Search

This study investigated the personal rapid transit (PRT) guideway network design problem at a strategic level. PRT is known as an on-demand mobility method serving the first or last mile of travel demand, which can fluctuate significantly during a day. This on-demand property may make the system vulnerable to changes in the pattern of demand. The robustness of the PRT guideway network refers to the capability to fit different demand patterns. The motivation of this study was to take the uncertainty of demand into consideration in the process of network design to enhance the robustness of the PRT guideway network in operation. This problem was formulated as a two-stage stochastic linear programming model. The model was applied to a case study with real travel demand data from Singapore. The results of the comparison study show that the network designed via the proposed model is robust and cost-effective.

This paper studies the daily storage yard manage problem arising in maritime container terminals, which integrates the space allocation and yard crane deployment decisions together with the consideration of container traffic congestion in the storage yard. The integrated problem is formulated as an integer linear programming model with the objective of minimizing the yard crane operating cost and the yard crane interblock movement cost. A divide-and-conquer solution approach is designed to solve the problem in an efficient manner in which harmony search and constraint satisfaction techniques are employed. Numerical experiments are conducted to validate the performance of the solution approach and the improvement from the integrated optimization method.

This paper investigates the network design problem arising from the regional hazardous waste management system. The problem is to identify the locations of various waste facilities, and determine the transportation routes of hazardous wastes and waste residues between those waste facilities. Aiming at minimizing jointly the total cost and total risk, the problem is formulated as a multi-objective mixed integer linear programming model. By exploiting the advantages of the model, three multi-objective optimization approaches are customized to find highly qualified non-dominated solutions. The effectiveness and efficiency of the approaches are examined both on a hypothetical case and a realistic case.

Quay-side berthing congestion is an emerging challenging issue that arises in busy container transshipment terminals and calls for effective management of terminal operations. This paper tackles the berthing congestion problem by introducing a proactive management strategy from the terminal’s perspective that adjusts the calling schedule of feeder vessels in such a way that the quay-side workload distribution in the temporal dimension can be balanced. Such a schedule template design problem is considered simultaneously with another two tactical level decision problems, berth template design (i.e., determining preferred berthing positions for vessels) and yard template design (i.e., allocating storage yard space to transshipment flows). This highly integrated problem is formulated as a set covering model. Heuristic methods based on column generation are developed to obtain near-optimal solutions in an efficient way. Computational experiments on real-world sized test instances demonstrate the efficiency and effectiveness of the proposed approach.

Transportation Research Part A: Policy and Practice, 69, ISSN:0965-8564, ISSN:1879-2375

Timetable design is crucial to the metro service reliability. A straightforward and commonly adopted strategy in daily operation is a peak/off-peak-based schedule. However, such a strategy may fail to meet dynamic temporal passenger demand, resulting in long passenger waiting time at platforms and over-crowding in trains. Thanks to the emergence of smart card-based automated fare collection systems, we can now better quantify spatial–temporal demand on a microscopic level. In this paper, we formulate three optimization models to design demand-sensitive timetables by demonstrating train operation using equivalent time (interval). The first model aims at making the timetable more dynamic; the second model is an extension allowing for capacity constraints. The third model aims at designing a capacitated demand-sensitive peak/off-peak timetable. We assessed the performance of these three models and conducted sensitivity analyzes on different parameters on a metro line in Singapore, finding that dynamical timetable built with capacity constraints is most advantageous. Finally, we conclude our study and discuss the implications of the three models: the capacitated model provides a timetable which shows best performance under fixed capacity constraints, while the uncapacitated model may offer optimal temporal train configuration. Although we imposed capacity constraints when designing the optimal peak/off-peak timetable, its performance is not as good as models with dynamical headways. However, it shows advantages such as being easier to operate and more understandable to the passengers.

This paper advances the field of network disruption analysis by introducing an application to a multi-modal transport network, capitalizing on the redundancies and improved connectivity of an integrated metro-bus network. Metro network resilience to disruptions can be enhanced by leveraging on public bus services. To ensure better acceptance among operators and commuters, we focus on introducing localized integration with bus services instead of designing an entirely new bus network to achieve the desired resilience to potential disruptions. This is accomplished by increasing the capacity of bus services that run in parallel with affected metro lines as well as those connecting to different metro lines. Our analysis starts with a network representation to model the integrated metro and bus system. A two-stage stochastic programming model is further developed to assess the intrinsic metro network resilience as well as to optimize the localized integration with bus services. The approach is applied to a case study based on the Singapore public transit system and actual travel demand data. The results show that the metro network resilience to disruptions can be enhanced significantly from localized integration with public bus services.

This paper investigates the Train Network Design Optimization problem arising from railroad industry which involves the integration of three inter-related decision sub-problems: train routing which is to identify origin, destination and itinerary for individual trains; block-to-train assignment detailing the block movements with trains and swaps between trains; and crew-to-train assignment specifying the crew services for train routes. A column generation based hierarchical approach with two stages is designed: the first stage generates a pool of promising train routes iteratively based on the crew segments by the column generation technique; and the second stage develops an integer linear programming model for the subsequent decisions including train route selection and block-to-train assignment. Numerical experiments with realistic test instances are conducted and the outcome demonstrates the capability of the proposed approach in solving the Train Network Design Optimization problem competently.

The feeder vessel management problem consists of designing preferred berthing positions and service time for cyclically visiting feeders, and allocating storage yard space to the transshipment flows between mother vessels and feeders. We consider the above three tactical decision problems simultaneously for a container transshipment terminal with an eye toward the quayside congestion and the housekeeping cost of container movements. The integrated problem is formulated as a mixed integer programming model and solved by a memetic heuristic approach. A comprehensive computational experiment is conducted to show the effectiveness of the heuristic and the improvement upon real-world terminal operations.

This paper presents an integer programming model for the terminal and yard allocation problem in a large container transshipment hub with multiple terminals. The model integrates two decisions: terminal allocation for vessels and yard allocation for transshipment container movements within a terminal as well as between terminals. The objective function aims to minimize the total inter-terminal and intra-terminal handling costs generated by transshipment flows. To solve the problem, we develop a 2-level heuristic algorithm to obtain high quality solutions in an efficient way. Computational experiments show the effectiveness of the proposed approach.

The quayside operation problem is one of the key components in the management system for a container terminal. In this paper, the integrated models proposed in the previous studies to address the quayside operation problem are examined and one of the potential frameworks is identified. A new method called combinatorial benders’ cuts algorithm is developed to solve the berth-level model in the framework. The computational experiment conducted in this research shows that the proposed approach is more efficient than the branch and cut algorithm embedded in CPLEX.

This paper studies the management of transshipment flows in a container terminal. Management consists of designing a visiting schedule template for feeder vessels and determining storage locations for transshipment containers. A mixed integer programming formulation is developed; it includes the objective of minimizing both the total distance traveled by transshipment flows between quayside and storage yard and the workload imbalance in time. To solve the problem, an algorithm is devised on the basis of Lagrangian relaxation to find near-optimal solutions within short computational times. Numerical experiments are conducted to assess (a) the effectiveness and efficiency of the algorithm and (b) the benefit from adjusting feeder visiting schedule.

This paper discusses the quay crane scheduling problem at indented berth, an extension to the current quay crane scheduling problem in the field of container terminal operation. A mixed integer programming model by considering the unique features of the quay crane scheduling problem at indented berth is formulated. For solution, decomposition heuristic framework is developed and enhanced by Tabu search. To evaluate the performance of the proposed heuristic framework, a comprehensive numerical test is carried out and its results show the good quality of the proposed heuristic framework.

This article explores the quay crane scheduling problem at an indented berth. The indented berth is known as an innovative implementation in the container terminals to tackle the challenge from the emergence of more and more mega-containerships. A mixed integer programming model by considering the non-crossing and safety distance constraints is formulated. A Tabu search heuristic is developed to solve the proposed problem. The computational results from this research indicate that the designed Tabu search is an effective method to handle the quay crane scheduling problem at an indented berth.

Most routing decisions or diversion plans in the literature are made based on the current traffic situation and little attention is paid to future. Such kinds of greedy strategy consistently suggest en route diversion at any point where the alternative route choice is better than the ongoing one. Hence, in the constantly varying real world, excessive diversions may appear and are much likely to create a sort of ‘zig-zag’ effect, eventually leading to a higher cost. A forward-looking diversion strategy considering both real-time and future information is proposed in this paper, and the potential benefits, limitations and difficulties of several diversion strategies are evaluated through microscopic traffic simulations. The research will help logistics companies to assess the benefits of en route guidance system in their day-to-day operations.

This paper involves a study of signal priority strategies for bus operations using Radio Frequency Identification (RFID) technology enabled devices. An actual suburban network of the Clementi area of Singapore, which including several signalized intersections, was used for microscopic traffic simulations conducted in this study. Based on the experimental results, the proposed system has the potential to improve public bus service level by reducing congestion of bus traffic flow through street intersections.

This paper involves a study of a real-time system for monitoring the security of commercial vehicles in road networks. Embedded in the security monitoring system is a commercial vehicle tracking and incident detection algorithm which relies on a combination of vehicle telemetry data obtained from Global Positioning Systems and on-board sensors to continuously monitor the route choice and car-following behavior of the driver. The performance of this algorithm has been tested in a microscopic simulation model, on a set of hypothetical scenarios, which included deviations from the approved routes, forced to travel at unreasonably low speeds, or even stopped at unexpected places in the network. The initial results indicate that the proposed system has good potential in detecting abnormal driving behaviors, with 100% detection rate, 6.0 seconds of mean detection time, and less than 1% false alarm rate.

This paper presents a study of taxi booking service in Singapore using a microscopic traffic simulation model embedded with a link-to-link shortest path algorithm. A novel trip-chaining strategy for taxi booking based on a customized algorithm of Pickup and Delivery Problem with Time Window (PDPTW) was proposed. The idea is to chain several bookings with demand time points which are spread out within a reasonable period of time, and with each pick-up point coinciding with or being within close proximity to the previous drop-off location. Based on the simulation results, the proposed system has the potential to improve the taxi booking service currently operating in Singapore.

The integrated problem for bay allocation and yard crane scheduling in transshipment container terminals was addressed. Unlike space allocation under the entire yard overview and slot assignment within a yard bay, the bay allocation problem focuses on a block and aims to allocate bay resource to fleets of transshipment containers in a more efficient way. The receiving operation and retrieving operation in the storage yards were considered simultaneously to achieve a more efficient operation of yard crane. The bay allocation and the yard crane scheduling were integrated as a whole process. A mixed integer programming model was developed for the problem formulation with the objective of minimizing total costs, including yard crane cost and delay cost. Considering the high complexity of the problem, a simulated annealing heuristic algorithm was developed to obtain near-optimal solutions. Numerical experiments were also conducted to test the efficiency of the proposed algorithm.

The quay crane scheduling problem studied in this article is to determine the handling sequence of ship bays for quay cranes assigned to a container ship considering handling priority of every ship bay. This article provides a mixed integer programming model for the considered quay crane scheduling problem that is 𝒩𝒫-complete in nature. An approximation algorithm is proposed to obtain near optimal solutions and a worst-case analysis for the approximation algorithm is performed. Computational experiments are conducted to examine the proposed model and solution algorithm. The computational results show that the proposed approximation algorithm is effective and efficient in solving the considered quay crane scheduling problem.

In this research, continuous and dynamic Berth Allocation Problem (BAP) is studied to minimize the total weighted flow time. Different from previous studies on continuous BAP, an efficient method is proposed to address the problem to identify the possible locations for next vessel in the Time-space diagram. Then two versions of Greedy Randomized Adaptive Search Procedure (GRASP) are developed to search for near optimal solutions. Both small and large scale numerical experiments are tested to examine the effectiveness of the proposed GRASPs by comparison with CPLEX and stochastic beam search, respectively.

Quay crane and yard truck scheduling are two important subproblems in container terminal operations which have been studied separately in previous research. This paper proposes a new problem for the integrated quay crane and yard truck scheduling for inbound containers. The problem is formulated as a mixed integer programming (MIP) model. Due to the intractability, a genetic algorithm (GA) and a modified Johnson's Rule-based heuristic algorithm (MJRHA) are used for the problem solution. In addition, two closed form lower bounds are given to evaluate the solution accuracy. Computational experiments show that the solution algorithm can efficiently handle the scheduling problem and that the integrated methods are very useful.

This paper presents a novel multiagent approach to automating taxi dispatch that services current bookings in a distributed fashion. The existing system in use by a taxi operator in Singapore and elsewhere, attempts to increase customer satisfaction locally, by sequentially dispatching nearby taxis to service customers. The proposed dispatch system attempts to increase customer satisfaction more globally, by concurrently dispatching multiple taxis to the same number of customers in the same geographical region, and vis-a-vis human driver satisfaction. To realize the system, a multiagent architecture is proposed, populated with software collaborative agents that can actively negotiate on behalf of taxi drivers in groups of size N for available customer bookings. Theoretically, an analysis of the boundary and optimal multiagent taxi-dispatch situations is presented along with a discussion of their implications. Experimentally, the operational efficiency of the existing and proposed dispatch systems was evaluated through computer simulations. The empirical results, obtained for a 1000-strong taxi fleet over a discrete range of N , show that the proposed system can dispatch taxis with reduction in customer waiting and empty taxi cruising times of up to 33.1% and 26.3%, respectively; and up to 41.8% and 41.2% reduction when a simple negotiation speedup heuristic was applied.

This paper proposes a novel integrated model for yard truck and yard crane scheduling problems for loading operations in container terminal. The problem is formulated as a mixed-integer programming model. Due to the computational intractability, two efficient solution methods, based on Benders’ decomposition, are developed for problem solution; namely, the general Benders’ cut-based method and the combinatorial Benders’ cut-based method. Computational experiments are conducted to evaluate the effectiveness of the proposed solution methods.

The integrated discrete berth allocation and quay crane scheduling problem studied in this article considers the relationship between them. This article provides a mixed integer programming model for the studied problem that is -complete in nature. A genetic algorithm is proposed to obtain near optimal solutions. Computational experiments are conducted to examine the proposed solution algorithm. The computational results show that the proposed genetic algorithm is effective and efficient in solving the studied problem.

An empirical and comparative evaluation of multiagent taxi dispatch with extended (E) runtime taxi availability is presented. A taxi in operation is said to be E-runtime available if it has a passenger alighting in ?x > 0 minutes' time or is empty, but has no new committed taxi request to service next. In a multiagent architecture, we consider a new operation policy wherein agents of E-runtime available taxis are allowed to negotiate in individual groups of size N for new taxi requests. The main objective is to present an evaluation of the multiagent system performance gains provided by different times-to-arrival of ?x, under a discrete range of demand rates for several iV-group sizes, as compared with the base case when ?x = 0. It is shown that the proposed policy can effectively reduce customer waiting time and empty taxi cruising time by up to about 60% and 96%, respectively, when the service demand is high for a 1000-strong taxi fleet. It is observed that the value selection for the policy parameter ?x is an important aspect for improving the general performance of multiagent taxi dispatch.

The yard truck scheduling and the storage allocation are two important decision problems affecting the efficiency of container terminal operations. This paper proposes a novel approach that integrates these two problems into a whole. The objective is to minimize the weighted sum of total delay of requests and the total travel time of yard trucks. Due to the intractability of the proposed problem, a hybrid insertion algorithm is designed for effective problem solutions. Computational experiments are conducted to examine the key factors of the problem and the performance of the proposed heuristic algorithm.

Yard truck scheduling and storage allocation, as two separate subproblems in port operations, have been extensively studied in the past decades. However, from the operational point of view, they are highly interdependent. This article proposes an integer programming model in which yard truck scheduling and storage allocation problems are formulated as a whole for heterogeneous import containers. Different stacking times at yard blocks is modelled as well. The objective of the proposed model is to reduce the congestion and waiting time of yard trucks in the terminal so as to decrease the makespan of discharging containers. Owing to the inherent computational complexity, a genetic algorithm and a greedy heuristic algorithm have been designed. Computational experiments show that the proposed genetic algorithm and greedy algorithm are both effective in solving the studied problem.

Load scheduling problems of the quay crane (QC) and yard crane (YC) are important issues in port container terminal operations. Because the two scheduling problems are closely related, this paper proposes a mathematical model to formulate the load scheduling problem of the QC and YC simultaneously. A genetic algorithm (GA) and a heuristic method [QC and YC scheduling heuristic (QYSH)] are designed for the problem solution. Results obtained through numerical experiments showed that the problem-oriented QYSH method significantly outperformed the GA and could be applied in real operations to conduct simultaneous quay crane and yard crane scheduling.

The design of reverse logistics network has attracted growing attention with the stringent pressures from environmental and social requirements. In general, decisions about reverse logistics network configurations are made on a long-term basis and factors influencing such reverse logistics network design may also vary over time. This paper proposes dynamic location and allocation models to cope with such issues. A two-stage stochastic programming model is further developed by which a deterministic model for multiperiod reverse logistics network design can be extended to account for the uncertainties. A solution approach integrating a recently proposed sampling method with a heuristic algorithm is also proposed in this research. A numerical experiment is presented to demonstrate the significance of the developed stochastic model as well as the efficiency of the proposed solution method.

This paper addresses the load scheduling problem of multiple yard crane systems in container terminals with buffer areas. In the problem several yard cranes are used to pick up a sequence of containers for a quay crane. The containers picked up ahead of schedule can be stored in buffer areas until they can be handled by the quay crane. An integer programming model is developed for the problem formulation. In the model the work schedules of different yard cranes are decided simultaneously to minimize the loading time. The yard crane scheduling problem is NP-complete by nature. Therefore a scheduling heuristic is designed to solve the proposed problem. Numerical experiments are conducted to compare the performance of the multiple yard crane systems in container terminals with and without buffer areas.

Truck scheduling and storage allocation, as two separate subproblems in port operations, have been deeply studied in past decades. However, from the operational point of view, they are highly interdependent. Storage allocation for import containers has to balance the travel time and queuing time of each container in yard. This paper proposed an integer programming model handling these two problems as a whole. The objective of this model is to reduce congestion and waiting time of container trucks in the terminal so as to decrease the makespan of discharging containers. Due to the inherent complexity of the problem, a genetic algorithm and a greedy heuristic algorithm are designed to attain near optimal solutions. It shows that the heuristic algorithm can achieve the optimal solution for small-scale problems. The solutions of small-and large-scale problems obtained from the heuristic algorithm are better than those from the genetic algorithm.

This paper discusses the logistics network design for end-of-lease computer products recovery by developing a deterministic programming model for systematically managing forward and reverse logistics flows. Due to the complexity of such network design problem, a two-stage heuristic approach is developed to decompose the integrated design of the distribution networks into a location–allocation problem and a revised network flow problem. The applicability of the proposed method is illustrated in a numerical study. Computational experiments demonstrate that high-quality solutions are obtained while modest computational overheads are incurred.

This special issue contains revised versions of selected papers originally presented at the 9th IEEE International Conference on Intelligent Transportation Systems (ITSC 2006) held in Toronto, Canada, on September 17-20, 2006.

The quay crane scheduling problem studied in this article is to determine a handling sequence of ship bays for quay cranes assigned to a container ship considering handling priority of every ship bay. This article provides a mixed integer programming model for the considered problem. A genetic algorithm (GA) is proposed to obtain near-optimal solutions. Computational experiments to examine the proposed model and solution algorithm are described. The computational results show that the proposed GA is effective and efficient in solving the considered quay crane scheduling problem.

Increasing security and maximizing throughput toward achieving a sustainable seaport is an immediate challenge confronting seaport operators, especially after the terrorist attacks of September 11, 2001. Clearly, introducing container-inspection processes in a seaport will impede the current workflow schedule of container operations and no doubt incur additional management and operation costs, even if the inspection equipment used is highly efficient. Hence, this paper aims to evaluate the impact of imposing container-inspection procedures on container port operations and to optimize port throughput at the same time. The extra handling effort required for security inspection can then be traded off against port efficiency requirements. In our study of a combined transshipment and import–export port, we first propose mathematical formulas to characterize the container operation in such a port. Based on these mathematical formulations, a genetic algorithm is presented to optimize the percentage of c...

The quay crane scheduling problem studied in this paper is to determine a handling sequence of holds for quay cranes assigned to a container vessel considering interference between quay cranes. This paper provides a mixed integer programming model for the considered quay crane scheduling problem that is NP-complete in nature. A genetic algorithm is proposed to obtain near optimal solutions. Computational experiments are conducted to examine the proposed model and solution algorithm. The computational results show that the proposed genetic algorithm is effective and efficient in solving the considered quay crane scheduling problem.

Reverse logistics—that is, the distribution activities involved in product returns—is receiving growing attention. Because of the difference and interaction between the forward and reverse distributions, how to integrate the forward and reverse channels has become an emerging issue. Few studies have attempted to integrate the forward and reverse distributions. Furthermore, the increasing opportunities for cost savings and customer satisfaction in such integrated distribution have prompted third-party logistics providers (3PLs) to get involved in forward and reverse logistics operations. This paper first develops a multiobjective model considering activities of the 3PLs for forward and reverse distributions simultaneously. Two objectives are included in the proposed model: (a) maximization of the returned products shipped from customers back to the collection facilities and (b) minimization of the total costs associated with the forward and reverse logistics operations. A fuzzy goal programming approach is applied to determine the compromise solution for the multiobjective model. A genetic algorithm with two subalgorithms is then developed to solve the problem. Numerical experiments are presented to demonstrate the applicability of the formulated model and the proposed solution method.

Reverse logistics—the distribution activities involved in product returns—is receiving growing attention. Because of the differences in and interaction between forward and reverse distribution, figuring out how to integrate the two channels has become an urgent issue. Many companies use inefficient, slow, and expensive processes for handling returned products, and many are either incapable of or unwilling to enter the reverse logistics market. Thus, more and more companies are outsourcing their distribution operations to third-party logistics providers (3PLs). In practice, 3PLs usually serve a number of clients that handle multiple products. Thus, the multiproduct aspect must be considered in any distribution network design for 3PLs. Nonetheless, almost all existing research in this field focuses only on the separate reverse distribution and homogeneous product. This paper addresses an integrated forward and reverse distribution network design problem for 3PLs that involves locating distribution facilities such as warehouses, collection centers, and hybrid warehouse–collection centers and determining the strategy for distributing multiple products between the clients of the 3PLs and the customers via the distribution facilities. A genetic algorithm with two greedy algorithms is used for solution purposes. Computational experiments demonstrate a great deal of promise for this solution method, as high-quality solutions are obtained while expending modest computational effort.

The design of product recovery networks has received growing attention in the past decades. Because of the number of processing activities involved and the variability in the value of the components, these problems present a high level of uncertainty. This paper proposes a stochastic programming-based approach in which a deterministic model for product recovery network design can be extended to account for uncertainties explicitly. Because almost all existing approaches for solving such problems are either restricted to deterministic situations or can deal only with a modest number of scenarios for the uncertain problem parameters, a solution approach integrating a recently proposed sampling method with an acceleration strategy is developed. A computational study involving a large-scale product recovery network is presented to demonstrate the significance of the developed stochastic model as well as the efficiency of the proposed solution approach.

As a key component of bus dwell time, passenger movement time in boarding and alighting processes plays an important role in affecting service reliability and service quality. Boarding and alighting processes are impacted by several factors, including vehicle configuration, friction between boarding and alighting passengers, bus floor level, and fare payment methods. The crowding effect refers to the interaction between passengers boarding, alighting, and on-board. In this paper, we introduce two indices, namely rank and occupancy, to describe the crowding effect. Rank reflects the sequence of passengers boarding and alighting, which reveals how passengers’ movement time changes in boarding/alighting process. Occupancy indicates level of crowdedness in vehicle, and quantifies the friction from on board passengers during boarding and alighting process. Based on time-stamped smart card transactions, we measured the inter-tapping interval between two successive boarding (alighting) passengers to identify the effect of rank and occupancy in determining boarding/alighting processes. Before our analysis, a cluster based outlier detection method is applied to guarantee data quality. Using large quantities of boarding/alighting observations from smart card data, we found that both rank and occupancy have strong effects on boarding and alighting processes. For occupancy, 60% is a critical threshold for boarding process, after which a dramatic increase of average boarding time is observed. This research may help better understand the passenger behavior in boarding and alighting processes.

Autonomous mobility on demand (AMOD) has emerged as a promising solution for urban transportation. Compared to prevailing systems, AMOD promises sustainable, affordable personal mobility through the use of self-driving shared vehicles. Our ongoing research seeks to design AMOD systems that maximize the demand level that can be satisfactorily served with a reasonable fleet size. In this paper, we introduce an extension for SimMobility — a high-fidelity agent-based simulation platform — for simulating and evaluating models for AMOD systems. As a demonstration case study, we use this extension to explore the effect of different fleet sizes and stations locations for a station-based model (where cars self-return to stations) and a free-floating model (where cars self-park anywhere). Simulation results for evening peak hours in the Singapore Central Business District show that the free-floating model performed better than the station-based model with a “small number” of stations; this occurred primarily because return legs comprised “empty” trips that did not serve customers but contributed to road congestion. These results suggest that making use of distributed parking facilities to prevent congestion can improve the overall performance of an AMOD system during peak periods.

This paper proposes an integrated Bayesian statistical inference framework to characterize passenger flow assignment model in a complex metro network. In doing so, we combine network cost attribute estimation and passenger route choice modeling using Bayesian inference. We build the posterior density by taking the likelihood of observing passenger travel times provided by smart card data and our prior knowledge about the studied metro network. Given the high-dimensional nature of parameters in this framework, we apply the variable-at-a-time Metropolis sampling algorithm to estimate the mean and Bayesian confidence interval for each parameter in turn. As a numerical example, this integrated approach is applied on the metro network in Singapore. Our result shows that link travel time exhibits a considerable coefficient of variation about 0.17, suggesting that travel time reliability is of high importance to metro operation. The estimation of route choice parameters conforms with previous survey-based studies, showing that the disutility of transfer time is about twice of that of in-vehicle travel time in Singapore metro system.

Short-term traffic flow prediction has long been regarded as a critical concern for intelligent transportation systems. On the basis of many existing prediction models, each having good performance only in a particular period, an improved approach is to combine these single predictors together for prediction in a span of periods. In this paper, a neural network model is introduced that combines the prediction from single neural network predictors according to an adaptive and heuristic credit assignment algorithm based on the theory of conditional probability and Bayes' rule. Two single predictors, i.e., the back propagation and the radial basis function neural networks are designed and combined linearly into a Bayesian combined neural network model. The credit value for each predictor in the combined model is calculated according to the proposed credit assignment algorithm and largely depends on the accumulative prediction perfor- mance of these predictors during the previous prediction intervals. For experimental test, two data sets comprising traffic flow rates in 15-min time intervals have been collected from Singapore's Ayer Rajah Expressway. One data set is used to train the two single neural networks and the other to test and compare the performances between the combined and singular models. Three indices, i.e., the mean absolute percentage error, the variance of absolute percentage error, and the probability of percentage error, are employed to compare the forecasting performance. It is found that most of the time, the combined model outperforms the singular predictors. More importantly, for a given time period, it is the role of this newly proposed model to track the predictors' performance online, so as to always select and combine the best-performing predictors for prediction.

In past research, several versions of hybrid genetic algorithm-simulation methodology have been proposed for scheduling of multiple lane closures that aims to minimize a network's total traffic delay. The genetic algorithm is used as a search engine for generation of lane closure schedule, while a microscopic traffic simulation model is employed to calculate the total network travel time under each lane closure scenario. A difficulty in implementing this methodology practically is the long computing time required, due to the many simulation runs needed to evaluate the average total network travel time of each feasible schedule. This paper applies the precondition technique, standard error criterion, and termination criterion to reduce the number of necessary simulation runs. As a further improvement, traffic simulations are distributed in different processors of a multiprocessor machine. To further reduce the computing time, a two-stage hybrid genetic algorithm methodology has been proposed in this paper. This two-stage methodology consists of a hybrid genetic algorithm-traffic assignment methodology as the first stage, followed by a hybrid genetic algorithm-distributed simulation methodology as the second stage. The traffic assignment model is used to replace the traffic simulation model in the estimation of total network travel time in stage 1. The applications of the improvement techniques have been demonstrated through a hypothetical problem involving 20 lane closure requests in a network consisting of 986 links, 397 nodes, and 22 origin-destination zones. Together, these improvement techniques contributed to up to 87% reduction in waiting time for a solution of the example problem.

The application of intelligent transportation system technologies to facilitate freight mobility requires dynamic routing decisions to reduce operational costs and enhance service levels. Although most route planning or scheduling methods studied in the literature assume fixed or static travel times for the networks, an attempt is made to focus on the dynamic routing decisions in response to real-time information. A microscopic traffic simulation model coupled with a link-to-link shortest-path search with turning penalties is suggested and developed to evaluate the potential benefits of the dynamic routing decisions, which considers both updated link travel times and the incident conditions detected. The simulation results are described and discussed. Some interesting issues for future research needs are also recommended. This research will help logistics companies assess the benefits of the routing advisory systems that they use for their day-to-day operations.

Application and extensions of a dynamic network equilibrium model to the Advanced Driver and Vehicle Advisory Navigation Concept (ADVANCE) Network are described in this paper. ADVANCE is a dynamic route guidance field test designed for 800 km(superscript 2) in the northwestern suburbs of Chicago. The dynamic route choice model employed in this paper is solved efficiently by a modified version of Janson's DYMOD algorithm. Realistic traffic engineering-based link delay functions, instead of the simplistic Bureau of Public Roads (BPR) function, are used to estimate link travel times and intersection delays for most types of links and intersections. Further, an expanded intersection representation is utilized, resulting in a network of nearly 23,000 links and 10,000 nodes. Time-dependent link flows, travel times, speeds and queue spillbacks are generated for the ADVANCE Network. The model was solved on a CONVEX-C3880. Convergence and computational results are presented and analyzed.

A traffic assignment‐simulation model for planning, rather than operational, purposes is proposed to evaluate what possible benefit might be obtained by adapting a traffic center‐based vehicle route guidance system (VRGS). The model consists of three modules: a multiple driver class traffic assignment module, a traffic signal control module, and a vehicle route guidance system module. Multiple driver class traffic assignment generates the time‐dependent K shortest paths based on “static”, “semi‐dynamic”, and “dynamic” link travel times for each of three classes of driver, namely, non‐equipped drivers, those who are equipped but have partially compliant drivers, and those who are equipped and have totally compliant drivers. Traffic signal settings adopt a table look‐up traffic‐responsive signal scheme to “optimize” signal performance locally. Different VRGS scenarios, which are presented for comparison, are delineated by the proportion of equipped vehicles, the proportion of equipped vehicles in compliance...

The decision of quay crane scheduling problem is crucial for port operators in port container terminals. In this paper, bay-based quay crane scheduling problem is studied and three heuristics designed for the studied problem are identified and their performances (measured both effectiveness and robustness) are examined through a numerical experiment. Herein, the robustness of the solving heuristics is defined as the quality of the methods, as measure by how well the solutions constructed by the methods can offset the adverse impact generated by the randomness of the input. Compared with previous studies in this field, the proposed benchmarking criteria (in terms of both effectiveness and robustness) to evaluate the performance of the heuristics is unique and practical.