Regional Transportation Network Travel Time Estimation Based on Transition of Traffic Flow Phase

Travel time is widely recognized as an important performance measure for assessing transportation system, it is a meaningful parameter in theory and practice. Previous researches about travel time mainly focus on two areas: first, the research object is the unit of the transportation network, and there is little research on the area transportation network travel time; second, previous researches about travel time mainly focus on travel time forecast which using history traffic data. But, in fact, the confidence of the forecasted travel time can't be tested because the instantaneous travel time can't be measured in practice. Not only that, previous researches about the travel time forecasting are under one traffic state (congestion state), they don't consider the impact of the traffic state variation. The traffic flow state of the area transportation network is mutative, and the travel time of the multi-state area transportation system can be estimated using the random process theory. In this paper, we divide the traffic flow into three states using C-mean fuzzy clustering method and develop the travel time function under every state. We also analyze the state changing of the unit of the area transportation network using semi-Markov random process theory and develop a travel time estimating model to evaluate the area transportation network travel time. We use two types of data to develop and verify the model. The first type of the data is traffic flow data (include traffic flo w velocity, speed and occupation). The second type of the data is the travel time, which was got by the plate matching method. Using the two types data of changzhou china of 1/3-10/3(2010), we develop the model. Using the data of changzhou china of 11/3- 11/3(2010), we verify the result. Estimates from the model are compared to the field-measured travel time; the paired t-test for the mean difference is conducted too. The results show that at the significance level of α=0.05, there is no significance difference between the estimated travel time and the field mean travel time.