臺灣博碩士論文加值系統

在台灣每天平均都有數百起車禍事故案件，車禍的發生通常來環境因素以及雙方車輛駕駛的人為因素，交通事故通常都會牽扯到金額賠償的問題，因此在做肇事鑑定的時候必須將雙方的責任歸屬說明得非常明確，其中分成五類：全部肇事原因、肇事主要因素、肇事次要因素、同為肇事原因、無肇事因素。本研究使用徑向基類神經網路作為系統基本架構，使用事故實際案例資料作為資料庫建立來源，並依照鑑定委員鑑定案件時所參考的因素選定環境因素以及車輛因素來輸入系統。實驗分成四階段來實行，並記錄訓練和預測的時間。額外設計出一套人機介面讓使用者能更方便使用以及推廣，在程式發展上也容易取得使用者的意見。

In Taiwan, there are hundreds of accidents every day recorded by government due to the human factor and environmental factor. The accident usually involved the money dispute; therefore the accident appraisal must indicate the bilateral parties’ blame clearly: all blame; major blame; minor blame and none blame. This study employed Radial Basis Function Neural Networks to build an expert system for appraisal of bilateral vehicle accident. The database was built from 489 accident cases in Taiwan from the year of 2004 to 2008. According to Committee’s analysis, there are 30 appraisal basses including environmental basses and vehicle basses chosen to be the input of the expert system. The training stage was carried out by four types. Validation stage was carried out by using 100 fixed cases and the correctness was recorded. The training and validation processes were completed in one second. With the design of the graphical user interfaces, the system could be easier to use and circulate. The more users’ reflections and problem response, the more system’s disadvantage could be improved.

ABSTRACT (CHINESE) i
ABSTRACT (ENGLISH) ii
ACKNOWLEDGEMENT iii.
CONTENTS iv
LIST OF FIGURES vi.
LIST OF TABLES viii.
LIST OF SYMBOLS ix

CHAPTER 1
INTRODUCTION 1

CHAPTER 2
APPRAISAL PROCESS 4
2-1 Relative Authority 5
2-2 Blame Level 10
2-3 Intersection 12
2-4 Violation Weightings 14
2-5 Relative Position of Bilateral 16
2-6 Position of Collision 21
2-7 Other Basses 27
2-8 Difficulty of Accident Analysis 28
2-9 Appraisal Basses 30

CHAPTER 3
THEORY
3-1 Neural Networks 32
3-2 Radial Basis Function Neural Networks 35
3-3 Other Training Methods for RBFNNs 41

CHAPTER 4
EXPERIMENTS AND RESULTS 44

CHAPTER 5
GRAPHICAL USER INTERFACES
5-1 Process of GUI 47
5-2 Designing a Graphical User Interfaces 49
5-3 Ways to Build MATLAB GUIs 50
5-4 Graphical User Interface Development Environment 53
5-5 The Interface of Appraisal System 56
5-6 The Process of Appraisal System 57

CHAPTER 6
CONCLUSIONS 58

REFERENCES 60


Figure. 2.1 The “yield” sign and flash red signal. 6
Figure. 2.2 The special marking. 6
Figure. 2.3 The number of lanes is two and three. 7
Figure. 2.4 The turn-way car does not yield the straight-way car. 8
Figure. 2.5 The red car is left side, the blue car is right side. 9
Figure. 2.6 The bike-forbidden lanes. 15
Figure. 2.7 Parties on the same way in the different lane. 18
Figure. 2.8 Parties on the same way in the same lane. 18
Figure. 2.9 Parties in opposite direction. 19
Figure. 2.10 The blue car exceeded its lane. 19
Figure. 2.11 Parties in bevel position. 20
Figure. 2.12 Eight parts of collision. 24
Figure. 2.13 Behind car crashed to rear part. 25
Figure. 2.14 Behind car crashed to waist part. 25
Figure. 2.15 Bike riders are dangerous in the outside lanes. 26
Figure. 3.1 Simplified view of feed forward artificial neural networks. 34
Figure. 3.2 The biological neurons. 34
Figure. 3.3 The basic structure of RBF neural networks. 37
Figure. 3.4 The hidden neuron in the hidden layer. 37
Figure. 3.5 The process of basses, neural network and result. 43
Figure. 4.1 Correctness test result. 46
Figure. 5.1 An empty GUIDE edit box. 54
Figure. 5.2 Push button. 55
Figure. 5.3 Pop-up menu. 55
Figure. 5.4 Static text. 55
Figure. 5.5 The interface of appraisal system. 56

Table 2.1 The classification of blames. 11
Table 4.1 The correctness rate. 46
Table 5.1 Possibilities in GUI. 52

[1] Chiou, Y.C. (2006). An artificial neural network-based expert system for the appraisal of two-car crash accidents. Accident Analysis and Prevention, 38, 777-785.
[2] Al-Ghamdi, A.S. (2002). Using logistic regression to estimate the influence of accident. Accident Analysis and Prevention, 34(6), 729-741.
[3] Greibe, P. (2003). Accident prediction models for urban roads. Accident Analysis and Prevention, 35(2), 273–285.
[4] Abbas, K.A. (2004). Traffic safety assessment and development of predictive models for accidents on rural roads in Egypt. Accident Analysis and Prevention, 36(2), 149-163.
[5] Ivan, J.N. (1997). Neural network representations for arterial street incident detection data fusion. Transportation Research, 3, 245-254.
[6] Hasnemi, R.R., Le Blanc, L.A., Rucks, C.T. and Shearry, A. (1995). A neural network for transportation safety modeling”, Expert Systems with Applications, 9(3), 247-256.
[7] Mussone, L., Ferrari, A. and Oneta, M. (1999). An analysis of urban collisions using an artificial intelligence model. Accident Analysis and Prevention, 31(6), 705-718.
[8] Hopfield, J.J. (1982). Neural networks and physical systems with emergent collective computational abilities. Biophysics, 79, 2554-2558.
[9] MathWorks (2011). User's Guide of MATLAB R2011a Documentation
[10] Aizerman, M.A., Braverman, E.M. and Rozonoer, L.I. (1964). Theoretical foundations of the potential function method in pattern recognition learning, Autom. Remote Control, 25, 821–837.
[11] Bashkirov, O.A., Braverman, E.M. and Muchnik, I.B., (1964). Potential function algorithms for pattern recognition learning machines, Autom. Remote Control, 25, 629–631.
[12] Sprecht, D.F. (1968). A Practical Technique for Estimating General Regression Surfaces, Report LMSC-6-79-68-6, Lockheed Missile and Space Co., Inc., Palo Alto, CA.
[13] Broomhead, D.S. and Lowe, D., (1988). Multivariable functional interpolation and adaptive networks, Complex Systems 2, 321–355.
[14] Moody, J.E. and Darken, C.J. (1989). Fast learning in networks of locally-tuned processing units, Neural Computing. 1 (2), 281–294.
[15] Mak, M.W., Allen, W.G. and Sexton, G.G., (1994). Speaker identification using multilayer perceptrons and radial basis function networks, Neural Computing, 6, 99–117.
[16] Bianchini, M., Frasconi, P. and Gori, M. (1995). Learning without local minima in radial basis function networks. IEEE Trans Neural Networks 6, 749–56.
[17] Park, J. and Sandberg, I.W. (1991) Universal approximation using radial basis functions network. Neural Computing, 3, 246–57.
[18] Zhang, A. and Zhang, L. (2004). RBF neural networks for the prediction of building interference effects. Computers&Structures, 82, 2333-2339
[19] Looney, Carl G. (2002). Radial basis functional link nets and fuzzy reasoning. Neural Computing, 48, 489-509.
[20] Leonard, J.A. and Kramer, M.A. (1991). Radial basis function networks for classifying process faults. IEEE Control Systems, 11, 31–8.
[21] Weiss, S.M. and Kulikowski, C.A. (1991). Computer systems that learn. San Mateo, CA, Morgan Kaufmann Publishers Inc.
[22] Looney, C. (1996). Demystifying feedforward neural networks as knowledge acquiring black boxes, IEEE Trans. Knowledge Data Eng. 8 (2), 211–226.
[23] Looney, C. (1997). Pattern Recognition Using Neural Networks, Oxford University Press.