Ali Oran

Online map-matching based on Hidden Markov model for real-time traffic sensing applications

In many Intelligent Transportation System (ITS) applications that crowd-source data from probe vehicles, a crucial step is to accurately map the GPS trajectories to the road network in real time. This process, known as map-matching, often needs to account for noise and sparseness of the data because (1) highly precise GPS traces are rarely available, and (2) dense trajectories are costly for live transmission and storage. We propose an online map-matching algorithm based on the Hidden Markov Model (HMM) that is robust to noise and sparseness. We focused on two improvements over existing HMM-based algorithms: (1) the use of an optimal localizing strategy, the variable sliding window (VSW) method, that guarantees the online solution quality under uncertain future inputs, and (2) the novel combination of spatial, temporal and topological information using machine learning. We evaluated the accuracy of our algorithm using field test data collected on bus routes covering urban and rural areas. Furthermore, we also investigated the relationships between accuracy and output delays in processing live input streams. In our tests on field test data, VSW outperformed the traditional localizing method in terms of both accuracy and output delay. Our results suggest that it is viable for low latency applications such as traffic sensing.

Spatiotemporal Patterns in Large-Scale Traffic Speed Prediction

The ability to accurately predict traffic speed in a large and heterogeneous road network has many useful applications, such as route guidance and congestion avoidance. In principle, data-driven methods, such as support vector regression (SVR), can predict traffic with high accuracy because traffic tends to exhibit regular patterns over time. However, in practice, the prediction performance can significantly vary across the network and during different time periods. Insight into those spatiotemporal trends can improve the performance of intelligent transportation systems. Traditional prediction error measures, such as the mean absolute percentage error, provide information about the individual links in the network but do not capture global trends. We propose unsupervised learning methods, such as k-means clustering, principal component analysis, and self-organizing maps, to mine spatiotemporal performance trends at the network level and for individual links. We perform prediction for a large interconnected road network and for multiple prediction horizons with an SVR-based algorithm. We show the effectiveness of the proposed performance analysis methods by applying them to the prediction data of the SVR.

Location and Routing Models for Emergency Response Plans with Priorities

In emergency planning, consideration of emergency priorities is a necessity. This paper presents new formulations of the facility location problem (FLP) and vehicle routing problem with time windows (VRPTW) with considerations of priority. Our models ensure that higher priority locations are considered before the lower priority ones, for both facility and routing decisions. The FLP is solved using an MIP solver, while a tabu search based metaheuristic is developed for the solution of the VRPTW. Under a set of possible emergency scenarios with limited emergency resources, our models were able to serve higher priority locations better than the much utilized Maximal Coverage Location Problem (MCLP) model. We also present preliminary work and results for an integrated location-routing analysis which improves service results further.

Unsupervised learning based performance analysis of n-support vector regression for speed prediction of a large road network

Many intelligent transportation systems (ITS) applications require accurate prediction of traffic parameters. Previous studies have shown that data driven machine learning methods like support vector regression (SVR) can effectively and accurately perform this task. However, these studies focus on highways, or a few road segments. We propose a robust and scalable method using v-SVR to tackle the problem of speed prediction of a large heterogeneous road network. The traditional performance measures such as mean absolute percentage error (MAPE) and root mean square error (RMSE) provide little insight into spatial and temporal characteristics of prediction methods for a large network. This inadequacy can be a serious hurdle in effective implementation of prediction models for route guidance, congestion avoidance, dynamic traffic assignment and other ITS applications. We propose unsupervised learning techniques by employing k-means clustering, principal component analysis (PCA), and self organizing maps (SOM) to overcome this insufficiency. We establish the effectiveness of the developed methods by evaluation of spatial and temporal characteristics of prediction performance of the proposed variable window v-SVR method.

An HMM-based map matching method with cumulative proximity-weight formulation

In this study, we introduce a hidden Markov model map matching method that is based on a cumulative proximity-weight formulation. This new formula is based on the line integral of a point-wise segment weight rather than the almost standard shortest distance based weight. The proposed method was tested using vehicle and map data from Seattle area. Several simulations were conducted so as to have a clear comparison of the new weight to the traditional one; and particular emphasis were given to matching of GPS data with long sampling periods and high level noise. Overall, possible improvements to MM accuracies by the new weight were identified. It was seen that the new weight could be a better option than the shortest distance based weight in the presence of low-frequency sampled and/or noisy GPS data.

A precise proximity-weight formulation for map matching algorithms

With the increased use of satellite-based navigation devices in civilian vehicles, map matching (MM) studies have increased considerably in the past decade. Frequency of the data, and denseness of the underlying road network still dictate the accuracy limits of current MM algorithms. One practical way that can improve the accuracy of most MM approaches is to use more precise weights for the candidate road segments. Because of the geometric nature of the MM problem, proximity-weights have been considered in almost every MM study. However, being formulated through the shortest distance measure, these weights are prone to inaccuracies. We propose a new, more precise, proximity-weight formulation based on a cumulative proximity function which only assumes that the positioning data displays Gaussian distribution errors. Proposed formulations are developed independent of any MM approach, and for this reason they can be used easily under any future MM algorithm.

An Integrated Likelihood Formulation for Characterizing the Proximity of Position Measurements to Road Segments

The analysis of spatial proximity between objects can yield useful insights for a variety of problems. A common application is found in map matching problems, where noisy position measurements collected from a receiver on a network-bound mobile object is analyzed for estimating the original road segments traversed by the object. Motivated by this problem, we take a detailed look at proximity measures that quantify the spatial closeness between points and curves in non-deterministic problems, where the given points are noisy observations of a stochastic process defined on a given set of curves. Starting with a critical review of traditional pointwise approaches, we introduce the integral proximity measure for quantifying proximity, so as to better represent the statistical likelihoods of a process’ states. Assuming a generic stochastic model with additive noise, we discuss the correct proximity function for the proximity measures, and the relationship between a posteriori probabilities of the process and the proximity measures for a comparison of both measures. Later, we prove that the proposed measure can provide better inferences about the process’ states, when the process is under the influence of uncorrelated bivariate Gaussian noise. Finally, we conduct an extensive Monte Carlo analysis, which shows significant inference improvements over traditional proximity measures, particularly under high noise levels and dense road settings.