Search

This paper reports the development of a public transport trip planner to help the urban traveller in planning and preparing for his commute using public transportation in the city. A Genetic Algorithm (GA) approach that handles realtime Global Positioning Systems (GPS) data from buses of the Metropolitan Transport Corporation (MTC) in Chennai City (India) has been used to develop the planner. The GA has been shown to provide good solutions within the problem's computation time constraints. The developed trip planner has been implemented for static network data first and subsequently extended to use real-time data. The "walk mode" and Chennai Mass Rapid Transit System (MRTS) have also been included in the geospatial database to extend the route-planner's capabilities. The algorithm has subsequently been segmented to speed up the prediction process. In addition, a temporal cache has also been introduced during implementation, to handle multiple queries generated simultaneously. The results showed that there is promise for scalability and citywide implementation for the proposed real-time route-planner. The uncertainty and poor service quality perceived with public transport bus services in India could potentially be mitigated by further developments in the route-planner introduced in this paper. [ABSTRACT FROM AUTHOR]

Link performance functions commonly used for traffic assignment are often based on Volume Delay Functions (VDF) developed for homogeneous traffic. However, VDFs relating stream speed to the volume of traffic based on homogeneous lane-based traffic are not adequate for traffic assignment in developing countries due to the heterogeneous nature of traffic that is characterized by a mix of a wide range of vehicle classes with significant differences in static and dynamic characteristics and an imperfect lane discipline. Unfortunately, the use of VDFs similar to those for homogeneous traffic flow situations imposes strong restrictions considering two respects: 1) travel times at path and link levels can be obtained for an aggregated stream but not for individual vehicle types; 2) the effect of varying composition and asymmetric interactions is captured only to a limited extent by converting all vehicles into equivalent Passenger Car Unit (PCU). Hence, this paper proposes the development of VDFs specific to different classes of heterogeneous traffic, as it is more realistic in traffic assignment than the use of the same VDF for all classes of vehicles in a link. This study is aimed at developing models to determine the speed of each vehicle class as a function of flow and composition for six lane roads with heterogeneous traffic based on data obtained from Chennai city, India. Heterogeneity in this study mainly refers to differences in vehicle types (two-wheeler, car, bus, etc.) participating in mixed traffic. To develop multiple user class VDFs, the speed and flow of each vehicle class for a wide range of traffic flow conditions need to be recorded. As this is not possible using field measurements, an established micro-simulation model (HETEROSIM) is used for determining speeds for each vehicle type by systematically varying the volume and composition levels over a range of values that represent relevant and practical traffic conditions observed in six lane divided roads in Chennai city. The proposed delay functions are different from standard single user class VDFs in three key respects: first, they enable more realistic behaviour by modelling differences in class wise speeds at a given volume and composition level; second, they allow for capturing asymmetric interactions of different vehicle types on an average speed of a given vehicle class. Finally, speed-flow relationships for each class are also allowed to vary across volume levels which enable the representation of differential interactions at different levels of congestion in mixed traffic. The need for homogenizing the volumes in terms of a single class is obviated. The models significantly outperformed single class VDFs in both calibration and validation datasets. Further, the proposed models are used for analyzing heterogeneous traffic characteristics. Empirical evidence of asymmetric interactions and the impact of composition on classwise performance are also found and quantified. Finally, two applications of the proposed models are demonstrated for the level of service analysis of different classes and impact analysis of excluding some classes. The proposed models may have applications such as determining class wise road user costs and performance measures (e.g. emissions) that depend on class-specific speeds. [ABSTRACT FROM AUTHOR]

Real-time bus travel time prediction has been an interesting problem since past decade, especially in India. Popular methods for travel time prediction include time series analysis, regression methods, Kalman filter method and Artificial Neural Network (ANN) method. Reported studies using these methods did not consider the high variance situations arising from the varying traffic and weather conditions, which is very common under heterogeneous and lane-less traffic conditions such as the one in India. The aim of the present study is to analyse the variance in bus travel time and predict the travel time accurately under such conditions. Literature shows that Support Vector Machines (SVM) technique is capable of performing well under such conditions and hence is used in this study. In the present study, nu-Support Vector Regression (SVR) using linear kernel function was selected. Two models were developed, namely spatial SVM and temporal SVM, to predict bus travel time. It was observed that in high mean and variance sections, temporal models are performing better than spatial. An algorithm to dynamically choose between the spatial and temporal SVM models, based on the current travel time, was also developed. The unique features of the present study are the traffic system under consideration having high variability and the variables used as input for prediction being obtained from Global Positioning System (GPS) units alone. The adopted scheme was implemented using data collected from GPS fitted public transport buses in Chennai (India). The performance of the proposed method was compared with available methods that were reported under similar traffic conditions and the results showed a clear improvement. [ABSTRACT FROM AUTHOR]