Data Mining - Concepts and Techniques.

Highway Capacity Manual改訂の動向--テイラ-教授の講演より

Neural networks have been extensively applied to short-term traffic prediction in the past years. This study proposes a novel architecture of neural networks, Long Short-Term Neural Network (LSTM NN), to capture nonlinear traffic dynamic in an effective manner. The LSTM NN can overcome the issue of back-propagated error decay through memory blocks, and thus exhibits the superior capability for time series prediction with long temporal dependency. In addition, the LSTM NN can automatically determine the optimal time lags. To validate the effectiveness of LSTM NN, travel speed data from traffic microwave detectors in Beijing are used for model training and testing. A comparison with different topologies of dynamic neural networks as well as other prevailing parametric and nonparametric algorithms suggests that LSTM NN can achieve the best prediction performance in terms of both accuracy and stability.

Long short-term memory neural network for traffic speed prediction using remote microwave sensor data

Since the early 1980s, short-term traffic forecasting has been an integral part of most Intelligent Transportation Systems (ITS) research and applications; most effort has gone into developing methodologies that can be used to model traffic characteristics and produce anticipated traffic conditions. Existing literature is voluminous, and has largely used single point data from motorways and has employed univariate mathematical models to predict traffic volumes or travel times. Recent developments in technology and the widespread use of powerful computers and mathematical models allow researchers an unprecedented opportunity to expand horizons and direct work in 10 challenging, yet relatively under researched, directions. It is these existing challenges that we review in this paper and offer suggestions for future work.

Short-term traffic forecasting: Where we are and where we’re going

Parallel control and management have been proposed as a new mechanism for conducting operations of complex systems, especially those that involved complexity issues of both engineering and social dimensions, such as transportation systems. This paper presents an overview of the background, concepts, basic methods, major issues, and current applications of Parallel transportation Management Systems (PtMS). In essence, parallel control and management is a data-driven approach for modeling, analysis, and decision-making that considers both the engineering and social complexity in its processes. The developments and applications described here clearly indicate that PtMS is effective for use in networked complex traffic systems and is closely related to emerging technologies in cloud computing, social computing, and cyberphysical-social systems. A description of PtMS system architectures, processes, and components, including OTSt, Dyna CAS, aDAPTS, iTOP, and TransWorld is presented and discussed. Finally, the experiments and examples of real-world applications are illustrated and analyzed.

Parallel Control and Management for Intelligent Transportation Systems: Concepts, Architectures, and Applications

The connected vehicle research program is a multimodal research initiative in the U.S. that envisions a fully connected transportation system with wireless communications linking vehicles, the infrastructure, and handheld smart devices. This paper designs and evaluates a reservation-based approach to intersection control that is designed to take full advantage of the unprecedented connectivity that the connected vehicle initiative promises to provide. The control approach, which is referred to herein as the “intelligent intersection” approach, builds on the previous work by Dresner and Stone by introducing new features to better account for several aspects of the real-world driving environment. To design and evaluate the “intelligent intersection,” a novel simulation test bed for connected vehicle applications is developed. The test bed integrates a microscopic traffic simulator with a network simulator and an emission analyzer. Using the integrated simulator, the mobility and environmental benefits of the intelligent intersection approach, compared with those of traditional control methods, are evaluated on two case studies: 1) an isolated intersection and 2) a real-world transportation network with multiple intersections. Results show that the proposed control approach offers significant mobility and environmental benefits. For example, for the second test case and using observed traffic volumes, the intelligent intersection reduced the average vehicle delay by 85%, fuel consumption by 50%, and emissions by 39%-50%. The study also demonstrates the utility of using the simulator test bed in the design and evaluation of connected vehicle applications.

Assessing the Mobility and Environmental Benefits of Reservation-Based Intelligent Intersections Using an Integrated Simulator

This book presents the essential information necessary for the successful planning of intelligent transportation systems (ITS). The book is appropriate for upper-level undergraduate and undergraduate courses, and also serves as a practical reference for transportation operations/planning practitioners. Chapters address the fundamentals of traffic flow and control, ITS user services, ITS applications and benefits, ITS architecture, ITS planning, ITS standards, ITS evaluation and ITS challenges and opportunities.

Fundamentals of Intelligent Transportation Systems Planning

This paper presents a framework for real-time highway traffic condition assessment using vehicle kinetic information, which is likely to be made available from vehicle-infrastructure integration (VII) systems, in which vehicle and infrastructure agents communicate to improve mobility and safety. In the proposed VII framework, the vehicle onboard equipment and roadside units (RSUs) collaboratively work, supported by an artificial intelligence (AI) paradigm, to determine the occurrence and characteristics of an incident. Two AI paradigms are examined: 1) support vector machines (SVMs) and 2) artificial neural networks (ANNs). Each RSU then assesses the traffic condition based on the information from multiple vehicles traveling on its supervised highway segment. As a case study, this paper developed a model of the VII-SVM framework and evaluated its performance in a microscopic traffic simulation environment for a highway network in Spartanburg, SC. The performance of the VII-SVM was compared with the performance of the corresponding VII-ANN framework, and both frameworks were found to be capable of classifying the travel experience using the kinetic data generated by each vehicle. The performance of the VII-SVM framework, in terms of its detection rate, false-alarm rate, and detection times, was also found to be superior to a baseline California-type incident-detection algorithm. Moreover, the framework provided additional information, including an estimate of the incident location and the likely number of lanes blocked, which will be helpful for implementing an appropriate response strategy. The proposed VII-AI framework thus provides a reliable alternative to traditional traffic sensors in assessing traffic conditions.

Real-Time Highway Traffic Condition Assessment Framework Using Vehicle–Infrastructure Integration (VII) With Artificial Intelligence (AI)

With the availability of large volumes of real-time traffic flow data along with traffic accident information, there is a renewed interest in the development of models for the real-time prediction of traffic accident risk. One challenge, however, is that the available data are usually complex, noisy, and even misleading. This raises the question of how to select the most important explanatory variables to achieve an acceptable level of accuracy for real-time traffic accident risk prediction. To address this, the present paper proposes a novel Frequent Pattern tree (FP tree) based variable selection method. The method works by first identifying all the frequent patterns in the traffic accident dataset. Next, for each frequent pattern, we introduce a new metric, herein referred to as the Relative Object Purity Ratio (ROPR). The ROPR is then used to calculate the importance score of each explanatory variable which in turn can be used for ranking and selecting the variables that contribute most to explaining the accident patterns. To demonstrate the advantages of the proposed variable selection method, the study develops two traffic accident risk prediction models, based on accident data collected on interstate highway I-64 in Virginia, namely a k -nearest neighbor model and a Bayesian network. Prior to model development, two variable selection methods are utilized: (1) the FP tree based method proposed in this paper; and (2) the random forest method, a widely used variable selection method, which is used as the base case for comparison. The results show that the FP tree based accident risk prediction models perform better than the random forest based models, regardless of the type of prediction models (i.e. k -nearest neighbor or Bayesian network), the settings of their parameters, and the types of datasets used for model training and testing. The best model found is a FP tree based Bayesian network model that can predict 61.11% of accidents while having a false alarm rate of 38.16%. These results compare very favorably with other accident prediction models reported in the literature.

Adel W. Sadek

Papers

Assessing the Mobility and Environmental Benefits of Reservation-Based Intelligent Intersections Using an Integrated Simulator

Fundamentals of Intelligent Transportation Systems Planning

Real-Time Highway Traffic Condition Assessment Framework Using Vehicle–Infrastructure Integration (VII) With Artificial Intelligence (AI)

A novel variable selection method based on Frequent Pattern tree for real-time traffic accident risk prediction

A combined M5P tree and hazard-based duration model for predicting urban freeway traffic accident durations.