隨著科技不斷的進步，智慧城市(Smart City)已經成為世界各大城市發展目標的其中之一，其發展的目標有無人駕駛車、車輛通訊等，對於車載無線通訊的使用，能讓車輛之間和車輛與路邊設備利用車載網路 (VANET)進行訊息的交換，讓車輛可以在各種交通狀況即時的應變，減少車輛的旅行時間，提升整體交通的運輸效率。傳統交通號誌在交通運輸系統下，對於車流量的管理有很大的限制，對於交通事件的發生，無法即時的反應和處理；並且在佈置跟營運的耗費，也提高了管理車流量所需的成本。因此本論文基於車載網路中的車間通訊 (Vehicle-to-Vehicle)，讓車輛互相交換信息，以利用虛擬交通號誌 (Virtual Traffic Light)進行車輛的管理，改善車流控制問題與管理的成本。綜合以上敘述，本論文提出車輛路徑決策虛擬交通號誌機制(Re-routing Decision Virtual Traffic Light, RDVTL)，考慮車輛道路行駛時的道路狀況，以及車輛種類不同的行為，以及加入車輛消耗能源的使用效率的考量，將車輛進行分流，解決交通壅塞問題。透過將道路中的車流量負載平衡的成果，增進道路的使用效率，並且提升車流的順暢度，透過車輛上的顯示裝置，除了告知駕駛路口號誌之外，另外提供車輛行駛路線的導航指引。最後，利用模擬實驗驗證機制的運行成果，在實驗模擬的結果顯示，本論文所提出之RDVTL機制相較於傳統交通號誌的車流控制方法，可以提升整體交通的行車速度57%和增加42%的路口運輸吞吐量，另外減少42%的行車旅行時間。;With the rapid development of novel technologies, Smart city has become one of the development goals in worldwide cities. The goal covers self-driving cars, vehicle communications, etc. Vehicular Ad Hoc Network (VANET) is used to connect vehicles so that vehicles can communicate with each other or even with the roadside unit. In addition, vehicles are able to simultaneously react any situation and to reduce vehicle travel time for achieving the efficiency of transportation. There are many limitations in managing traffic flows based on the control method of traditional traffic lights. The traditional traffic light control method not only wastes management and deployment costs but also tackles traffic accidents immediately. Therefore, this thesis aims to improve traffic control and to reduce management cost by utilizing Virtual Traffic Light based on the Vehicle-to-Vehicle communication in VANETs. In conclusion, a Re-routing Decision Virtual Traffic Light (RDVTL) mechanism is proposed in the thesis. In order to disperse traffic and resolve the problem of traffic congestion, the RDVTL not only takes road condition and different types of vehicles’ behavior into consideration but also the energy consumption of vehicles. Based on the proposed traffic load balancing in RDVTL, transportation system can be enhanced on road utilization and traffic smoothness. Moreover, the display device on vehicles is able to inform traffic light and to provide navigation for vehicles. The experimental results show that the proposed RDVTL mechanism improves 57.52% average velocity, increases 42.79% transportation throughput, and reduces 54.42% travel time compared with the traditional traffic signal.
