減少GPS探偵車回報數量之交通資訊系統設計

Abstract

提供用路人即時交通資訊可避過塞車、節省行車時間及能源消耗。由於智慧型手機裝載GPS接收器與無線資料傳輸技術且逐漸普及，利用智慧型手機作為GPS探偵車回報車速取得即時交通資訊成為可行的方式。相對於傳統固定式車輛偵測器而言，GPS探偵車大幅降低佈建與維護成本。然而目前GPS探偵車的回報方法可分為：固定週期回報或是固定地點回報，無論是哪種，皆會造成回報資訊超載的問題，若使用條件式回報可降低回報量，但傳統的條件式回報方式交通資訊中心不易計算出精確之平均旅行時間。本研究基於條件式回報支方式，提出交通資訊中心提供下個交通號誌週期之最大旅行時間與最小旅行時間給GPS探偵車，探偵車根據收到的資料與自身旅行資料相比做選擇性回報，能有效降低交通資訊回報次數和成本，並掌握精確的交通資料。為了分析系統效能，我們使用運輸模擬軟體VISSIM來進行實驗模擬。研究成果證實，我們所提供的資料的方法可大量減少探偵車資訊回報之數量，減輕交通資訊中心之負載，並能精確掌握交通狀況之波動。

Providing real-time traffic information can help road users to avoid congestion, save traveling time and reduce fuel consumption. With the increasing popularity of smartphones, which have GPS receiver and wireless data communication capability, smartphones on traveling vehicles can be used as probes to obtain the real-time traffic information, such as speed and travel time. The conventional report policy of GPS-equipped probes can be classified into two categories: periodical report and segment-based report. Both approaches may cause a large amount of redundant report messages if there are too many probes on the same road segment. Using conditional report can reduce the communication overhead, but it’s difficult for traffic information center (TIC) to compute the accurate average travel time from limited conditional reports. We propose a report policy based on conditional report policy, and our TIC predicts the maximum and minimum travel time for the next traffic light cycle, broadcasts the predicted travel times to probes. The probes report to the TIC only when their travel times are close to the maximum or the minimum travel time. To evaluate the performance of the system, we have performed simulations using traffic model simulator VISSIM. The simulation results indicate that our approach significantly reduces the number of traffic reports and the loading of the TIC, and we can accurately catch the change of the travel time during traffic congestions.