Estimation Of Travel Time At Signalized Intersection Based On Data Fusion Of Two Types Of Floating Cars

Signalized intersection is a key node in the urban road network, the travel time in this region can effectively reflect the state of transportation and service level of urban roads. However, the urban road network traffic is full of random itself, especially in the area of signalized intersection. Because of the controlling of the traffic lights each of the direction in the intersection obtains right of way periodically, which further exacerbats the region's traffic randomness, resulting its travel time with a very large uncertainty. Therefore, accurate and reliable estimating its travel time in signalized intersection, can provide more reliable data base in designing and planning the signalized intersection. Besides, it can also promote the level of service in signalized intersection, which has important practical significance.Aimed at the uncertainty of travel time at the signalized intersection, the paper used bus and taxi GPS data to propose a method to estimate its travel time of the single vehicle effectively. Then the paper explored the characteristics of the travel time between buses and taxis at signalized intersection to find a way to enhance the accuracy of the estimation. The main contents include:(1)Time estimation of a single vehicle at signalized intersection. By analysising the kinetic characteristic of the vehicle driving through signalized intersection, the paper determined the scope of the lights. On this basis, it analysed the distribution of GPS data in upstream and downstream of the intersection, then proposed a way to estimate its travel time of the single vehicle effectively. Finally, it used the actual GPS data to verify the effectiveness of the method. The results showed that the method is better than traditional methods in estimating the travel time at the intersection.(2)Estimation of the average travel time at signalized intersection. By analysising the impacts of signal lights on the travel time of vehicles at signalized intersection, the paper determined the optimal estimating period of the average travel time in the region. Aimed at the clustering of travel time of vehicles at signalized intersection, the paper fused the same type of floating cars to eliminate the impact of individual factors on one floating car. Further according to two types of the floating cars, the paper proposed a method to estimate the average travel time at signalized intersection based on the data fusion of two types of floating cars. Finally, it used VISSIM simulation to verify the effectiveness of the method. The results showed that the method the method is better than the method of using one type of floating car.Lastly, introducing the design and implement of the system, and appling to the average travel time at signalized intersection called CaiFuZhongXin in ChongQing. The results showed that comparing with the contrast method, this paper had a higher estimated accuracy.