Accurate and timely traffic flow information is important for the successful deployment of intelligent transportation systems. Over the last few years, traffic data have been exploding, and we have truly entered the era of big data for transportation. Existing traffic flow prediction methods mainly use shallow traffic prediction models and are still unsatisfying for many real-world applications. This situation inspires us to rethink the traffic flow prediction problem based on deep architecture models with big traffic data. In this paper, a novel deep-learning-based traffic flow prediction method is proposed, which considers the spatial and temporal correlations inherently. A stacked autoencoder model is used to learn generic traffic flow features, and it is trained in a greedy layerwise fashion. To the best of our knowledge, this is the first time that a deep architecture model is applied using autoencoders as building blocks to represent traffic flow features for prediction. Moreover, experiments demonstrate that the proposed method for traffic flow prediction has superior performance.

Traffic Flow Prediction With Big Data: A Deep Learning Approach

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

For the last two decades, intelligent transportation systems (ITS) have emerged as an efficient way of improving the performance of transportation systems, enhancing travel security, and providing more choices to travelers. A significant change in ITS in recent years is that much more data are collected from a variety of sources and can be processed into various forms for different stakeholders. The availability of a large amount of data can potentially lead to a revolution in ITS development, changing an ITS from a conventional technology-driven system into a more powerful multifunctional data-driven intelligent transportation system (D2ITS) : a system that is vision, multisource, and learning algorithm driven to optimize its performance. Furthermore, D2ITS is trending to become a privacy-aware people-centric more intelligent system. In this paper, we provide a survey on the development of D2ITS, discussing the functionality of its key components and some deployment issues associated with D2ITS Future research directions for the development of D2ITS is also presented.

Data-Driven Intelligent Transportation Systems: A Survey

Accurate and real-time traffic forecasting plays an important role in the intelligent traffic system and is of great significance for urban traffic planning, traffic management, and traffic control. However, traffic forecasting has always been considered an “open” scientific issue, owing to the constraints of urban road network topological structure and the law of dynamic change with time. To capture the spatial and temporal dependences simultaneously, we propose a novel neural network-based traffic forecasting method, the temporal graph convolutional network (T-GCN) model, which is combined with the graph convolutional network (GCN) and the gated recurrent unit (GRU). Specifically, the GCN is used to learn complex topological structures for capturing spatial dependence and the gated recurrent unit is used to learn dynamic changes of traffic data for capturing temporal dependence. Then, the T-GCN model is employed to traffic forecasting based on the urban road network. Experiments demonstrate that our T-GCN model can obtain the spatio-temporal correlation from traffic data and the predictions outperform state-of-art baselines on real-world traffic datasets. Our tensorflow implementation of the T-GCN is available at https://www.github.com/lehaifeng/T-GCN .

T-GCN: A Temporal Graph Convolutional Network for Traffic Prediction

Accuracy and robustness with respect to missing or corrupt input data are two key characteristics for any travel time prediction model that is to be applied in a real-time environment (e.g. for display on variable message signs on freeways). This article proposes a freeway travel time prediction framework that exhibits both qualities. The framework exploits a recurrent neural network topology, the so-called state-space neural network (SSNN), with preprocessing strategies based on imputation. Although the SSNN model is a neural network, its design (in terms of input- and model selection) is not “black box” nor location-specific. Instead, it is based on the lay-out of the freeway stretch of interest. In this sense, the SSNN model combines the generality of neural network approaches, with traffic related (“white-box”) design. Robustness to missing data is tackled by means of simple imputation (data replacement) schemes, such as exponential forecasts and spatial interpolation. Although there are clear theoretical shortcomings to “simple” imputation schemes to remedy input failure, our results indicate that their use is justified in this particular application. The SSNN model appears to be robust to the “damage” done by these imputation schemes. This is true for both incidental (random) and structural input failure. We demonstrate that the SSNN travel time prediction framework yields good accurate and robust travel time predictions on both synthetic and real data.

Accurate freeway travel time prediction with state-space neural networks under missing data

In recent years, travel time reliability has become one of the key performance indicators of transportation networks and corridors around the globe. Travel time reliability indicators are mostly related to properties of the day-to-day travel time distribution on for example a freeway corridor. On the basis of empirical data a number of key characteristics of this day-to-day distribution can be identified. Most importantly, this distribution is not only very wide but also heavily skewed. The (economic) consequences of this skew are substantial. For example, it is shown that in some peak periods the 5% most “unlucky drivers” incur almost five times as much delay as the 50% most fortunate travelers. We argue this implies first of all that (besides the variance of travel times) skew must be considered an important contributing factor to travel time unreliability. Secondly, it suggests that most of currently used unreliability measures (which are predominantly based on travel time variance), should be used and interpreted with some reservations, since they only account for a part of the costs (that is, delays) of unreliability. This is further substantiated by a comparison on the basis of empirical data from a densely used freeway in the Netherlands between a new travel time reliability measure based on both width and skew, and a number of travel time reliability measures commonly used in practice. The analysis clearly illustrates the inconsistency between all measures, both old and new. In illustration, in cases where the commonly used misery index dubs a particular departure period very unreliable, another common measure (buffer time) considers these periods relatively reliable. Although without objective and quantitative criteria (e.g. economic or societal costs) a choice for any of these measures in road network performance analyses will remain subject to debate, this article provides empirically underpinned arguments to prefer measures incorporating the skew of the travel time distribution.

Travel time unreliability on freeways: Why measures based on variance tell only half the story

An approach to freeway travel time prediction based on recurrent neural networks is presented. Travel time prediction requires a modeling approach that is capable of dealing with complex nonlinear spatio-temporal relationships among flows, speeds, and densities. Based on the literature, feedforward neural networks are a class of mathematical models well suited for solving this problem. A drawback of the feed-forward approach is that the size and composition of the input time series are inherently design choices and thus fixed for all input. This may lead to unnecessarily large models. Moreover, for different traffic conditions, different sizes and compositions of input time series may be required, a requirement not satisfied by any feedforward data-driven method. The recurrent neural network topology presented is capable of dealing with the spatiotemporal relationships implicitly. The topology of this neural net is derived from a state-space formulation of the travel time prediction problem, which is in l...

Freeway Travel Time Prediction with State-Space Neural Networks: Modeling State-Space Dynamics with Recurrent Neural Networks:

Providing travel time information to travelers on available route alternatives in traffic networks is widely believed to yield positive effects on individual drive behavior and (route/departure time) choice behavior, as well as on collective traffic operations in terms of, for example, overall time savings and-if nothing else-on the reliability of travel times. As such, there is an increasing need for fast and reliable online travel time prediction models. Previous research showed that data-driven approaches such as the state-space neural network (SSNN) are reliable and accurate travel time predictors for freeway routes, which can be used to provide predictive travel time information on, for example, variable message sign panels. In an operational context, the adaptivity of such models is a crucial property. Since travel times are available (and, hence, can be measured) for realized trips only, adapting the parameters (weights) of a data-driven travel time prediction model such as the SSNN is particularly challenging. This paper proposes a new extended Kalman filter (EKF) based online-learning approach, i.e., the online-censored EKF method, which can be applied online and offers improvements over a delayed approach in which learning takes place only as realized travel times are available.

/pdf/online-learning-solutions-for-freeway-travel-time-prediction-2hpfvf1m7m.pdf

Online Learning Solutions for Freeway Travel Time Prediction

Freeway traffic state estimation and prediction are central components in real-time traffic management and information applications. Model-based traffic state estimators consist of a dynamic model for the state variables (e.g., a first- or second-order macroscopic traffic flow model), a set of observation equations relating sensor observations to the system state (e.g., the fundamental diagrams), and a data-assimilation technique to combine the model predictions with the sensor observations [e.g., the extended Kalman filter (EKF)]. Commonly, both process and observation models are formulated in Eulerian (space-time) coordinates. Recent studies have shown that this model can be formulated and solved more efficiently and accurately in Lagrangian (vehicle number-time) coordinates. In this paper, we propose a new model-based state estimator based on the EKF technique, in which the discretized Lagrangian Lighthill-Whitham and Richards (LWR) model is used as the process equation, and in which observation models for both Eulerian and Lagrangian sensor data (from loop detectors and vehicle trajectories, respectively) are incorporated. This Lagrangian state estimator is validated and compared with a Eulerian state estimator based on the same LWR model using an empirical microscopic traffic data set from the U.K. The results indicate that the Lagrangian estimator is significantly more accurate and offers computational and theoretical benefits over the Eulerian approach.

J.W.C. van Lint

Papers

Accurate freeway travel time prediction with state-space neural networks under missing data

Travel time unreliability on freeways: Why measures based on variance tell only half the story

Freeway Travel Time Prediction with State-Space Neural Networks: Modeling State-Space Dynamics with Recurrent Neural Networks:

Online Learning Solutions for Freeway Travel Time Prediction

Real-Time Lagrangian Traffic State Estimator for Freeways