Online Learning Solutions for Freeway Travel Time Prediction

Abstract

Abstract-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.Index Terms-Advanced traffic information systems (ATIS), extended Kalman filter, online learning, recurrent neural networks, state space neural networks, traffic information, travel time prediction.

Experienced travel time prediction for congested freeways

“…These methods can be broadly classified in two major categories; parametric methods (e.g. linear regression (Zhang and Rice, 2003), time series models (Yang, 2005;Min and Wynter, 2011), Kalman filtering (Okutani and Stephanedes, 1984;Van Lint, 2008)) and non-parametric methods (neural network models (Ledoux, 1997;Vlahogianni et al, 2005;Van Lint, 2006), support vector regression (Vanajakshi and Rilett, 2007), simulation models (Liu et al, 2006)). In the past years, neural network models have gained attention in transportation field and are frequently applied in traffic state prediction.…”

Modelling the impact of traffic incidents on travel time reliability

“…Figure 2.1 illustrates the factors influencing variations in travel time. Some studies addressed major sources of large variability in travel time including traffic incidents, work zones, adverse weather, traffic control, special events, road geometry and fluctuation in normal traffic (Cambridge Systematics Inc. and Texas Transportation Institute, 2005;Emam and Al-Deek, 2006;Hendren et al, 2006;Van Lint et al, 2008;Transportation Research Board, 2009). Factors causing fluctuations in either demand or supply might have inter-relationships and are not independent.…”

“…Various types of travel time models have been developed for the purpose of estimating and predicting travel time and, consequently, estimating variability of travel time (Zhang and Rice, 2003;Van Lint, 2004;Jiann-Shiou, 2005;Li, 2006;Liu et al, 2006;You and Kim, 2007;Van Lint, 2008;Fei et al, 2011).…”

Mitigating Congestion by Integrating Time Forecasting and Realtime Information Aggregation in Cellular Networks

“…Later, a dynamic route guidance system (DRGS) [74] was proposed for estimating and predicting travel time and then weighted to get optimal routes. In order to further improve the performance of the prediction model, researchers have applied various techniques to this problem such as Bayesian framework [75], Linear regression [76], Kalman filtering [77], neural networks [78], and kernel estimator [79].…”

“…In 1973, Messer, Dudek and Friebele [73] announced the possibility of time prediction in their work. After that, many researchers [75][76][77][78][79] have concentrated their efforts on improving the accuracy of the predicted time. [61,62].…”