運用車載通訊最佳化車輛吞吐量與號誌公平性之十字路口動態交通控制

Abstract

在現代的城市中，壅塞的交通嚴重地影響我們生活品質與環境。沒有效率的交通控制方式導致每年浪費超過一億加侖的燃料。本論文提出一個動態交通控制機制，可透過車載通訊來取得即時車輛資訊，例如：轉彎資訊、車輛在車道中的位置，以達到路口車流量最大化與公平車流分配。藉由車載通訊，路口的交通號誌可以在車輛轉彎之前收集所有的轉彎資訊。我們所提出的交通排程演算法考慮了車輛型別與轉彎資訊來安排綠燈給較高車輛通過率的車流，而且會考慮公平性來安排綠燈給較低車輛通過率的車流。模擬實驗結果顯示我們提出的機制優於現存的方法，此外，我們亦設計與實作了一個雛型系統以減少在車輛在路口怠速等待所產生的溫室氣體。

Traffic congestion in modern cities seriously affects the living quality and environments. Inefficient traffic management leads to fuel wastage in excessive of billion gallons per year. In this paper, we propose a dynamic traffic control framework using vehicular communications to utilize fine-grained information, such as turning intensions and lane positions of vehicles, for maximizing the intersection throughput and providing fairness among traffic flows. With vehicular communications, the traffic controller of an intersection can collect all turning information before vehicles make their turns. Our proposed signal scheduling algorithm takes the mixing of vehicles with different turning intensions into consideration. In addition, while allocating the green light to the traffic flows with the higher passing rates for throughput maximization, our framework also allocates the green light to the ones with the lower passing rates for fairness provision. Simulation results show that the proposed framework outperforms existing works. We also design and implement a prototype for minimizing greenhouse gases emitted by idling vehicles at intersections.