Dynamic Modeling of Lane Flow and Density Distribution in Congested Freeways

This research aims to verify the hypothesis that lane traffic distribution initiates traffic regimes occurrence in freeways, and in particular to unveil the inter-dependence of lane vehicle allocation and congestion formation. The developed methodological framework consists of (i) a clustering approach to separate traffic regimes and (ii) a multi-regime modeling approach to identify the critical traffic variables that influence the traffic distribution in freeway lanes. The prediction of regime occurrence, through lane-related parameters, could be implemented in control algorithms, in order to ameliorate their efficiency by timely activating a policy. A neural network algorithm is invoked, in order to detect in an unbiased approach homogeneous groups of the prevailing traffic regimes during peak periods, and a denoised input space. Based on the optimized clusters of traffic states and peak periods and on the rationale that congestion propagates from left to right in the onset of peak period indicating a transition to another state, lane density distribution ratio (LDDR) and density of the right or left lane, depending on the state, are introduced as promising determinant response variables. LDDR appears to be site-independent and is intermittently occurring during congestion and free flow conditions. To efficiently forecast stream dynamics, their parameterization is effectuated with lane-related variables. Static and dynamic regression models are deployed for congested or saturated regimes, for free flow regimes and also for the entirety of traffic states. Static models are formed, pursuing a simple approach that would ensure feasibility of implementation. Statistical assessment indicates that the prediction models yield significant prediction accuracy.

Dynamics of Lane Distribution Flows in Freeways

André-Gilles Dumont, Nikolaos Geroliminis, Sofia Samoili

Lane density and lane flow distribution are utilised in the current study as characteristics of traffic dynamics for control and planning purposes in freeways with ITS systems. As part of an on-going research, the Lane Flow Distribution Ratio (LFDR) and the Lane Density Distribution Ratio (LDDR) were introduced as main scalability factors for congestion evolution estimation. The preliminary exploratory analysis from LFDR, LDDR, traffic and spatiotemporal parameters along with link fundamental diagrams, indicated dependence between lane vehicle allocation and impending congestion emergence, in terms of traffic levels. This led to a partial parameterisation of driver’s behaviour during peak hours, that was studied in several contexts, namely at mainline, on-ramp and off-ramp locations. The assessment of the developed model was conducted based on observations derived from a freeway site in California. The site is operating during peak hours with high congestion level. Further on this approach, decision thresholds for a number of ITS policies could be derived, to adapt accordingly to control policies (eg. hard shoulder operation, VSL, ramp metering).

2013

Dynamic Modeling Of Lane Distribution Impact In Complex Freeways

André-Gilles Dumont, Nikolaos Geroliminis, Sofia Samoili

Traffic dynamics have been the focal point of research aiming to provide a better understanding of traffic phenomena and to be integrated in traffic management for recurrent congestion. The aim of the study is to ameliorate complex highway and/or freeway networks performance by forecasting traffic regimes based on lane vehicle allocation, traffic characteristics and their synergies on impending traffic conditions for control and modeling purposes. In order to define the prevailing traffic regimes, model-based Bayesian clustering was employed on flow-density relationships for exploratory data mining. In addition, the stochastic cluster analysis was invoked to uncover the time span of morning and evening peak hours, refining and reducing the input data space and enhancing noise reduction. On the contrary to distance-based clustering, in a preliminary phase, two explicit homogeneous groups (congested/not congested regimes for morning and evening peak hours) were probabilistically determined via the Bayesian propositional logic, regarding the number of existing regimes as well as the peak hours period. With this approach, a metric cluster definition was avoided, as two hypotheses were formed, represented by two models. The null hypothesis H_0:κ=1, with κ the true unknown number of clusters, is examined versus the hypothesis H_1: κ=k, with k a pre-specified number of clusters. Assuming that each model is equally likely to be expressed, the posterior probability of null hypothesis was estimated, thus the Bayesian factor (BF) included, with small values signalizing evidence against H_0. To approximate the BF, Markov chain Monte Carlo (MCMC) techniques were evoked. Partitioned sampling from the mixture distribution of traffic flow and density was obtained through the Metropolis-Hastings algorithm, so as to constraint the draws to high probability areas. Based on the optimised reduced data space of the respective regimes, piecewise/segmented modeling was introduced to address the parameterization of network dynamics. Piecewise models were studied to delineate stream patterns, as they are additive models that inherit linear models properties, enhancing thus description’s flexibility. Lane vehicle allocation and spatiotemporal variables were provided as representative parameters of driver’s behaviour (Table 1), where lddrR is the lane density distribution ratio of the right lane, lddrL the lane density distribution ratio of the left lane, ΔVMR-ML the speed difference between median right and median left lane, ΔVMR-R the speed difference between median right and right lane, ΔVML-L the speed difference between median left and left lane, and kR, kL, kMR, kML the density of the right, left, median right and median left lanes respectively. Initially, static models were developed, for both peak and off peak hours, with the lane density distribution ratio (LDDR) as response variable, as well as the differences of speed per adjacent and non-adjacent lanes and lane density. The proposed models indicated strong precision, particularly for off-peak hours. However, potentially, the fluctuating behaviour within traffic states transitions during peak hours, led to a dynamic modeling approach with time-related parameters. The persistence on lane scale yielded significant improvement on accuracy, confirming the aforementioned presumption. The models’ assessment was conducted with empirical data from a complex freeway network, with successive on and off ramps and four to six lanes per direction in the mainline (Fig. 1), that features pronounced congested condition during peak hours. The study site is located in district 11 of San Diego County, U.S. (I5-N and I5-S). The traffic dataset was consisted of aggregated measurements (intervals of Δt=5 min), of flow, speed and occupancy per lane, and it is part of the Caltrans Performance Measurement System (PeMS). Further on this on-going study, several types of networks will be evaluated either with ITS control policies or deprived, aiming to assess the stationarity of the static and dynamic models.

2014

Stochastic Modeling of Lane Distribution and Dynamic Control of Managed Lanes for Highways

Sofia Samoili

Recurrent congestion during increasingly extended daily peak periods is an escalating phenomenon with multidimensional impact.Congestion mitigation in highways is polarized between invasive infrastructure interventions and Intelligent Transportation Systems (ITS) that promote ameliorated performance with sustainable economic and spatiotemporal requirements.Nevertheless, traffic intra-class variability, adaptability of driversâbehavior to ITS and inter-lane behavior variation challenge the strategiesâperformance and advocate congestion that could be anticipated by timely operation of designated control policies.In this scope, a novel multi-level algorithm is introduced that provides unbiased definition of peak periods and prevailing traffic regimes, through stochastic clustering; forecasts lane traffic distribution in congested and uncongested traffic regimes through lane-scale parameterisation; optimises managed lanes (ML) systems control through multi-objective functions with traffic and economic interdependencies, so as to balance LOS and operational costs.In the first level, separate stochastic clustering procedures capture spatial patterns of lane stream dynamics and temporal patterns of time span during which max capacity is attained, which unbiasedly define prevailing traffic regimes and peak periods.Data mining reveals underlying spatial association between lane traffic distribution and traffic regimes emergence that is assessed in the subsequent level, through spatiotemporal parameterization per lane.Multivariate modeling is integrated to anticipate the sequence between separate regimes, as ensued by clustering, namely to capture patterns of lanes vehicle allocation during free flow and congested regimes, and forecast impending traffic behavior that could proactively trigger the efficient implementation of control policies.Static models are developed, to ensure simple feasible implementation to reactive control management systems, and dynamic models for integration into real-time proactive systems A novel introduced parameter, the lane density distribution ratio (LDDR), and the density of the destination-lane for congested conditions or of the origin-lane for uncongested, are addressed as promising determinant response variables Both are proven site-independent and occur intermittently during congestion and free flow conditions At the lower level, multi-objective optimisation of a MLâs system operation is conducted, on account of the explanatory variables of the developed forecasting models and the operational costs of such systems, so as to ensure timely operation of a policy and so congestion alleviation A reactive ML system is subject to the proposed optimisation scheme, where observed underutilization of the ML, undermines the systemâs performance The integrated approach is assessed based on the efficiency of the designated control policy in preventing congested conditions, and it concludes with the set of Pareto optimal operation thresholds appointment, through maximisation of throughput per lane and minimisation of operational costs The procedure is considered innovative, as relevant frameworks are not acknowledged in literature for ML systems, and decisions for their timely operation are solely empirically driven The proposed algorithm is integrated at a reactive hard shoulder running (HSR) system, and is evaluated through a developed API and simulation.Finally,discrete choice ordered probit models of usersâ adaptation a

EPFL2015

Stochastic Modeling of Lane Distribution and Dynamic Control of Managed Lanes for Highways

Sofia Samoili

EPFL2015

Dynamic Modeling Of Lane Distribution Impact In Complex Freeways

André-Gilles Dumont, Nikolaos Geroliminis, Sofia Samoili

2014