混合車流格位傳遞模式之建立與驗證

Abstract

Daganzo (1994; 1995)提出格位傳遞模式(Cell transmission model , CTM)，此模式為中觀車流模式，能有效模擬不同交通狀態之車流行為，CTM能反映簡單連續流模式(LWR)及密度與流量關係，CTM模式與微觀車流模式相比，所需運算時間較少，且CTM模式已能廣泛應用在高速公路與號誌化路口。傳統CTM模式係假設單一車流，如高速公路上汽車所有車型大小皆相同，若要利用CTM模式模擬市區道路混合車流(汽車與機車)，需考慮混合車流特性。在亞洲許多國家，機車與汽車混合車道為市區道路常見的現象，機車車型較小，若干騎士不會遵循車道分配行駛，甚至利用汽車與汽車間空隙行駛，當等候車隊產生時，可藉由汽車停等的空隙繼續前進，因此，當汽車佔滿下游格位時，機車依然能進入下游格位，機車無規律性的移動行為，對大車型之側向空間，可以增加格位的機車儲車空間，進行混合車流模擬時，如未考慮上述車流行為及不同車型車種的互動關係，會使模擬結果與實際車流狀態產生差距。 基此，本研究建立混合車流格位傳遞模式(mixed traffic cell transmission models, MCTM)模擬汽機車混合車流行為，兩車種以不同的車流行為競爭道路最大容量與剩餘儲車空間，因此，MCTM參數包含最大流率、最大儲車空間及剩餘儲車空間，這些參數根據不同混合比例與交通時相動態調整與分配，反映自由流、同步流與壅擠流車流行為。透過本研究實際觀察，建立兩種MCTM模式。第一個模式係參考上一格位汽車與機車之比例，決定汽車或機車能分配之道路資源(如容量與儲車空間)，也就是說汽車比例比機車大，汽車對道路資源的競爭優勢就更大，第二個模式進一步利用熵值(entropy)調整速率，因為兩車種運行的干擾，會隨著汽機車混合比例增加而增加。 本研究建立之MCTM驗證，拍攝台北市區三車道交通資料，拍攝時間涵蓋自由流到壅擠流時段，公車載運旅客行為，對混合車流有顯著的影響，為排除此互動關係，採用有公車專用道路段，因此選擇羅斯福路二段。驗證結果顯示，本研究建立的兩種MCTM模式在號誌時相及混合比例的變化下，皆可以準確的模擬，平均絕對誤差百分比(mean absolute percentage error, MAPE)值小於30%，此模擬結果於不同的空間(不同的格位)和不同的時間(綠燈時間 vs. 紅燈時間,離峰 vs. 尖峰)進行比較也有準確的結果。

Cell transmission model (CTM), proposed by Daganzo (1994; 1995), is a mesoscopic traffic flow model which can efficiently simulate traffic hydrodynamics under various traffic conditions. CTM can provide a convergent approximation to the Lighthill and Whiteham (1955) and Richard (1956) (LWR) models and covers the entire fundamental diagram. Due to its low computation requirements compared with micro-simulation models, CTM is widely applied to simulate the traffic flow behaviors at freeways and signalized intersection. The conventional CTM was designed for pure traffic, i.e. all vehicles are of the same size such as cars on freeways. Incorporation of more realistic CTM rules into the simulation of mixed traffic (various types of vehicles such as cars and motorcycles) on urban streets is comparatively less addressed. However, it is ubiquitous in many Asian urban streets that motorcycles oftentimes move concurrently with the cars by sharing the “same lane”. Some erratic motorcyclists do not even follow the lane disciplines at all. They may make lateral drifts breaking into two moving cars. Once blocked by the front vehicles, they even make wide transverse crossings through the gap of two stationary cars in the same lane, in order to move forward. That is, even the downstream cell is fully occupied by cars, motorcycles can still move forward. These erratic behaviors of motorcycles in the lateral clearance among larger vehicles would definitely increase the storage capacity of a cell. Without considerations of such behaviors and interaction between different sized vehicles, the traffic simulation might obtain a misleading result. Based on this, this study proposes mixed traffic cell transmission models (MCTM) to replicate the behaviors of mixed traffic consisted of cars and motorcycles. Both types of vehicles exhibit rather different traffic behaviors in competing roadway capacity and remaining storage space. Thus, the parameters of the MCTM, including maximal flow rate, maximal storage capacity, and remaining storage capacity, should be dynamically adjusted and allocated between cars and motorcycles according to mixture ratio of vehicles types and the traffic phases – free flow, synchronized, and congested. Based on our field observations, two MCTM models are developed. The first MCTM model uses the ratio of car to motorcycle in the last upstream cell to determine the amount of roadway resources (i.e. capacity and storage capacity) allocated to cars, and vice versa for motorcycles. That is, the larger value of car to motorcycle ratio is, the more advantageous of car in competing the roadway resource is. The second MCTM model further incorporates an entropy index to adjust the traffic speed, since the interferences between two types of vehicles will be rapidly increased as the mixture ratio of cars and motorcycles becomes higher. To validate the proposed MCTM models, videotaping traffic data of three corridors in Taipei city covering a full traffic spectrum from free-flow to congested are collected. Since the traffic behaviors will be significantly affected by the passengers embarking and disembarking maneuvers of bus traffic, to exclude such interactions, the corridors with a bus exclusive lane, Sec. 4, Roosevelt Road, are selected. The results show that the both proposed MCTM models can accurately simulate the traffic flows under various traffic phases and mixture ratios with an average MAPE (mean absolute percentage error) below 30%. The simulation accuracy is also compared spatially (different cells) and temporally (green time vs. red time, off-peak hour vs. peak hour).