以模糊類神經理論與混沌理論構建高速公路事件偵測演算法
9
0
0
全文
(2)
(3) . Freeway Incident Detection Algorithms: Fuzzy-Neural-Based and Chaotic-Based Approaches .
(4). . StudentYeh-Chieh Huang AdvisorLawrence W. Lan. . . ! " # $ % & '. A Dissertation Submitted to Institute of Traffic and Transportation College of Management National Chiao Tung University in partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in Engineering. July 2005 Hsinchu, Taiwan, Republic of China. . i.
(5) .
(6)
(7)
(8). . . . . . . . . . . . . .
(9). . . . . . . #. $. :. 4. S. %& 5. M. N. b. . '. . (. 6. 7. d. s]. t. u. v. R. 1. w.
(10). x. ? T. >. U g. y. ). > S. `. . <= . ef. ;. OPQ c. *. %+. . @A. 6. hi. z. {. . 7 j. |. ,. -. B.
(11). C. g. . 6. G. 7. Z. D. ~. 1 E. Y B. 0. D X. A. }. ./. VW k. 1. . 2 F. 6. 3. 1 7. l. . 4 G. Z. 5 A. <H. B. [. . / \. I. . 6 H. m. 9 K. G. !" 8. J _. n. 7. I. ]^ J. . L A. o. . B. p. ` q. a r. . . S . . . . ¡. . . ®. ¡. ,. Ø. µ. Ù. . ç ×. 8 ß. ¬. /. . è X. . õ. f g. äå. J. . hø. ` l. ~. U. L. ù. ê. g. Oë. ì û. Ó /. Ô. ç ii. Ó. Ô. Ü ð. . . . ñ. '. ¶. ó. >. ,. 6. . . ¼. ?. % <!. 3 ö. Ì. N l. 2 õ. l *. Ë M. . . µ Ê. C. <X . . . <. 7. ã. l L. º» (. ô µ. ` ×. %+ ò. Æ. Ö. K. . Ã. . ). Â. <. ¿. e1 . *. ¤. d . J. P . ® . J. /. . I. . . H -. . µ. . & . . . ¬ . . &. ) ï. É . Õ d. d î. S. 5. . . Á ¾. 1 £. . ½ u. . í. ¸. . d. . ¢ 4. . ¼ Í. . Q. ¶. º» ·. È /. : â. V w. . . . <J. . µ. ¡. . « . . * Ò. á é. ú. d ®. . 4. <J. ³. . <l Æ. ¡ Ñ. Û. ¤ k. c Ð. . <. ô. b Ç. . Å º». /. <Ú S. . l. a. . T . à. Ç S. B. . V¹. µ. . ´. Á. . ª . ¸. t . ¬. À. . © . L. E. ³. · . ¹ ¥. ². K. . . h¨ ±. ¿. L. 5. I. , . §. <. . . 9 ¦. -. ³. 8 . l. K . %+ . . ¡. ~ . . g J. º» . ß 9. I. Ï5 . ×. H /. Î. 9. . d. <H. 8. . . l. 2. L. <¥ ¦. 9. . ¤ °. 3. ` . £ 8. 0. Ä. +. ¢ . ¾. . . ¯. <O. . . . ½. . Ý. ¬. äå X. Þ æ ÷.
(12) Ç. \ 7. â C. . . 6 5. ~. ¼ æ. . . . . 6. 7. Ï5 ç. 8 `. Í. . . . . . . . !. . . . . iii. . . . 7 ®. . . . Ï5. /. Ò . . U Q. S . µ. T í. a . . S ´. . . . . 7. . ¹ . 6. g. ý ~. . >. _. Ï5 . Ê. . â. . . . \ . 7. . À . 6. 7. 9 ¼. Q. 6. 8 ã. /. /. 7 . ç. Q. 6 . æ. í. ,. . S. +. . ~ a. w. äå Ü. 9 >. f. . þ æ.
(13) è. . È . * 9. . ý . d. 8 . _. äå . ç . ü. . . ]. 7. . 7. Ê g. . 6 . . 0. äå !Ï . ~. . . ¡. ³. äå. æ. ç. .
(14) Freeway Incident Detection Algorithms: Fuzzy-Neural-Based and Chaotic-Based Approaches Student: Huang, Yeh-Chieh. Advisor: Dr. Lawrence W. Lan. Institute of Traffic and Transportation National Chiao Tung University. Abstract Safety and efficiency is the most important issue in transportation. To reduce traffic flow delay, damage to life and properties and social cost, and to increase efficiency and safety in transportation system, this research focuses on non-recurring congestion and develops the automatic incident detection algorithm. Existent automatic incident detection algorithms may encounter one or some of the following difficulties: the detection performance is subject to the settings of algorithm threshold, traffic flow condition (medium and heavy traffic flows normally have higher detection rate) and the distance between two adjacent detectors; Most detection algorithms are not transferable in that parameters and thresholds of an algorithm must be recalibrated and revalidated to be valid for different locations or times; Quantity and quality of traffic flow data is subject to detector types. In dealing with the uncertain contexts, both neural networks and fuzzy inference have been proven as powerful tools. The FNN approaches have the advantages of learning capability to avoid subjectively setting of the parameters and possessing high fault tolerance due to the distributed memory of parameters separately stored on each link of the network. To capture the change in traffic dynamics through network training, this study presumes that the rolling-trained procedure in FNN might be imperative in augmenting the incident detection performance. Thus, the present research attempts to develop a rolling-trained fuzzy neural network (RTFNN) approach for freeway incident detection. Its underlying logic is to establish a proper fuzzy neural network and then adaptively adjust the network parameters using the most up-to-date traffic data in response to the prevailing traffic conditions so as to improve the detection performance over the conventional FNN approach. In practice, the complexity of traffic dynamics is iv.
(15) characterized with uncertain and nonlinear nature. The chaos abnormality diagnosis algorithm proposed in this paper attempts to use the change in chaotic traffic parameters, including largest Lyapunov exponent, capacity dimension, correlation dimension, relative Lz complexity, Kolmogorov entropy, delay time, and Hurst exponent to examine the existence of traffic incident. The RTFNN approach was found to have the highest potential, compared with FNN approach, to achieve a better incident detection performance. The Chaotic diagnosis approach performed the best detection rate, which covering low flow condition to heavy flow condition, but it also suffered the worst false alarm rate. However, all the tests results have indicated the feasibility of attaining the real-time automatic incident detection using the fuzzy-neural-based approaches and chaotic diagnosis approach. Furthermore, these approaches are promising and, in expectation, can be integrated into the hybrid incident detection algorithm with chaotic-based approach, which has the capability of identification, for initial diagnosis and with fuzzy-neural-based, which has the capability of classification, for confirmation of incidents in future exploration. Keywords: Incident detection, Fuzzy neural network, Fuzzy inference, Rolling-trained procedure, Chaotic diagnosis. v.
(16) CONTENT PAGES ABSTRACT (CHINESE).….…………………………………………………………...i ABSTRACT…………………………..………………………………...……………...iii ACKNOWLEDGEMENTS……………………………………………...…………….v CONTENT…………………………………………………………………...………...vi LIST OF TABLES……………………………………………………………...…….viii LIST OF FIGURES………………………………………………………...………….ix CHAPTER 1 INTRODUCTION……………………….…………………...……….1 1.1 Motivation………………………………………………………………...…………1 1.2 Objectives and Scope………………………………………………………....……...2 1.3 Framework of the Research……………………………………………….………....3 1.4 Chapters Organization………………………………..………...................................5 CHAPTER 2 LITERATURE REVIEW…………………………………...………..7 2.1 Macroscopic AID Algorithms………………………………………………………..7 2.1.1 Pattern Recognition Algorithm. ………………………………………………7 2.1.2 Statistical Approach…………………………...………………………………9 2.1.3 Catastrophe Theory…………………………………………………………..11 2.1.4 Artificial Intelligent Based Approach…………...…………………………...12 2.1.5 Fuzzy Set Theory…………………………………………………………….14 2.1.6 Chaotic Theory…………………………………...………………………….15 2.2 Microscopic AID Algorithms…………………………...………………………….16 2.2.1 Low Volume Approach…………………………...………………………….16 2.2.1 Probe Vehicle Method…………………………...…………………………..17 2.2.2 Image Processing Based Approach………………...………………………..17 2.3 Some Comments……………………………………………………………………18 CHAPTER 3 PROPOSED METHODOLOGIES………...…………………...….20 3.1 The Structure of Fuzzy Neural Network………………………...………………….23 3.1.1 Layers of Fuzzy Neural Network…………………...…...………….……….23 3.1.2 Back-propagation Training Algorithm……………...……………………….29 3.2 The Rolling-trained Procedure……………………...……...………………………32 3.3 Chaotic Diagnosis…………………………………………………………………..33 3.4 Definitions of AID Performance……………………………………………………37 CHAPTER 4 DATA FOR CALIBRATION AND VALIDATION……………..…39 4.1 Field Observation……………………...………………………………...…………39 4.2 Paramics Calibration…………………………………………………………..........41 4.3 Simulation Scenarios and Results…………………………………………………..42 CHAPTER 5 RESULTS OF FNN………………………………….……...……….46 5.1 Off-line Test Results……………………………………………………...………...46 5.2 Statistical Tests…………………………………………………………...………...50 vi.
(17) 5.3 Sensitivity of Network Structure………………………………………...…………57 5.4 Summary……………………………………………………………………………61 CHAPTER 6 RESULTS OF RTFNN………………………………………………63 6.1 Off-line Test Results………………………………...……………………………...63 6.2 Comparison with FNN……………………………………………………………...65 6.3 Sensitivity of Rolling Horizons and Training Sample Sizes……………...………..68 6.4 Summary……………………………………………………………………………71 CHAPTER 7 RESULTS OF CHAOTIC DIAGNOSIS…………………...………73 7.1 Examination for Chaos……………………………………………………...……...73 7.2 Determination of threshold for max……………………………………………….79 7.3 Off-line Test Results……………………………………………………...………...80 7.4 Comparison with RTFNN…………………………………………………………..81 7.5 Summary……………………………………………………………………………84 CHAPTER 8 CONCLUSIONS………………...………...………………………...86 8.1 General Conclusions…………………………………………………………..……86 8.2 Further Extensions….………………………………………………………………91 REFERENCES……………………………………...………………………………...93 APPENDIX…………………………………………..……………………………….100 VITA….. ……………………………………….………………….…………….100. vii.
(18) LIST OF TABLES. Table 4-1 Table 4-2 Table 5-1 Table 5-2 Table 5-3 Table 6-1 Table 6-2 Table 6-3 Table 6-4 Table 6-5 Table 7-1 Table 7-2 Table 7-3. PAGES Test for difference between inner lane and outer lane……………………..41 Test for difference between observed and Paramics simulated data…...….42 The off-line test results of the FNN incident detection………………...….48 Summary of the statistical tests………………………...………………….52 The result of sensitivity analysis……………………...…………………...58 The detection performance of RTFNN for 36 incident scenarios……...….64 Test for the difference of detection performance between RTFNN and FNN approaches…………………………………………………………………67 Detection rates for Case (I)………………………………………………...69 Detection rates for Case (II)………………………………………...……..69 Average detection rates for Case (III)……………………………………...70 Chaotic parameters for 30-second flow rate time series…………...……...75 Changes in Largest Lyapunov Exponents and Detection Performance……81 Test for the difference of detection performance between RTFNN and chaotic diagnosis approaches………………………………………………83. viii.
(19) LIST OF FIGURES. Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure. 1-1 3-1 3-2 3-3 3-4 3-5 4-1 4-2 4-3 4-4 5-1 5-2 5-3. PAGES Research procedure………………………………………………………5 The approach-based fuzzy neural network structure……….………...24 The lane-based fuzzy neural network structure……………….……...25 The trapezoid membership function……………………………………27 The rolling-trained procedure…………………………………………..33 Framework of chaotic diagnosis for incident detection…...………...34 The schematic of the incident case on Taiwan Freeway No. 1………40 Comparison of observed data and simulated data……………………...43 Observed 30-second flow rates on the studied freeway………………..44 Flow variation for one incident scenario…………………………..….45 The detection rate of different incident locations………………………49 The false alarm rate of different incident locations…………………….49 The time to detection of different incident locations…………………...49. Figure Figure Figure Figure. 5-4 The relation of detection rate and incident location……………………55 5-5 The relation of time to detection and incident location………………...56 5-6 Sensitivity analysis of FNN incident detection………………………...60 6-1 Comparison of detection performance for each simulation hour between RTFNN and FNN approaches………………………………………….66 Figure 6-2 Graph of detection rate vs. false alarm rate for 36 incident scenarios….68 Figure 6-3 Detection rates in each simulation hour for Case (I)…………………...70 Figure 6-4 Detection rates in each simulation hour for Case (II)……………….….70 Figure 6-5 Average detection rates of 36 incident scenarios for Case (III)………..71 Figure 7-1 The 30-second flow rate time series at downstream……………….…...74 Figure 7-2 Phase space reconstruction for the 30-second flow rate time series….....76 Figure 7-3 Lyapunov exponent plots…………………………...…………………….77 Figure 7-4 Capacity dimension plots…………………………………………………78 Figure 7-5 Correlation dimension plots……………………………………………....78 Figure 7-6 Delay time plots………………………………...………………………...78 Figure 7-7 Hurst exponent plots…………………………...………………………....79 Figure 7-8 Comparison of detection performance for each simulation hour between RTFNN and chaotic diagnosis approaches…………………………….…82 Figure 7-9 Graph of detection rate vs. false alarm rate between RTFNN and chaotic diagnosis approaches for 36 incident scenarios…………...……………..83. ix.
(20)
相關文件
incapable to extract any quantities from QCD, nor to tackle the most interesting physics, namely, the spontaneously chiral symmetry breaking and the color confinement..
• Formation of massive primordial stars as origin of objects in the early universe. • Supernova explosions might be visible to the most
(Another example of close harmony is the four-bar unaccompanied vocal introduction to “Paperback Writer”, a somewhat later Beatles song.) Overall, Lennon’s and McCartney’s
Microphone and 600 ohm line conduits shall be mechanically and electrically connected to receptacle boxes and electrically grounded to the audio system ground point.. Lines in
When ready to eat a bite of your bread, place the spoon on the When ready to eat a bite of your bread, place the spoon on the under plate, then use the same hand to take the
– evolve the algorithm into an end-to-end system for ball detection and tracking of broadcast tennis video g. – analyze the tactics of players and winning-patterns, and hence
The 3SEQ maximum descent statistic describes clus tering patterns in sequences of binary outcomes, a nd is therefore not confined to recombination analy sis... New Applications (1)
A multi-objective genetic algorithm is proposed to solve 3D differentiated WSN deployment problems with the objectives of the coverage of sensors, satisfaction of detection
