority determination and consequently successfully prevent the conflict between the driver and the lane keeping assist controller during a lane change maneuver.

摘要............................................................................................................................... I
ABSTRACT .................................................................................................................. II
目錄.............................................................................................................................. III
圖目錄........................................................................................................................... V
表目錄....................................................................................................................... VIII
第一章 緒論.................................................................................................................. 1
1研究背景與動機.............................................................................................. 1
2文獻探討.......................................................................................................... 2
2.1 使用HMM判別駕駛意圖之文獻探討.............................................. 2
2.2使用BN或DBN應用於駕駛意圖相關文獻探討............................. 4
2.3使用其它方式判別駕駛意圖相關文獻探討....................................... 5
2.4 文獻結果討論...................................................................................... 6
3研究架構與工作項目...................................................................................... 7
1-4預期貢獻 ......................................................................................................... 8
第二章 基礎理論.......................................................................................................... 9
1隱馬可夫模型.................................................................................................. 9
1.1隱馬可夫模型基本組成元素............................................................. 10
1.2 評估HMM模型................................................................................ 11
1.3推論最佳狀態序列 (Inference) ......................................................... 14
1.4學習 (Learning) ................................................................................. 16
2 動態貝氏網路 (Dynamic Bayesian Network, DBN) .................................. 18
2.1 貝氏網路 (Bayesian Network, BN) .................................................. 18
2.2動態貝氏網路基礎與其應用文獻探討............................................. 19
2.3動態貝氏網路基本組成元素............................................................. 20
2.4動態貝氏網路參數學習..................................................................... 21
2.5 DBN推論 ........................................................................................... 21
2.6 HMM 與 DBN 之比較 [27] ........................................................... 22
IV
3駕駛人狀態判別............................................................................................ 22
3.1 兩種疲勞狀態判別機制.................................................................... 22
3.2 卡爾曼濾波器.................................................................................... 23
5 控制權重調節器........................................................................................... 23
第三章 實驗設計與資料處理.................................................................................... 25
1 線上駕駛模擬器硬體架構........................................................................... 25
2 實驗設計與資料取得................................................................................... 28
3 實驗數據資料處理....................................................................................... 30
3.1 建立狀態序列.................................................................................... 30
3.2 資料截取與合併................................................................................ 31
3.3 資料分群............................................................................................ 33
4 HMM及DBN網路結構圖 .......................................................................... 39
4.1 HMM之網路結構圖.......................................................................... 39
4.2 DBN之網路結構圖 ........................................................................... 42
第四章 實驗結果分析與討論.................................................................................... 47
1 駕駛意圖推論決定規則 (Decision Rules).................................................. 47
2 資訊源探討................................................................................................... 47
2.1 單一資訊源........................................................................................ 47
2.2 前車資訊及雙資訊源........................................................................ 50
2.3 多資訊源............................................................................................ 54
3 疲勞資訊改善效果....................................................................................... 56
4 線上DO意圖識別與控制調節器之效能驗證實驗 ................................... 60
第五章 結論與未來展望............................................................................................ 65
1 總結與結論................................................................................................... 65
2 未來工作與展望........................................................................................... 66
參考文獻...................................................................................................................... 67
附錄A 線上實驗規劃 ................................................................................................ 71
附錄B 程式架構Pseudo-Code .............................................................................. 76

                                

[1]. 楊釧暉, “結合駕駛人資訊之車道維持輔助控制器設計”, 國立臺灣科技大學機械工程系碩士論文, 台灣台北, 2009
[2]. Pentland, A., and Liu, A., “Modeling and prediction of human behavior,” Neural Computation, 11:229-242, 1999.
[3]. Kuge, N., Yamamura, T., Shimoyama, O., and Liu, A., “A driver behavior recognition method based on a driver model framework”, Technical report, Society of Automotive Engineers, Inc., 1998.
[4]. Oliver, N., and Pentland, A.P., “Graphical models for driver behavior recognition in a smartcar”, Proceeding of the IEEE Intelligent Vehicles Symposium, pages 7-12, 2000.
[5]. Polling, D., Mulder, M., van Paassen, M.M., and Chu, Q.P., “Inferring the driver’s lane change intention using context-based dynamic Bayesian networks”, Disciplinary Group Control and Simulation Faculty of Aerospace Engineering, Delft, Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics, v 1, p 853-858, 2005, IEEE Systems, Man and Cybernetics Society, Proceedings - 2005 International Conference on Systems, Man and Cybernetics
[6]. Mizushima, and Takuya, “Direct yaw-moment control adapted to driver behavior recognition”, 2006 SICE-ICASE International Joint Conference, p 534-539, 2006, 2006 SICE-ICASE International Joint Conference
[7]. Mitrovic, D., “Reliable method for driving events recognition”, IEEE Transactions on Intelligent Transportation Systems, v 6, n 2, June 2005, p 198-205
[8]. Berndt, H., Emmert, J., and Dietmayer, K., “Continuous Driver Intention Recognition with Hidden Markov Models”, ITSC 2008 11th International IEEE Conference on Intelligent Transportation Systems, Oct 2008, p1189-1194
68
[9]. Nobuyuki, K., Tomohiro, Y., and Osamu, S., “A driver behavior recognition method based on a driver model framework”, Society of Automotive Engineers Publication, 1998.
[10]. Tezuka, S., Soma, H., and Tanifuji, K., “A Study of Driver Behavior Inference Model at Time of Lane Change using Bayesian Networks”, Industrial Technology, 2006. ICIT 2006. IEEE International Conference on 15-17 Dec. 2006 Page(s):2308 - 2313 Digital Object Identifier 10.1109/ICIT.2006.372650
[11]. McCall, J.C., Wipf, D.P., Trivedi, M.M., and Rao, B.D., “Lane Change Intent Analysis Using Robust Operators and Sparse Bayesian Learning”, Intelligent Transportation Systems, IEEE Transactions on V8, Issue 3, Sept. 2007 Page(s):431 - 440 Digital Object Identifier 10.1109/TITS.2007.
[12]. McCall, J.C. and Trivedi, M.M.,” Driver Behavior and Situation Aware Brake Assistance for Intelligent Vehicles,” Proceedings of the IEEE, v2 , 2007 ,p374-387
[13]. Kumagai, Toru, Akamatsu, and Motoyuki, “Prediction of human driving behavior using dynamic bayesian networks”, IEICE Transactions on Information and Systems, Volume E89-D, Issue 2,Feb 2006, p857-860
[14]. Cheng, S. Y., Park, S., and Trivedi, M. M., “Multi-spectral and multi-perspective video arrays for driver body tracking and activity analysis,” Computer Vision and Image Understanding: Special Issue on Advances in Vision Algorithms and Systems Beyond the Visible Spectrum, 2006.
[15]. Gunnarsson, J., Svensson, L., Bengtsson, E., and Danielsson, L., “Joint Driver Intention Classification and Tracking of Vehicles”, IEEE Nonlinear Statistical Signal Processing Workshop, Sept 2006 ,p95-98
[16]. Ohashi, L., Yamaguchi, T., and Tamai, I., “Humane automotive system using driver intention recognition”, SICE 2004 Annual Conference , v2 , Aug 2004 , p1164-1167
[17]. Zhao, G., Peng, C., and Wang, X., “Intelligent control for AMT based on driver’s intention and ANFIS decision-making” WCICA 2008. 7th World Congress on Intelligent Control and Automation, June 2008 , p1-6
69
[18]. Itoh, M., Yoshimura, K., and Inagaki, T., “Inference of Large Truck Driver’s Intent to Change Lanes to Pass a Lead Vehicle via Analyses of Driver’s Eye Glance Behavior in the Real World”, SICE, 2007 Annual Conference 17-20 Sept. 2007 Page(s):2385 - 2389 Digital Object Identifier 10.1109/SICE.2007.4421387
[19]. Risack, R., Mohler, N., and Enkelmann, W., “A video-based lane keeping assistant”, in Proc. IEEE Intell. Veh. Symp., Oct. 2000, pp. 506–511.
[20]. Salvucci, D.D., “Inferring driver intent: A case study in lane-change detection”, in Proc. Human Factors Ergonom. Soc. 48th Annu. Meeting, 2004, pp. 2228–2231.
[21]. Salvucci, D.D., Mandalia, H.M., Kuge, N., and Yamamura, T., “Lane-change detection using a computational driver model”, Human Factors: Human Factors: The Journal of the Human Factors and Ergonomics Society, Vol. 49, No. 3, 532-542 (2007)
[22]. Rabiner, Lawrence R., “Tutorial on hidden Markov models and selected applications in speech recognition”, Proceedings of the IEEE, v 77, n 2, p 257-286, Feb 1989
[23]. 顧正偉, “利用多觀察值型隱馬可夫模型進行人體動作辨識 ”, 國立交通大學資訊工程系所,碩士論文93
[24]. Li, X., “Training hidden Markov models with multiple observations - a combinatorial method”, IEEE Transactions on Pattern Analysis and Machine Intelligence, v 22, n 4, p 371-377, 2000
[25]. Kessentini, Y., Paquet, T., and Benhamadou, A., “A multi-stream hmm-based approach for off-line multi-script handwritten word recognition”, In Proc. 10th, pages 147–152, 2008.
[26]. Somervuo, P., “Speech Recognition Using Context Vectors And Multiple Feature Streams”, Helsinki University of Technology, Master Thesis, 1996
[27]. Murphy, K., “Dynamic Bayesian Networks: Representation, inference and Learning”, Phd Thesis. UC Berkeley, Computer Science Division, July 2002.
70
[28]. 彭孟璿, “線上駕駛人建模與駕駛狀態判別之正確性評估,” 國立臺灣科技大學機械工程系碩士論文, 台灣台北, 2007.
[29]. 黃品誠, “國內穩定跟車模式之建構”, 逢甲大學交通工程與管理學系碩士在職專班碩士論文, 2005
[30]. Kevin Murphy's Hidden Markov Model (HMM) Toolbox, URL: http://www.cs.ubc.ca/~murphyk/Software/HMM/hmm.html
[31]. Kevin Murphy's Bayes Net Toolbox (BNT), URL: http://code.google.com/p/bnt/

簡易檢索 / 詳目顯示

相關論文