No abstract

A hybrid model for predicting urban arterial travel time on the basis of so-called state-space neural networks (SSNNs) and the extended Kalman filter (EKF) is presented. Previous research demonstrated that SSNNs can address complex nonlinear spatiotemporal problems. However, SSNN models require off-line training with large sets of input–output data, presenting three main drawbacks: ( a) great amounts of time and effort are involved in collecting, preparing, and executing these training sessions; ( b) as the input–output mapping changes over time, the model requires complete retraining; and ( c) if a different input set becomes available (e.g., from inductive loops) and the input–output mapping has to be changed, then retraining the model is impossible until enough time has passed to compose a representative training data set. To improve SSNN effectiveness, the EKF is proposed to train the SSNN instead of conventional approaches. Moreover, this network topology is derived from the urban travel time prediction problem. Instead of treating the neural network as a “black-box” model, the design explicitly reflects the relationships that exist in physical traffic systems. It allows the interpretation of neuron weights and structure in terms of the inherent mechanism of the network process with clear physical meaning. Model performance was tested on a densely used urban arterial in the Netherlands. Performance of this proposed model is compared with that of two existing models. Results of the comparisons indicate that the proposed model predicts complex nonlinear urban arterial travel times with satisfying effectiveness, robustness, and reliability.

Compared with the ring-barrier framework used for ring structures (or phasing plans) in signalized control of intersections in the United States, the Dutch framework has no explicit barriers, but only a requirement to respect pairwise conflicts. This paper describes how ring structures can be modeled with pairwise conflicts as a starting point. Modeling techniques were extended to account for offset constraints such as leading pedestrian intervals in which the start or end of one traffic movement was constrained by the start or end of another one that was otherwise compatible. One practical drawback of the more flexible Dutch framework is that it permits so many more possible ring structures that it can be prohibitive to evaluate them all manually. Therefore, this paper describes VRIGEN, an automated method that overcomes this drawback by identifying and evaluating all possible ring structures. Finally, this paper presents several examples in which barrier-free ring designs allow signals to cycle more quickly and efficiently, with improvements in safety and delay for pedestrians and bicyclists. Most of these examples feature pedestrian phases that are allowed to overlap while their parent vehicular phases are not.

When making trips in urban environments, cyclists lose time as they stop and idle at signalized intersections. The main objective of this study was to show how augmenting the situational awareness of traffic signal controllers, using observations from moving sensor platforms, can enable prioritization of cyclists and reduce lost time within the control cycle in an effective way. We investigated the potential of using observations from connected autonomous vehicles (CAVs) as a source of new information, using a revised vehicle-actuated controller. This controller exploits CAV-generated observations of cyclists to optimize the control for cyclists. The results from a simulation study indicated that with a low CAV penetration rate, prioritizing cyclists by tracking reduced cyclist delays and stops, even with a small field of view. As the delay of car directions were not taken into account in this study, the average car delay increased considerably with an increasing number of cyclists. Future work is needed to optimize the control that balances the delays and stops of cyclists and cars.