Journal of Transportation Engineering, Part A: Systems

The taxi–passenger problem is more and more prominent in terminal stations; passengers usually wait in a long queue for taking a taxi in rush hour, and otherwise taxis wait in a queue for passengers. Nevertheless, there are few ways for dispatchers handling taxis’ and passengers’ queues to save time. In this paper, we build a new taxi–passenger dispatching model to solve this problem. The model contains four parts: information collection, passenger dispatching model with dispatcher control, taxi dispatching model, and emergency dispatching strategies. Different from traditional thought in the problem of double-ended queue theory, it is an analytical model with a simulation method. The numerical result of the passenger and taxi dispatching model can be used in station engineering and management. The method is based on queue theory, the simulation model, and technologies on the sensors and wireless communication. From the passenger dispatching results, we obtain the optimal release curve to direct the optimal number of passengers to be released in different arrival rates. The taxi dispatching results shows the strategy point of dispatching taxis. And the simulation results with the dispatching model perform much better than the traditional method in four scenarios. Sensitivity analysis is conducted that shows the different pattern of optimal release curve in variable conditions.

Several macroscopic approaches exist for modeling the flow of the vehicular traffic in the developed economies, from kinematic wave models using first-order or higher-order systems of partial differential equations to spatially and temporally discretized models like cell transmission models and noncontinuum models that treat traffic as a collection of dynamic systems. The adaptation of these modeling approaches to the traffic in developing economies is made difficult by the latter’s high levels of heterogeneity and weak lane discipline. However, most of the existing studies on heterogeneous or multiclass traffic implicitly assume lane discipline. In the present article, the notion of heterogeneity in the contexts of developed and developing economies is discussed. It is argued that the differences in their aggregate traffic behaviors are a direct result of the presence of the small-sized, highly maneuverable vehicles like motorcycles and auto-rickshaws. In addition, vehicles are classified into two types: car-following and gap-filling, and two different forms of fundamental relationships are derived using some simplifying assumptions. Finally, a heuristic cell transmission model that is capable of reproducing the vehicle creeping phenomenon that is a salient feature of the heterogeneous, disordered traffic is presented, and its results are compared with those of a numerical scheme for a multiclass Lighthill-Whitham-Richards model.

Pedestrian safety is a major concern throughout the world because pedestrians are considered to be the most vulnerable roadway users. This paper sought to identify the main factors in pedestrian-vehicle crashes that increase the risk of a fatality or severe injury. Pedestrian-vehicle crashes which occurred in urban and suburban areas in Jordan between 2009 and 2011 were investigated. Extracted rules from Bayesian networks were used to identify factors related to severity of pedestrian-vehicle crashes. To obtain as much information as possible about these factors, three subsets were used. The first and second subsets contain all types of collisions (pedestrian and nonpedestrian), in which the first subset used collision type as a class variable and the second subset used injury severity. The third subset contains pedestrian collisions only and used injury severity as the class variable. The results indicate that when using collision type as the class variable, better performance was obtained and that the following variables increase the risk of fatality or severe injury: roadway type, number of lanes, speed limit, lighting, and adverse weather conditions.

Route travel time varies with vehicles and traffic demand. Besides the average route travel time, route travel time reliability in the form of travel time distribution is indispensable. However, the sample size of Complete Route Travel Times (TTC) is rather small for many reasons. Existing methods using convolution distribution rely on strong assumptions about the correlation structure or the link travel time distributions; other methods relying on scaled Partial Route Travel Times (TTP) may extend the estimation bias. To overcome these issues, we present an estimation method for route travel time distribution by fusing kinds of route travel time information from Automatic Number Plate Recognition (ANPR) data. The proposed method firstly improves the data quality for estimation in four steps, including route redefinition, observation extraction, path inference, and scaling. Secondly, using TTP data, it convolutes the empirical travel time distributions on all the partial routes divided at the breakpoints identified by the Hopkins statistics. Thus, the link correlations are considered and the assumption about the correlation structure is eschewed. Thirdly, the convolution distribution and TTC information are fused to estimate the actual route travel time distribution based on Bayes’ theorem and Shannon’s information entropy. Finally, estimation results using different methods are compared to evaluate the developed model.

In several emerging countries such as India, premium quality buses (buses with enhanced comfort and security) are introduced with an aim to attract private car owners (choice riders). However, lack of ridership and high subsidy are the two key concerns regarding the operation of premium buses. This article documents a rational approach to address these concerns by improving the service characteristics of premium buses giving due consideration to the requirements of choice riders. The approach includes three major components: (1) development of generalized cost (GC) and demand model for choice riders, (2) identifying improvement levels for premium service attributes, and (3) development of a simulation model for evaluating alternative improvement scenarios. The case study on selected routes suggested that marginal fare increment and higher level of service are instrumental in eliminating the requirement of subsidy. In addition, even with the current fare, it is possible to improve the service, and thereby substantially reduce the burden of subsidy to the government. The work also highlights the importance of accounting travel behavior of users in service design, which is contrary to the concept of fare-centric design of bus services prevailing in several emerging countries.

Wide moving jams (WMJ) have been reported for over a decade in the three-phase theory. However, existing studies about wide moving jams have mostly been based on simulation data or limited empirical data. In addition, wide moving jams have been studied by only a few researchers, and identification of wide moving jams is still an important issue. Based on a large amount of real-world floating car data (FCD) in Beijing, this paper discusses the identification of wide moving jams and evaluates their impact on traffic flows. The identification of wide moving jams was studied based on the spatiotemporal characteristics of link-based travel speeds. It was found that wide moving jams were widely recognized. Next, the characteristics of traffic speed and the propagation speed of wide moving jams were studied. The average vehicle speed inside wide moving jams was $8.0\mathrm{km}/\mathrm{h}$, and the average propagation speed of wide moving jams was $-8.7\mathrm{km}/\mathrm{h}$. Finally, the influence of wide moving jams on traffic flows was quantitatively studied for the 3rd Ring Road in Beijing. This paper contributes to the identification of wide moving jams based on link-based FCD with specific space-time granularity instead of the trajectory-based FCD that has been used in previous studies. Wide moving jams are common occurrences, but they are difficult to identify due to the inappropriate space-time granularity of the data. Wide moving jams show different characteristics when the space-time granularity of the data is different. The engineering applications of the findings in this paper include two aspects: (1) traffic control to avoid traffic breakdown induced by wide moving jams, and (2) predictive routing for travelers according the propagation speed of wide moving jams. The influence of wide moving jams on the probability of traffic breakdown and capacity in wide moving jams are further studied.

Road network vulnerability analysis is helpful in the improvement of vulnerable links with proper maintenance investments and management strategies. This paper addresses issues that have received limited attention in past studies: estimation of travel demand after link disruption and analysis of accessibility variation. An activity-based model was used to estimate travel demand changes due to link closure, and link importance was evaluated using a set of vulnerability indicators. Accessibility changes induced by link closure are presented and discussed. Vulnerability analysis was conducted for the road network of the municipality of Dolo in northern Italy. Considering the spatial distribution of activities and trips, the results obtained with the activity-based model were more reliable than those obtained with the fixed demand model, which makes the unrealistic assumption of unchanged travel demand after network degradation. These findings are relevant for appropriate resource allocation strategies, which depend on correct link vulnerability analysis and ranking.