Edouard Ivanjko

Advancements in technology and the emergence of vehicle-to-everything communication encourage new research approaches. Continuously sharing data through the onboard unit, connected vehicles (CVs) have proven to be a valuable source of real-time microscopic traffic data. Utilizing CVs as mobile sensors is a key driver for traffic safety improvement and increasing the effective operative road capacity. Data obtained from CVs can be effectively processed using speed transition matrices (STMs) while preserving spatial and temporal characteristics. This research proposes a new approach to adaptive traffic signal control utilizing STMs and a cooperative multi-agent learning system for the environment of CVs. To confirm its effectiveness, the concept is tested in a simulated environment of an intersection network, comparing different CVs’ penetration rates and cooperation coefficients between agents.

Traffic assignment in large urban areas is an old but increasingly important problem because of the rapid growth of the world population and traffic demands. Many algorithms have been developed but their convergence rates and complexities are still prohibitive for real-time applications. The recently developed MinSum algorithm introduces a new approach. It is a highly efficient discrete-domain optimization algorithm for system-optimized route assignment between two city zones. Its complexity (the number of critical operations) is O(R³), where R is the number of routes. Nonetheless, there is still room for improvements, and this paper presents two modified MinSum variants, heuristic and approximate, that are significantly faster and of lower complexity, while retaining MinSum’s prominent features. Heuristic variant MinSumH is up to five times faster than MinSum and its complexity theoretically remains O(R³), though experiments indicate that it is closer to O(R²). Approximate variant MinSumA is faster by up to over 100 times and reduces the complexity to O(R). Both proposed variants are progressively faster as R grows. Due to their high convergence rate and exceptionally low complexity, along with other prominent features, the proposed algorithms are ready for real-time system-optimal traffic assignment in a real urban environment.

The machine learning-based approaches for analysing the mobility needs of users are currently the most prevalent approach in the mobility-on-demand (MoD) analysis. Their efficiency relies on the comprehensiveness and consistency of training datasets. However, this is also the biggest challenge, as high-quality training data are often difficult to obtain. Thus, the Variational Autoencoders (VAE) are investigated as potential generators of synthetic samples for the augmentation of MoD-based datasets. This MoD-based dataset is created using real-world taxi trip data recorded in the Manhattan district of New York City, USA. This augmentation by synthetic samples can potentially enable larger, balanced, and more consistent datasets for machine learning analysis of MoD-based data. The proposed VAE approaches are compared with common dimensionality reduction techniques and standard autoencoders concerning their efficiency in 2-dimensional clustering based on collected MoD-based data. The proposed 2-dimensional convolution VAE framework has achieved clustering results comparable with the other analysed approaches. Thus, it generates synthetic samples, known as “deepfakes”. They are added in different percentages to the initial dataset based on real-world MoD-based data. Thus, this creates augmented datasets of the initial one. The models for predicting the cluster of each sample are used to evaluate the impact of those augmented datasets on their accuracy and learning convergence compared to the initial dataset. Results have shown that the accuracy and learning convergence are improved if those predictive models are trained on an augmented dataset which includes up to 10% of synthetic samples for each cluster.

Frequent traffic jams are a common problem in large urban areas and despite real-time operation, contemporary traffic control (i.e., navigation) systems do not fully succeed in solving this problem. This situation motivates traffic assignment research aimed at optimal usage of infrastructure, that is, minimization of travel time. However, achieving this goal in real-time requires a departure from traditional approach and an enhancement of the predictive and preventive abilities of control and navigation systems. To this aim, the paper presents an efficient solution for an optimized route assignment given the set of available routes between two city zones. Unlike other methods, mostly working on link (graph edge) level, the proposed solution works on route (graph path) level and is based on estimation of the expected (average) travel times. This enables the application of two efficient algorithms for optimization of vehicle distribution according to two important criteria, that is, both Wardrop’s principles. The most prominent characteristics of the proposed algorithms are linear complexity with respect to the number of vehicles (traffic flow), linear and quadratic complexity with respect to the number of routes for user equilibrium and system optimization, respectively, simplicity, and absence of any numerical issues, which makes them suitable for real-time traffic management. The theoretical results have been confirmed by diverse evaluations using synthetic models and in a case study using the city of Zagreb as a real world example. The proposed concept is universal, inexpensive, and quickly implementable as an upgrade to the existing traffic control or navigation systems.

Mixed traffic flows are opening up new areas for research and are seen as key drivers in the field of data and services that will make roads safer and more environmentally friendly. Understanding the effects of Connected Vehicles (CVs) and Connected Autonomous Vehicles (CAVs), as one of the vehicle components of mixed traffic flows, will make it easier to avoid traffic congestion and contribute to the creation of innovative applications and solutions. It is notable that the literature related to the analysis of the impact of mixed traffic flows on traffic signal control in urban areas rarely considers mixed traffic flow containing CVs, CAVs, and Human Driven Vehicles (HDVs). Therefore, this paper provides an overview of the relevant research papers covering the topic of urban Traffic Signal Control (TSC) and mixed traffic flows. Best practices for intersection state estimation and TSC in the case of mixed traffic flows in an urban environment are summarized and possible approaches for utilizing CVs and CAVs as mobile sensors and actuators are discussed.