Chien Hung Wei

Neural network based temporal feature models for short-term railway passenger demand forecasting

Accurate forecasts are the base for correct decisions in revenue management. This paper addresses two novel neural network structures for short-term railway passenger demand forecasting. An idea to render information at suitable places rather than mixing all available information at the beginning in neural network operations is proposed. The first proposed network structure is multiple temporal units neural network (MTUNN), which deals with distinctive input information via designated connections in the network. The second proposed network structure is parallel ensemble neural network (PENN), which deals with different input information in several individual models. The outputs of the individual models are then integrated to obtain final forecasts. Conventional multi-layer perceptron (MLP) is also constructed for comparison purposes. The results show that both MTUNN and PENN outperform conventional MLP in the study. On average, MTUNN can obtain 8.1% improvement of MSE and 4.4% improvement of MAPE in comparison with MLP. PENN can achieve 10.5% improvement of MSE and 3.3% improvement of MAPE in comparison with MLP.

A computerized feature selection method using genetic algorithms to forecast freeway accident duration times

: This study presents a feature selection method that uses genetic algorithms to create two artificial neural network-based models that provide a sequential forecast of accident duration from the time of accident notification to the accident site clearance. These two models can provide the estimated duration time by plugging in relevant traffic data as soon as an accident is notified. To select data feature, the genetic algorithm is designed to decrease the number of model inputs while preserving the relevant traffic characteristics. Using the proposed feature selection method, the mean absolute percentage error for forecasting accident duration at each time point is mostly under 29%, which indicates that these models have a reasonable forecasting ability. Thanks to this model, travelers and traffic management units can better understand the impact of accidents. This study shows that the proposed models are feasible in the Intelligent Transportation Systems context.

Development of Freeway Travel Time Forecasting Models by Integrating Different Sources of Traffic Data

Artificial neural network (ANN) techniques are applied to build a travel time estimation model. The model exhibits a functional relation between real-time traffic data as the input variables and the actual bus travel time as the output variable. A great quantity of traffic data is collected from intercity buses equipped with global positioning systems, vehicle detectors along the roadway, and the incident database. For model development, data from neighboring sections and time intervals are considered to present the time-space relation of traffic. To account for the various methods of specifying freeway sections, four criteria are employed to partition the freeway into comparable units. These are based on interchanges, similar distances, travel times, and geometry. The southern part of the number one national freeway in Taiwan is selected as the case study. In most sections of the four partitions, the mean absolute percentage errors (MAPEs) of the predicted travel time are under 20%, which indicates a good forecasting effect. For practical use purposes, the path travel time is obtained from the section models with a dynamic forecast concept. Through the validation process, the MAPEs of the travel times at each O-D path (from original point to destination point) are known to be mostly under 20%. These results suggest that this dynamic forecasting approach is practical and reliable for modeling travel time characteristics.

Optimal All-Stop, Short-Turn, and Express Transit Services Under Heterogeneous Demand

As a major mode choice of commuters, public transit plays an important role in relieving congestion on urban transportation networks. A model is developed for cost-efficient operation that optimizes all-stop, short-turn, and express transit services and the associated frequencies by considering heterogeneous demand. For a transit line with given stations and origin–destination demand, the objective total cost function is optimized by considering a set of constraints that ensure frequency conservation and sufficient capacity subject to operable fleet size. A numerical example is designed that is based on a real-world rail transit line to demonstrate the applicability and effectiveness of the developed model. Results show that optimized integrated service patterns and associated service frequencies significantly reduce total cost.

A dynamic system-optimum control model for commuting traffic corridors

Abstract The need to implement effective traffic control in commuting corridors has long been recognized by transportation professionals. However, most existing focused either on optimizing freeway ramp metering rates or providing coordinated surface street signals, without taking account of the vital interaction between these two subsystems. As a result, it is not unusual that an effective control strategy for freeway operation may cause significant detrimental effects to the adjacent surface streets. On the other hand, the access to freeway ramps is often impeded by the formation of congestion or bottlenecks on surface streets due to the increasing peak-period traffic demand and ineffective signal operation. This paper presents a dynamic system-optimal control model (DSOCM) for commuting corridors which consist of both freeway and surface street segments. The proposed DSOCM considers the complex interactions among the freeway, surface street and diversion flows, and allows the system operators to compute the optimal time-dependent ramp metering rate and signal setting over the selected time horizon. Depending on the input reliability, DSOCM need not be executed at every control interval as long as the differences between the projected and actual traffic conditions are within the acceptable range. An effective and coordinated control operation for integrated traffic systems can then be achieved.

Analysis of artificial neural network models for freeway ramp metering control

Abstract Traffic along a freeway varies not only with time but also with space. It is thus essential to model dynamic traffic patterns on the freeway in order to derive appropriate metering control strategies. Existing methods cannot fulfill this task effectively. Due to the learning capability, artificial neural network models are developed to simulate typical time series traffic data and then expanded to capture the inherent time–space interrelations. The augmented-type network is proposed that includes several basic modules intelligently affiliated according to traffic characteristics on the freeway. Inputs to neural network models are traffic states in each time period on the freeway segments while outputs correspond to the desired metering rate at each entrance ramp. The simulation outcomes indicate very encouraging achievements when the proposed neural network model is employed to govern the freeway traffic operations. Also discussed are feasible directions for further improvements.

VEHICLE CLASSIFICATION USING ADVANCED TECHNOLOGIES

Applying advanced technologies to existing problem domains is a highly desirable approach in many research areas. Among these techniques, image processing has been shown useful in transportation fields for such tasks as traffic pattern recognition, data collection, accident detection, and pavement evaluation. The integrated model with artificial neural networks (ANNs) has promising potential applications. The image processing and ANN model are combined to explore the feasibility of vehicle classification in real-world situations. Three methods were developed during the research process: ground segmentation, background subtraction, and window segmentation. The first two methods were used to separate the objects of scene and nonscene from the actual traffic image. To reduce the complexity of neural networks, the image was divided into 16 windows and three characteristics (occupation rates of vehicles, of horizontal image lines, and of vertical image lines) of each window were extracted to generate 48 factor...

Joint Optimization of Temporal Headway and Differential Fare for Transit Systems Considering Heterogeneous Demand Elasticity

This study presents an approach to jointly optimize temporal service headway and differential fare for an intercity transit system considering heterogeneous demand elasticity. The research optimization problem is combinatorial and difficult to solve analytically. A genetic algorithm is developed to search for the optimal solution, including temporal headway and differential fare, which maximizes the daily profit considering service capacity sufficiency and operable fleet size. A real world case study, the Taiwan High Speed Rail Line, is applied to demonstrate the applicability of the developed model and the efficiency of the solution algorithm to search for the optimal solution that yields the maximum profit operation. Results from the sensitivity analyses indicate that the optimized number of partitions of travel distance and headway decreases as the demand increases, which results in the increase of profit. The developed model can be applied to evaluate and plan an intercity rail system.

Integrated transit services for minimum cost operation considering heterogeneous demand

Abstract In large metropolitan areas, public transit is a major mode choice of commuters for their daily travel, which has an important role in relieving congestion on transportation corridors. The purpose of this study is to develop a model which optimizes service patterns (SPs) and frequencies that yield minimum cost transit operation. Considering a general transit route with given stops and origin-destination demand, the proposed model consists of an objective total cost function and a set of constraints to ensure frequency conservation and sufficient capacity subject to operable fleet size. A numerical example is provided to demonstrate the effectiveness of the developed model, in which the demand and facility data of a rail transit route were given. Results show that the proposed model can be applied to optimize integrated SPs and headways that significantly reduce the total cost, while the resulting performance indicators are generated.

Comprehensive Performance Evaluation Framework for Urban Transport Policies: Integration of Investment, Pricing, Regulation, and Subsidy Options

This study describes a framework that is used to analyze urban transport policies in Kaohsiung, the second largest city in Taiwan. The framework is comprehensive in the sense that the full range of transport-related policy elements—investment, pricing, regulation, and subsidies—are explicitly integrated to generate policy alternatives for evaluation. In contrast, past planning processes have tended to focus solely on infrastructure investment. The relationships among these four policy elements were examined in the case study. The results show that an integrated, goal-related transport policy with the four elements is more effective than the traditional, single-element policy. The proposed policy evaluation framework can be implemented in the commonly used urban transportation planning process.