近年來緣於 ITS 在國內外之蓬勃發展，對於都市交通號誌控制系統而言，未來將可從道路上密佈的車輛偵測器以及車內資訊系統之定位回報等多重途徑，獲得極為充分的道路交通資訊；因此能夠逐秒逐車的分析計算，同時針對意外事件作出反應之適應性號誌控制系統，將會因民眾無止盡的對交通績效不斷提昇之需求，而逐漸取代傳統的定時及動態控制方式，而成為普及化的交通號誌控制系統。

　　然而回顧目前國內外所發展的適應性號誌控制邏輯，如：OPAC、MOVA、SCOOT、SCAT、RealBand、COMDYCS-III 等，皆是以個別路口為基礎來進行最佳化；或是設定過多的假設條件，來簡化多路口的適應性決策問題。其中 OPAC 係獨立路口之最佳化決策；而SCOOT 多使用於幹道走廊之最佳化，其績效較採路網控制方式為佳；又RealBand 係限用於車隊較易辨識的交通狀況下，因此當交通出現壅塞，以致於難以辨識出車隊時，就無法獲得較佳的控制績效；至於 COMDYCS-III 則係改良自 SAST 模式，而以獨立路口作為控制主體；後期雖然引進遺傳演繹法來發展網路控制邏輯，但還是脫離不了以個別路口為單位的加總運算方式。

　　另一方面，上述這些控制邏輯中與系統績效關係密切之部分控制參數，皆須經由事前的交通調查加以計算，而無法因應控制區域內旅次型態之改變而即時加以調整，以致造成控制決策對現場交通狀況的掌握偏離實際情況，如此反有可能使得系統績效落後於其他的控制策略。
　　
　　綜上所述，本研究乃嘗試應用類神經網路之未知函數形態問題求解特性，以使號誌系統具有自我調適(Self-Adaptive)、學習(Learning)、與容錯(Fault Tolerance)等能力，裨得以進行適應性號誌控制模式之性能革命。至於本研究所構建之模式則係基於路網全域最佳化的觀點，以線上自動學習所控制區域之交通特性的方式，並經由偵測器的量測結果為依據，來尋求控制績效的最佳化。本研究之成果將可處理目前適應性控制邏輯的多數發展瓶頸，因此具發展及推廣之潛力。

　　Due to the rapid development of ITS in the world recently, for traffic signal control system in the urban cities, it is possible to gain sufficient road traffic information from multiple approaches, such as vehicle detectors on the roads and Global Position with In-Vehicle Information Systems. This can lead to analytical calculation of per vehicle per second basis. At the same time, traffic adaptive signal control system can react to accidents and thus satisfy the demand on constantly enhancing the traffic excellence among the general public. It will also replace the traditional ways of fixed hour and dynamic control and become the popular traffic signal control system.
　　However, if looking back on the logic development of traffic adaptive signal control system in the domestic and worldwide at present, such as OPAC、MOVA、SCOOT、SCAT、RealBand、COMDYCS-III, they are designed to conduct optimal on the basis of individual intersection. Or, they are set to have too many conditions to simplify the problems of adaptive strategies of multi-intersections. Among them, OPAC is the best optimal strategy of independent intersections while SCOOT is most used for optimal of arterial corridors and its performance is best presented by network control. As for RealBand, it is limited only under the traffic condition which is easier to identify. Therefore, when there is a traffic jam and hard to identify cars, it is difficult to gain better control. COMDYCS-III is modified from SAST model and uses independent intersections as main control body. Later, although Genetic Algorithms is imported to develop the logic of network control, it can not exclude itself from the mathematical calculation of using individual intersections as units.
　　On the other hand, some controlling reference of these control logic and system performance have to be calculated before traffic investigations. They cannot be made any adjustments according to the change of travel patterns in the control areas. This would cause an unreal situation when the control strategy understands the traffic situation on spot and also possibly makes the system performance fall beyond the other control strategies.
　　According to the discussion above, this research aims to apply the theory of applied solve unknown function of Neural Network to make the traffic system possess abilities of self-adaptive, learning and fault tolerance, in an attempt to revolve the function of traffic adaptive signal control system. As for the model adopted in this research, it is based on the viewpoint of global optimal and characteristics of traffic of online self-learning control area. Based on the results of detectors, it is hoped to look for the optimal of control performance. The outcome of this research can solve the bottlenecks of development of traffic adaptive signal control system at present, and therefore, has potential for further development and promotion.

1. 王勝石，影像處理與類神經綱路整合模式應用於車種分類之研究，國立成功大學交通管理科學研究所碩士論文，民國84年。

2. 田川昇，利用基因演算法與類神經網路建立台灣西南海域深部地層滲透率模式之研，國立成功大學資源工程學系碩士論文，民國91年。

3. 何志宏，臺灣地區先進交通管理系統 (ATMS) 中都市交通號誌控制邏輯標準化與系統建置標準作業程序之研究，交通部運輸研究所，民國89年。

4. 何志宏等，先進式微觀車流模擬器—PARAMICS模擬軟體應用於高速公路與市區道路交控系統整合策略研究，交通部運輸研究所，民國90年7月。

5. 何志宏等，先進式微觀車流模擬器—PARAMICS應用於台灣地區發展ITS模擬網路之模式校估測試研究，交通部運輸研究所，民國88年。

6. 何志宏等，最新全動態交通號誌控制技術開發計畫，國立成功大學交通管理科學研究所，民國八十年二月。

7. 吳悅慈，都會區內高速公路走廊交通疏導改道策略之構建與模擬研究—以圓山與台北交流道間之高速公路走廊為例，國立成功大學交通管理科學研究所碩士論文，民國91年。

8. 吳耿毓，類神經網路應用於自動化高速公路匝道儀控之研究，國立成功大學交通管理科學研究所碩士論文，民國85年。

9. 呂政龍，新的幹道連鎖號誌模式之研究—結合公車綠燈帶寬的概念，私立淡江大學運輸管理研究所碩士論文，民國89年。

10. 李樑堅，建立微觀車流模擬模式以發展交通適應性號誌控制邏輯之研究，成大交研所博士論文，民國八十二年六月。

11. 李穎，類神經網路應用於國道客運班車旅行時間預測模式之研究，國立成功大學交通管理科學研究所碩士論文，民國91年。

12. 沈毓泰, 類神經網路在非線性系統控制之應用, 南台工商專校學報 22期 頁199-209, 1996

13. 林士傑，高速公路旅行時間預測模式之研究，國立成功大學交通管理科學研究所碩士論文，民國90年。

14. 林育瑞，利用類神經網路構建機車車流模式之研究，國立成功大學交通管理科學研究所碩士論文，民國91年。

15. 邱顯鳴，結合車道變換率之事件偵測新演算法研究，台灣大學土木工程研究所碩士論文，民國84年。

16. 柯仁傑，C++程式結構化測試方法，資訊與電腦，1998年1月，pp70~p73

17. 張明惠，四種現代化適應性號誌控制邏輯(OPAC、MOVA、SAST、COMDYCS-Ⅲ)之比較研究，成大交研所碩士論文，民國八十二年六月。

18. 張堂賢，電腦化交通號誌控制器進階功能之研發策略擬訂，交通部科技顧問室，民國91年。

19. 張鈞萍，發展緊急救援車輛之行車路徑導引系統與號誌優先通行控制邏輯之研究，國立成功大學交通管理科學研究所碩士論文，民國91年。

20. 許智淵，基因演算法於類神經網路之應用，國立成功大學航空太空工程研究所碩士論文，2000年

21. 郭哲肇，利用人工智慧發展電梯動態控制系統之研究，成功大學交通管理科學研究所碩士論文，民國83年。

22. 陳奇銘，動態競爭及二分類神經網路，國立成功大學電機工程研究所博士論文，1996

23. 曾國雄、邱裕鈞、許書耕，主線柵欄是收費站最佳區位遺傳演算尋優法與逐步尋優法之比較分析，中國土木水利工程學刊，第九卷第一期，民國八十六年，pp171~178。

24. 焦李成，神經網路系統理論，儒林出版社，1991年11月初版

25. 黃威龍，前饋式類神經網路軟體於系統識別與控制之研究，國立成功大學航空太空工程學系博士論文，民國91年。

26. 黃泰林，構建智慧型適應性網路號誌控制模式之研究，成大交研所博士論文，民國八十三年六月。

27. 楊正甫，物件導向分析與設計，台北市，松岡 2000。

28. 楊智能，都市幹道號誌化交通流佔量指標之研究，逢甲大學建築及都市計畫研究所碩士論文，民國90年。

29. 葉怡成，類神經網路-模式應用與實作，儒林出版社，1995年

30. 廖啟業，應用類神經網路及模糊理論於自學式控制器之研究 = Apply neural network and fuzzy theorem to self-learning controller design，國立成功大學工程科學研究所碩士論文，民國83年。

31. 蔣封文，應用車隊擴散理論於構建網路型適應性號誌控制模式之研究，國立成功大學交通管理科學研究所碩士論文，民國89年。

32. 鄭博王，以遺傳演算法及類神經網路應用於旅行推銷員問題上，淡江大學運輸科學研究所碩士論文，民國86年。

33. 謝欣宏，台鐵司機員排班與輪班問題之研究－以基因演算法求解，國立成功大學交通管理科學研究所碩士論文，民國91年。

34. 魏健宏、黃國平、陳昭宏，應用人工神經網路發展高速公路意外事件自動偵測模式，運輸計劃季刊第25卷第2期，209- 234頁，民國85年。

35. 魏健宏、楊雨青，高雄港轉口貨櫃運量預測-以類神經網路評選輸入變數，運輸學刊第11卷第3期，1-20，民國88年。

36. 魏健宏、楊雨青，智慧型運輸系統交通參數資料融合方法之研究-應用類神經網路，第一屆台灣ITS國際研討會論文集，B1-26-B1-41頁，民國88年。

37. 蘇木春, 張孝德，機器學習 :類神經網路、模糊系統以及基因演算法則，全華科技，1995年

38. A. Blanco, M. Delgado, M. C. Pegalajar, A genetic algorithm to obtain the optimal recurrent neural network, International Journal of Approximate Reasoning (2000), pp67-83

39. Baass-KG, Lefebvre-S，〝Analysis Of Platoon Dispersion With Respect To Traffic Volume.〞， Transportation Research Record No. 1194，1988，pp64-76.

40. Barada, S; Singh, H, Generating Optimal Adaptive Fuzzy-Neural Models of Dynamical Systems with Applications to Control, IEEE Transactions on Systems Man and Cybernetics - Part C -Applications and Reviews vol: 28 iss: 3, 1998

41. Bretherton, D (1996), ‘Current developments in SCOOT: version 3’, Transportation and Research Record 1554, pp 48-52.

42. Christina M. Andrews,S.Manzur Elahi,James E.Clark，〝Evalution of New Jersey Route 18 OPAC/MIST Traffic-Control System〞，Transportation Research Record No. 1603，1997，pp150-155.

43. D. E. Goldberg, Genetic Algorithm in Search, Optimization and Machine Learning, Addison Wesley, New York, 1989.

44. D.F. Cook, C.T. Ragsdale, R.L. Major, Combineing a neural network with a genetic algorithm for process parameter optimization, Engineering Applications of Artifical Intelligence 13, 2000, pp391-396

45. Daniel P. Freedman and Gerald M. Weinberg , Handbook of Walkthroughs, Inspections, and Technical Reviews,。

46. David Bretherton, Keith Wood, Neil Raha，〝Traffic Monitoring and Congestion Management in the SCOOT Urban Traffic Control System.〞，Transportation Research Record No. 1634，1998，pp118-124.

47. Ding, Y., Chien, S. I., and Wei, C. H., “Dynamic Transit Arrival Time Prediction Using Link-Based and Stop-Based Artificial Neural Networks (Revised Version),” Submit to Transportation Research, October 2000.

48. Douglas C., Montgomery, Design and analysis of experiments Fourth Edition, WILEY, 1996

49. Douglas C. Montgomery, Designed and analysis of experiments 4/E, John Wiley & Sons, 1997.

50. Eil Kown,Yorgos J.Stephanedes，〝Develop of an Adaptive Control Strategy in a Live Intersection Laboratory〞，Transportation Research Record No. 1634，1998，pp123-129.

51. Fred Glover, James P. Kelly, Manuel Laguna, Genetic algorithms and tabu search: hybrid for optimization, Computers Ops Res. Vol. 22 No.1 (1995 ), pp111-134

52. Gen, and R. Cheng, Genetic algorithms and engineering optimization, John Wiley & Sons, 2000.

53. Gupta, Jatinder and Sexton, Randall, Comparing backpropagation with a genetic algorithm for neural network training, Omega, 27(1999) ,pp679-684.

54. Haupt, L. Randy, and S. E. Haupt, Practical genetic algorithms, Wiley, New York, 1998.

55. Holland, J. H., Adaptation in Nature and Artifical System, University of Michigan press, Ann Arbor, 1975.

56. Ivan, J. N. et al., “Real-Time Data Fusion for Arterial Street Incident Detection Using Neural Networks,” Transportation Research Record, No. 1497, pp. 27-35, 1995.

57. James C. Spall; Daniel C. Chin, Traffic-Responsive Signal Timing for System-Wide Traffic Control,, Transportation Research, Part C. 1997/06/08. 5C(3/4) pp153-63

58. Jiuyi Hua; Ardeshir Faghri, Development of Neural Signal Control System--Toward Intelligent Traffic Signal Control., Transportation Research Record. 1995. (1497) pp53-61

59. Manar-A, Baass-KG，〝Traffic Platoon Dispersion Modeling On Arterial Streets.〞，Transportation Research Record No. 1566，1996，pp49-53.

60. Nathan H. Gartner, Philip J. Tarnoff, and Christina M. Andrews，〝Evaluation of Optimized Policies for Adaptive Control Strategy〞，Transportation Research Record No.1324，1991，pp105-114.

61. Palacharla, P. V. and Nelson, P. C., “On-Line Travel Time Estimation using Fuzzy Neural Network,” 2nd ITS World Congress, 1995.

62. Parisini, T; Zoppoli, R, Neural Approximations for Multistage Optimal Control of Nonlinear Stochastic Systems, IEEE Transactions on Automatic Control vol: 41 iss: 6 Order: LIB, IEEE Institute of Electrical and Electronics, 1996

63. Park, D. et al., “Spectral Basis Neural Networks For Real-Time Travel Forecasting,” Journal of Transportation Engineering, pp. 515-523, Nov/Dec, 1999.

64. R.D. Bertherton, K.Wood and G.T.Bowen，〝SCOOT Version 4〞，Transportation Engineeing & Control ，39(7)，1998，pp425-427.

65. Randall S. Sexton, Robert E. Dorsey, John D. Johnson, Optimization of Neural Networks - A Comparative-Analysis of the Genetic Algorithm and Simulated Annealing, European Journal of Operational Research 1999, Vol 114, Iss 3, pp 589-601

66. Randall S. Sexton, Robert E. Dorsey, John D. Johnson, Toward global optimization of neural networks: A comparison of the genetic algorithm and backpropagation, Decision Support Systems 22 (1998), pp171-185

67. Rumbaugh, Blaha, Premerlani, Eddy, Lorensen. "Object-Oriented Modeling And Design ". N.J.: Prentice Hall, 1991

68. S.Manzur Elahi, A. Essan Radwan, and K. Michael Goul,〝Knowledge-Based System for Adaptive Traffic Signal Control〞, Transportation Research Record No.1324, 1991, pp115-122.

69. Samerkae Somhom, Abdolhamid Modares, Takao Enkawa, Competition-Based Neural-Network for the Multiple Traveling Salesmen Problem with Minmax Objective, Computers & Operations Research 1999, Vol 26, Iss 4, pp 395-407

70. Shivakumar Vaithyanathan, Laura I. Burke, Michael A. Magent, Massively parallel analog tabu search using neural networks applied to simple plant location problems, European Journal of Operational Research 93 (1996) 317-330

71. Song, Q; Leland, R P, An Optimal Control Model of Neural Networks for Constrained Optimization Problems, Optimal Control Applications and Methods vol: 19 iss: 5 Order: LIB, 1998

72. Stephanedes,Y.J. and Liu,X. "Neural Network in freeway control", proceedings of Pacific Rim TransTech Conference, 1993.7.

73. T.T. Chow, G.Q. Zhang, Z. Lin C.L. Song, Global optimization of absorption chiller system by genetic algorithm and neural network, Energy and Building 34, 2002, pp103-109

74. Z. Michalewicz, Genetic algorithms + data structures = evolution programs, 3rd rev. and extended ed., Springer-Verlag, 1996.

75. Zhang, H., Ritchie, S. G., and Lo, Z., “Macroscopic Modeling of Freeway Traffic Using an Artificial Neural Network,” Transportation Research Record, No. 1588, 110-119, 1997.