This article discusses how the artificial neural network (ANN) is one advance approach to freeway travel time prediction. Various studies using different inputs have not come to consensus on the effects of input selections. In addition, very little discussion has been made on the temporal–spatial aspect of the ANN travel time prediction process. In this study, the authors employ an ANN ensemble technique to analyze the effects of various input settings on the ANN prediction performances. Volume, occupancy, and speed are used as inputs to predict travel times. The predictions are then compared against the travel times collected from the toll collection system in Houston, Texas. The results show speed or occupancy measured at the segment of interest may be used as the sole input to produce acceptable predictions, but all three variables together tend to yield the best prediction results. The inclusion of inputs from both upstream and downstream segments is statistically better than using only the inputs from current segment. It also appears that the magnitude of prevailing segment travel time can be used as a guideline to set up temporal input delays for better prediction accuracies. The evaluation of spatiotemporal input interactions reveals that past information on downstream and current segments is useful in improving prediction accuracy whereas past inputs from the upstream location do not provide as much constructive information. Finally, a variant of the state-space model (SSNN), namely time-delayed state-space neural network (TDSSNN), is proposed and compared against other popular ANN models. The comparison shows that the TDSSNN outperforms other networks and remains very comparable with the SSNN. Future research is needed to analyze TDSSNN's ability in corridor prediction settings.

Development of Recurrent Neural Network Considering Temporal-Spatial Input Dynamics for Freeway Travel Time Modeling

Tracking congestion throughout the network road is a critical component of Intelligent transportation network management systems. Understanding how the traffic flows and short-term prediction of congestion occurrence due to rush-hour or incidents can be beneficial to such systems to effectively manage and direct the traffic to the most appropriate detours. Many of the current traffic flow prediction systems are designed by utilizing a central processing component where the prediction is carried out through aggregation of the information gathered from all measuring stations. However, centralized systems are not scalable and fail provide real-time feedback to the system whereas in a decentralized scheme, each node is responsible to predict its own short-term congestion based on the local current measurements in neighboring nodes. 
We propose a decentralized deep learning-based method where each node accurately predicts its own congestion state in real-time based on the congestion state of the neighboring stations. Moreover, historical data from the deployment site is not required, which makes the proposed method more suitable for newly installed stations. In order to achieve higher performance, we introduce a regularized Euclidean loss function that favors high congestion samples over low congestion samples to avoid the impact of the unbalanced training dataset. A novel dataset for this purpose is designed based on the traffic data obtained from traffic control stations in northern California. Extensive experiments conducted on the designed benchmark reflect a successful congestion prediction.

Scalable Deep Traffic Flow Neural Networks for Urban Traffic Congestion Prediction

Increasing mobility and congestion results in an increase in travel time variability and in a decrease in reliability. Reliability becomes an important performance measure for transportation facilities. A variety of performance measures have been proposed to quantify it. Many of these indicators are based on percentiles of travel time. The knowledge of the distribution of travel time is needed to properly estimate these values. Congestion distorts the distribution and particular statistical distributions are needed. Different distributions have been proposed in the literature. In a previous paper, we presented a comparison of six statistical distributions used to model travel time. These six distributions are the Lognormal, Gamma, Burr (extended by Singh-Maddala), Weibull, a mixture of two Normal distributions and a mixture of two Gamma distributions. In this paper a probabilistic modeling of travel time which takes into account the levels-of-service is given. Levels of service are identified, then travel time distributions are modeled by level of service. This results in a very good fit between the empirical and modeled distributions Moreover, the adjustment was improved, thanks to the calibration of “Bureau of Public Roads” functions, linking the travel time to the traffic flow by level of service. The superiority of the Singh-Maddala distribution appears in many cases. This has been validated, thanks to travel time data from the same site at another period. However the parameters of the distributions vary from one year to another, due to changes in infrastructure. The transferability of the approach, not performed, will be based on travel time data on another site.

Estimating Travel Time Distribution under different Traffic conditions

Tracking congestion throughout the network road is a critical component of Intelligent transportation network management systems. Understanding how the traffic flows and short-term prediction of congestion occurrence due to rush-hour or incidents can be beneficial to such systems to effectively manage and direct the traffic to the most appropriate detours. Many of the current traffic flow prediction systems are designed by utilizing a central processing component where the prediction is carried out through aggregation of the information gathered from all measuring stations. However, centralized systems are not scalable and fail provide real-time feedback to the system whereas in a decentralized scheme, each node is responsible to predict its own short-term congestion based on the local current measurements in neighboring nodes. We propose a decentralized deep learning-based method where each node accurately predicts its own congestion state in realtime based on the congestion state of the neighboring stations. Moreover, historical data from the deployment site is not required, which makes the proposed method more suitable for newly installed stations. In order to achieve higher performance, we introduce a regularized euclidean loss function that favors high congestion samples over low congestion samples to avoid the impact of the unbalanced training dataset. A novel dataset for this purpose is designed based on the traffic data obtained from traffic control stations in northern California. Extensive experiments conducted on the designed benchmark reflect a successful congestion prediction.

Scalable deep traffic flow neural networks for urban traffic congestion prediction

Despite recent development efforts in many other advanced traffic surveillance systems, inductance loops are still the most widely used traffic detectors. This paper points out the deficiency of point measurements obtained from loop detectors in estimating travel times under congested traffic conditions and proposes a theoretically sound and practically applicable travel time estimation algorithm that uses the same loop detector data. The main idea of the algorithm is based on the concept of section density that can be easily obtained by observing in-and-out traffic flows between two point detection stations, with careful correction for detector errors. Travel times estimated from the proposed method are compared to those of other methods via both simulated and real traffic data. While the method that estimates travel time based on spot speeds tends to underestimate section travel times due to its inability to capture traffic congestion between detector stations, the proposed section-density based method provides fairly accurate travel time estimates.

Section Travel Time Estimation from Point Detection Data - eScholarship

Travel time is an important parameter in evaluating the operating efficiency of traffic networks, in assessing the performance of traffic management strategies, and as input to many intelligent transportation systems applications such as advanced traveler information systems Travel time can be obtained directly from instrumented test vehicles, license plate matching, probe vehicles etc, or from indirect methods such as inductance loop detectors Because of the widespread deployment of loop detectors, they are one of the most widely used inputs to travel time estimation techniques There are different methods available to calculate the travel time from loop detector data, such as extrapolation of the point speed values, statistical methods, and models based on traffic flow theory However, most of these methods fail during the transition period between the normal and congested flow conditions The present study proposes several modifications to an existing traffic flow theory based model for travel time 

Lellitha D. Vanajakshi

Papers

Improved Flow-Based Travel Time Estimation Method from Point Detector Data for Freeways