利用自我映射組織圖與增強式學習法於可調式交通號誌控制系統的機器學習模型

簡易檢索 / 詳目顯示

回結果列表

研究生：	高映慈 Kao, Ying-Cih
論文名稱：	利用自我映射組織圖與增強式學習法於可調式交通號誌控制系統的機器學習模型 A Machine Learning Model for Adaptive Traffic Light Control System with Self-Organizing Map and Reinforcement Learning
指導教授：	吳誠文 Wu, Cheng-Wen
口試委員:	李昆忠 Lee, Kuen-Jong 謝明得 Shieh, Ming-Der 黃稚存 Huang, Chih-Tsun
學位類別：	碩士 Master
系所名稱：
論文出版年：	2017
畢業學年度：	106
語文別：	英文
論文頁數：	51
中文關鍵詞：	機器學習、自我映射組織圖、增強式學習、交通燈號控制
外文關鍵詞：	machine learning, self-organizing map, reinforcement learning, traffic light control
相關次數：	點閱：2 下載：0
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

大城市中的交通壅塞問題日益嚴重，導致人們耗費更長的等待時間、更嚴重的空氣汙染、更多的汽油消耗，而交通壅塞的成本也因此提高。由於交通壅塞成本主要與旅程等待時間有關，所以提出了許多可調式交通控制(ATC)系統，用來減少旅行等待時間。然而，在過去的方法中，需要大量的感測器和硬體設備，因此很難部署在大區域中。幸好，先進物聯網(IoT)技術的出現使得我們有可能找到更有效且有效率的解決方案。此外，隨著人工智慧(AI)平台的蓬勃發展，尤其是機器學習(ML)和人工神經網路(ANN)模型的提出，使得我們有望利用機器學習或人工神經網路模型搭配物聯網技術來改進可調式交通控制系統。
在本篇論文中，我們提出了一個可調式交通號誌控制系統的機器學習模型。首先，我們假設可以從在道路上的車子上的物聯網感測裝置收集交通數據。接著，我們提出的機器學習模型會接收在雲端分析好的數據並產生一個最佳的交通號誌週期作為輸出。最後，將交通號誌週期轉換為交通號誌設定信號，傳送到物聯網致動裝置上，而交通號誌控制器即是物聯網致動裝置。為了驗證我們所提出的機器學習模型，我們建立了一個交通模擬器，而模擬器中的交通流量模型是基於台北市主要路口的高峰期交通流量開放數據集來建立。在二十四小時的模擬期間中，與固定式交通號誌控制系統相比，我們所提出的模型平均可以減少55.7%的等待時間和12.76%的最大道路佔有率。我們也針對不同交通流量進行模擬，我們所提出的模型在等待時間和最大道路佔有率方面的表現始終優於固定式交通號誌控制系統。等待時間和最大道路佔有率是交通壅塞問題的兩大因素，在我們的模擬中，這兩個指標的結果顯示我們所提出的模型可以減輕交通壅塞問題。

Increasing urban congestion problems in big cities lead to more delay time, more air pollution, and more gasoline consumption, so the congestion cost keeps rising. Because the congestion cost is mainly related to travel delay time, some Adaptive traffic control (ATC) systems have been proposed to reduce the travel delay time. However, previous solutions need a great number of sensors and hardware devices, so it is hard to deploy in a large area. Fortunately, the advent of advanced Internet of Things (IoT) technologies has made possible more effective and efficient solutions of the congestion issue. Besides, with the availability of the booming Artificial Intelligence (AI) platforms, particularly Machine Learning (ML) and Artificial Neural Network (ANN) models, there is hope that the ATC system can be improved with the IoT approach, adopting the ML/ANN models.
In this thesis, we propose a machine learning model for adaptive traffic light control system. First, we assume traffic data is collected from internet-connected IoT sensing devices in vehicles on road. Next, the proposed machine learning model receives the data analyzed in cloud and generates an optimal traffic light period as output. Finally, the optimal traffic light period is transformed to traffic light setting signals to be delivered to the IoT actuating devices, i.e., the traffic light controllers. For verifying the proposed model, we build a traffic simulator, and the traffic model in our simulator is based on an open data set, which is about traffic flow at peak period on main road intersections in Taipei. For a 24-hour simulated period, the proposed model reduces 55.7% waiting time and 12.76% maximum road occupancy on average as compared with the fixed traffic light control system. We also simulate different traffic levels, and our model performs consistently better than the fixed traffic light controller in the overall waiting time and maximum road occupancy. Waiting time and maximum road occupancy are two main factors of traffic congestions, and the results by these two metrics in our simulation show that the proposed model is able to alleviate the traffic congestion problem.

摘要 i
Abstract ii
Contents iv
List of Figures    vi
List of Tables vii
Chapter 1 Introduction 1
1 Motivation 1
2 Related Works 3
3 Proposed Approach and Results 4
4 Organization 5
Chapter 2 Proposed Machine Learning Model 6
1 Architecture of Proposed Machine Learning Model 6
2 Introduction to Self-Organizing Map (SOM) 8
3 Details of the SOM Model 10
4 An Example of RL: Q-Learning 12
5 Details of the RRL Model 14
Chapter 3 Traffic Simulator 16
1 Overview 16
2 Traffic Data 17
2.1 Data Pre-processing 17
2.2 Distribution of Traffic Data 19
3 Models 22
3.1 Urban Traffic Map    22
3.2 Traffic Light Controller 26
3.3 Traffic Generator    26
3.4 Cloud 28
Chapter 4 Experimental Results 32
1 Convergence of RRL 32
1.1 Updating Function    32
1.2 Learning Function 32
1.3 Observations 33
2 Reward Function of RRL model 36
3 Size of SOM 37
4 Verification 38
4.1 Random Road Length 39
4.2 24-hour Simulated Period 40
4.3 Different Traffic Level 40
Chapter 5 Conclusions and Future Work 42
1 Conclusions 42
2 Future Work    43
Bibliography 45
Appendix 47


                                

[1] D. Schrank, B. Eisele, T. Lomax, and J. Bak, “2015 Urban Mobility Scorecard,” Published jointly by The Texas A&M Transportation Institute and INRIX, Aug. 2015.
[2] R. Louf and M. Barthelemy, “How congestion shapes cities: from mobility patterns to scaling,” Scientific Reports 4, no. 5561, July 2014.
[3] A. G. Sims and K. W. Dobinson, “The Sydney coordinated adaptive traffic (SCAT) system philosophy and benefits,” IEEE Transactions on Vehicular Technology, vol. 29, issue 2, May 1980, pp. 130-137.
[4] D. I. Robertson and R. D. Bretherton, “Optimizing networks of traffic signals in real time-the SCOOT method,” IEEE Transactions on Vehicular Technology, vol. 40, issue 1, Feb. 1991, pp. 11-15.
[5] D. Evans, “The Internet of Things-How the Next Evolution of the Internet Is Changing Everything,” Cisco White Paper, Apr. 2011.
[6] A. Castillo and A. Thierer, “Projecting the Growth and Economic Impact of the Internet of Things,” Social Science Research Network (SSRN), Jun. 2015, pp. 2618794.
[7] M. B.Trabia, M. S.Kaseko, and M. Ande, “A two-stage fuzzy logic controller for traffic signals,” Transportation Research Part C: Emerging Technologies, vol. 7, issue 6, Dec. 1999, pp. 353-367.
[8] C. P. Pappis and E. H. Mamdani, “A Fuzzy Logic Controller for a Traffic Junction,” IEEE Transactions on Systems, Man, and Cybernetics, vol. 7, issue 10, Oct. 1977, pp.707-717.
[9] S. Araghi, A. Khosravi, and D. Creighton, “Optimal Design of Traffic Signal Controller Using Neural Networks and Fuzzy Logic Systems,” 2014 Int. Joint Conf. on Neural Networks (IJCNN), July 2014, pp.42-47.
[10] M. B. W. de Oliveira and A. de Almeida Neto, “Optimization of Traffic Lights Timing Based on Multiple Neural Networks,” 2013 Tools with Artificial Intelligence (ICTAI), Nov. 2013, pp. 825-832.
[11] M. Saito and J. Fan, “Artificial Neural Network–Based Heuristic Optimal Traffic Signal Timing,” Computer-Aided Civil and Infrastructure Engineering, vol. 15, issue 4, July 2000, pp. 293-307.
[12] M. A. Khamis and W. Gomaa, “Enhanced multiagent multi-objective reinforcement learning for urban traffic light control,” 2012 11th International Conference on Machine Learning and Applications (ICMLA), Dec. 2012, pp. 586-591.
[13] B. Bakker, S. Whiteson, L. Kester, and F. Groen, “Traffic light control by multi-agent reinforcement learning systems,” Interactive Collaborative Information Systems, 2010, pp. 475-510.
[14] Taipei City Traffic Engineering Office, “2016 Investigation on Traffic Flow and Characteristics in Taipei City.” (in Chinese)
[15] T. Kohonen, “The self-organizing map,” in Proc. of the IEEE, vol. 78, issue 9, Sep. 1990, pp. 1464-1480.
[16] U. Fidan, N. Ozkan and I. Calikusu, “Clustering and classification of dermatologic data with Self Organization Map (SOM) method,” 2016 Medical Technologies National Congress (TIPTEKNO), Oct. 2016, pp. 1-4.
[17] T. Kohonen, K. Lagus, J. Salojärvi, J. Honkela, V. Paatero, and A. Saarela, “Self Organization of a Massive Document Collection,” Transactions on Neural Networks, vol. 11, no.3, May 2000, pp.574-585.
[18] C. J. C. H. Watkins and P. Dayan, “Q-learning,” Machine Learning, vol.8, issue 3-4, May 1992, pp. 279-292.
[19] L. A. Hoel, N. J. Garber, and A. W. Sadek, “Transportation Infrastructure Engineering: A Multimodal Integration, SI Version,” pp. 587.
[20] M. L. Delignette-Muller and C. Dutang, “Package ‘fitdistrplus’,” from https://cran.rproject.org/web/packages/fitdistrplus/index.html.
[21] H. Akaike, "A new look at the statistical model identification," IEEE Transactions on Automatic Control, vol. 19, no. 6, Dec 1974, pp. 716-723.
[22] “Traffic Signs, Markings and Lights Installation Rules,” Laws & Regulations Database of The Republic of China, Jun. 2017. (in Chinese)

簡易檢索 / 詳目顯示

相關論文