Comparison of gap-based and flow-based control strategies using a new controlled stochastic cellular automaton model for traffic flow

Abstract

Autonomous vehicles are widely considered an imperative element of future transportation systems, and their adoption is expected to reduce traffic congestion. What is an effective vehicle control to reduce it? This study explores the impact of different vehicle control strategies on traffic flow through simulations. To achieve this, we introduce a novel stochastic traffic flow model called the controlled stochastic optimal velocity (CSOV) model, which incorporates vehicle control effects. Two distinct control strategies are applied within the CSOV model: gap-based control (GC) and flow-based control (FC). The GC strategy regulates the velocity to equalize the gap between the front and rear vehicles. Conversely, the FC strategy regulates the velocity to maintain a consistent front and rear flow rate. The results show that there were density regions where the GC strategy improved the flow rate. However, only the GC strategy with the weak control resulted in a lower flow rate compared to when there was no control. Conversely, the FC strategy consistently improved the flow rate regardless of control strength, yielding more robust results. Furthermore, when the two controls achieve similar flow rates, the FC strategy provided a more robust velocity distribution for density change than the GC strategy.