Traffic State Estimation on a Two-Dimensional Network by a State-Space Model

Probe vehicle-based traffic state estimation method with spacing information and conservation law

Probe vehicle data are increasingly becoming more attractive for real-time system state estimation in transportation networks. This paper presents analytical models for the real-time estimation of queue lengths at traffic signals using the fundamental information (i.e., location and time) that probe vehicles provide. For a single queue with Poisson arrivals, analytical models are developed to evaluate how error changes in queue length estimation as the percentage of probe vehicles in the traffic stream varies. When the overflow queue is ignored, a closed-form solution is obtained for the variance of the estimation error. For the more general case with the overflow queue, a formulation for the error variance is presented, which requires the marginal probability distribution of the overflow queue as the input. In addition, an approximate model is presented for the latter case, which yields results that are comparable with the exact solution. Overall, the formulations presented here can be used to assess the error in queue length estimation from probe data without conducting simulation runs for various scenarios of probe vehicle market-penetration rates and congestion levels.

Traffic state estimation by backward moving observers: An application and validation under an incident

Urban traffic networks comprise a combination of various links. These networks are complicated as they have numerous intersections, meaning that using an analytical approach or parametric models to estimate driving speeds on arterial or highway roads results in low accuracy. In this study, a model is developed to estimate the link speed using speed data collected by a probe vehicle driven across different urban traffic links, which have interrupted flows. We discover multimodal distributions of travel speeds in each link’s probe vehicle data and use them to separate the vehicle groups and calculate the mean speed of links. This strategy makes it possible to obtain more detailed data, which are used to determine the traffic state and increase the accuracy of the model. The developed nonlinear model, suitable for low correlations of consecutive links’ speed data, is built on a recurrent neural network. Moreover, this study merges three machine-learning techniques to apply low correlations between link properties and speed states. The model developed lowered the mean absolute error by 35.9% on average when compared with the long short-term memory with raw data: 46.8% for the slow state, 55.7% for the state change, and 48.0% for the sudden change over 10 km/h.

23 - A Study on the Performance of Probe Vehicle Data

Methods of estimating dynamic origin–destination (O-D) matrices for urban networks from probe vehicle data are explored. A speed–density function is derived and fitted for different types of roads with the use of the maximum likelihood method. Both a Bayesian method that carefully incorporates prior information and an ordinary method are used to estimate link flows from probe vehicle speed. A dynamic traffic assignment–based bilevel generalized least squares (GLS) estimator considering the distance between the estimated and target O-D matrices as well as the distance between the calculated and observed link flows is formulated to estimate dynamic O-D matrices from estimated link flows. In the iterative solution procedure, the upper level is solved with the extended Bell algorithm, and the microscopic dynamic traffic assignment system VISSIM is applied to produce the assignment matrix in the lower level. A medium-sized signalized network in Tokyo is modeled in a case study in which Bayesian and ordinary methods are compared both in link flow estimation and O-D matrix estimation. Further, the bilevel GLS estimator and bilevel ordinary least squares estimator are implemented and then compared in O-D estimation. The results validate the proposed bilevel GLS estimator.

On urban arterials, travel time estimation is challenging especially from various data sources. Typically, fusing loop detector data and probe vehicle data to estimate travel time is a troublesome issue while considering the data issue of uncertain, imprecise and even conflicting. In this paper, we propose an improved data fusing methodology for link travel time estimation. Link travel times are simultaneously pre-estimated using loop detector data and probe vehicle data, based on which Bayesian fusion is then applied to fuse the estimated travel times. Next, Iterative Bayesian estimation is proposed to improve Bayesian fusion by incorporating two strategies: 1) substitution strategy which replaces the lower accurate travel time estimation from one sensor with the current fused travel time; and 2) specially-designed conditions for convergence which restrict the estimated travel time in a reasonable range. The estimation results show that, the proposed method outperforms probe vehicle data based method, loop detector based method and single Bayesian fusion, and the mean absolute percentage error is reduced to 4.8%. Additionally, iterative Bayesian estimation performs better for lighter traffic flows when the variability of travel time is practically higher than other periods.

Road link traffic speed pattern mining in probe vehicle data via soft computing techniques

Traffic Monitoring immediately after a major natural disaster as revealed by probe data – A case in Ishinomaki after the Great East Japan Earthquake

As vehicle-to-vehicle and vehicle-to-infrastructure communications technologies are evolving, data from vehicles equipped with location and wireless technologies provide new opportunities to observe traffic flow dynamics more precisely. A new methodology is presented for estimating the dynamics of vehicular queues at signalized intersections on the basis of the event data generated by probe vehicles. The methodology uses shock wave theory (i.e., the Lighthill–Whitham–Richards theory) to estimate the evolution of the back of the queue over time and space from the event data generated when probe vehicles join the back of the queue. The time and space coordinates of these events are used to develop a formulation for determining the critical points needed to create time–space diagrams or shock waves to characterize the queue dynamics. The methodology is applied to sample data generated from the microscopic traffic simulation software VISSIM. It is found that the proposed methodology is effective in estimating queue dynamics at traffic signals.

Recent advances in information and communication technologies enable us to collect large spatio-temporal data of road traffic and to provide continuous, specific, historical, and wide-area datasets. Probe vehicle data are one of the most widely used passive datasets that can potentially observe wide ranges in the network, as they do not require fixed sensors. Such datasets potentially observe various phenomena which are caused by various reasons. However, it is not easy to detect significant incidents from such phenomenon due to the multi-dimensional form of the dataset and the large size required to scan the entire dataset. Incident detection is one of the significant approaches for monitoring of traffic flow and data cleansing to acquire “typical traffic situation” for analysis of measures and calibrations of models. Several previous studies have attempted to determine the local changes in the traffic flow. However, a few studies investigated a method to explore the significant incidents influencing wide-ranges of the spatio-temporal dimension which are possibly caused by the changes on either or both the service and demand side. This study proposes a data mining method to detect unusual situations in large-scale road network from a large volume of probe vehicle dataset. The method employs a nonparametric probabilistic function and the Markov random field to consider the spatio-temporal correlations in the probe vehicle dataset. The method is applied to an actual dataset provided by 40000 commercial vehicles all over Japan.

Development and evaluation of arterial incident detection models using fusion of simulated probe vehicle and loop detector data

Accurate estimation of travel times from single-loop detectors

Network-wide traffic state estimation using a mixture Gaussian graphical model and graphical lasso

The purposes of this study were to investigate accuracy and reliability of travel time information and to assess the effectiveness of probe vehicle in providing traffic information (travel time) under various traffic conditions. A widely used traffic microsimulation model, Paramics V5, was used to generate travel time data on a hypothetical network that was vuilt according to geometrics of corridor, traffic flow parameters and configuration of signalized intersection. Average travel time on links and routes were analyzed for various percentage of probe vehicles (market penetration) and traffic flow conditions (level of congestion) and compared to the ‘true’ average travel time. Moreover, the variation of travel time data obtained from probe vehicles was examined from ‘bootstrapping’ technique where this method of resampling represents the varying condition for probe vehicles. It was found that the plots of probe vehicle average travel time versus ‘tree’ average travel time for link and route case agreed well which imply the obtainment of travel time accuracy by probe vehicles in wide range of traffic volumes (situations). However, the probe vehicle performed less accurately when traffic congestion was forming. A case study of network and traffic condition similar to a real world condition was carried out to demonstrate the quality of travel time data from probe vehicles in selected routes (OD). It was clear that when having all vehicles in the link to be probes, the OD estimates are rather rlilable.