Modeling

Car-following (CF) algorithms are crucial components of traffic simulations and have been integrated into many production vehicles equipped with Advanced Driving Assistance Systems (ADAS). Insights from the model of car-following behavior help researchers to understand the causes of various macro phenomena that arise from interactions between pairs of vehicles. Car-following Models encompass multiple disciplines, including traffic engineering, physics, dynamic system control, cognitive science, machine learning, deep learning, and reinforcement learning. This paper presents an extensive...

In the past decade, motivated by the putative failure of naive self-play deep reinforcement learning (DRL) in adversarial imperfect-information games, researchers have developed numerous DRL algorithms based on fictitious play (FP), double oracle (DO), and counterfactual regret minimization (CFR). In light of recent results of the magnetic mirror descent algorithm, we hypothesize that simpler generic policy gradient methods like PPO are competitive with or superior to these FP, DO, and CFR-based DRL approaches. To facilitate the resolution of this hypothesis, we implement and release the...

The research described in this paper employed simulation modeling and empirical observations in an effort to: (1) Identify the traffic flow phenomena that characterize freeway weaving section capacity; and (2) determine appropriate traffic flow rate values that reflect weaving section capacity. The INTRAS microscopic simulation model was calibrated and validated using empirical data collected at a weaving site. Increasing traffic demands were then sequentially input into repeated simulation runs to identify the boundary between uncongested and congested operation. Where a weaving...

This paper explores the application of ordered probit modeling, an econometric technique commonly used for the analysis of rating data in situations when respondents are asked to rate items (e.g., an object, service, or product), one at a time. To demonstrate the application of this methodology, an existing data set, originally collected to measure the perceived complexities for an array of Blissymbols, was used. Findings from the use of the ordered probit model are compared with those resulting from the earlier research that used regression procedures. Results from the use of ordered...