Ying Lee

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.

A Computerized Feature Reduction Using Cluster Methods for Accident Duration Forecasting on Freeway

This study creates two Artificial Neural Network-based models and provides a sequential forecast of accident duration from the accident notification to the accident site clearance. With these two models, the estimated duration time can be provided by plugging in relevant traffic data as soon as an accident is notified. To reduce data feature, cluster method can decreases the number of model inputs and preserves the relevant traffic characteristics with fewer inputs. This study shows proposed models are feasible ones in the Intelligent Transportation Systems (ITS) context.

Evaluating the Effects of Highway Traffic Accidents in the Development of a Vehicle Accident Queue Length Estimation Model

Accurate estimation of vehicle accident queue lengths can assist traffic managers in providing improved plans for traffic control. Few studies that have analyzed the influence of highway incidents on traffic flow have discussed the length of vehicle queues during incidents. Therefore, the present study applied artificial neural network and regression methods to construct models for estimating the length of vehicle accident queues. The factors influencing the occurrence of accidents are numerous and complex. To estimate vehicle accident queue lengths, the models incorporated various data from accidents, including accident characteristics, traffic data, illumination, weather conditions, and road geometry characteristics. All raw data were collected from two public agencies and were integrated and cross-checked. Before model development, a correlation analysis was performed to reduce the scale of interrelated features or variables. In the model evaluation, the mean absolute percentage error of the most suitable model was below 50%, indicating that the proposed model and procedure provide a reasonable estimate of vehicle accident queue lengths.

Non-parametric machine learning methods for evaluating the effects of traffic accident duration on freeways

Traffic accidents usually cause congestion and increase travel-times. The cost of extra travel time and fuel consumption due to congestion is huge. Traffic operators and drivers expect an accurately forecasted accident duration to reduce uncertainty and to enable the implementation of appropriate strategies. This study demonstrates two non-parametric machine learning methods, namely the k-nearest neighbour method and artificial neural network method, to construct accident duration prediction models. The factors influencing the occurrence of accidents are numerous and complex. To capture this phenomenon and improve the performance of accident duration prediction, the models incorporated various data including accident characteristics, traffic data, illumination, weather conditions, and road geometry characteristics. All raw data are collected from two public agencies and were integrated and cross-checked. Before model development, a correlation analysis was performed to reduce the scale of interrelated features or variables. Based on the performance comparison results, an artificial neural network model can provide good and reasonable prediction for accident duration with mean absolute percentage error values less than 30%, which are better than the prediction results of a k-nearest neighbour model. Based on comparison results for circumstances, the Model which incorporated significant variables and employed the ANN method can provide a more accurate prediction of accident duration when the circumstances involved the day time or drunk driving than those that involved night time and did not involve drunk driving. Empirical evaluation results reveal that significant variables possess a major influence on accident duration prediction.