Cooperative On-Ramp Merging with Time-Varying Vehicle-to-Vehicle Communication Delay Compensation via a Model-Free Approach

Cooperative merging strategies enabled by vehicle-to-vehicle (V2V) communication have shown promise in addressing congestion, fuel inefficiency, and collision risks. However, their performance can be severely degraded by time-varying and uncertain communication delays-an issue often overlooked in existing research, which primarily focuses on merging sequence determination and trajectory planning. Furthermore, practical considerations such as heterogeneous vehicle dynamics, varying road conditions, and real-time implementation complexities are frequently neglected. This paper presents a model-free, online planning framework for cooperative on-ramp merging of connected and automated vehicles (CAVs), explicitly accounting for time-varying V2V communication delays. Without relying on detailed vehicle dynamics, the proposed method introduces a data-driven delay compensation scheme. A co-simulation platform integrating high-fidelity vehicle dynamics, traffic simulation (SUMO), and V2V communication within MATLAB/Simulink is developed to evaluate the proposed method. Simulation results demonstrate that unaddressed V2V communication delays significantly impair merging performance. In contrast, the proposed framework enhances intervehicle distance tracking and maintains low CO2 emissions and fuel consumption, under communication delay across different communication frequencies. In conclusion, its lightweight design also facilitates real-time implementation, making it well-suited for deployment in practical CAV systems.. Authors: Khan, Muhammad Waleed [Univ. of Texas at Dallas, Richardson, TX (United States)]; Zhou, Anye [Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)] (ORCID:0000000301455579); Cook, Adian [Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)] (ORCID:0000000160825395); Beck, Joe [Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)] (ORCID:0000000342571138); Summers, Tyler [Univ. of Texas at Dallas, Richardson, TX (United States)]. DOE Contract: AC05-00OR22725. Subjects: Accounting; Autonomous vehicles; Carbon dioxide; Collisions (Physics); Delays; Design; Emissions; Fuel consumption; Fuels; Matlab; Roads; Simulation; Simulation results; Traffic; Trajectories (Physics); V2V communication; Vehicle dynamics; Vehicles; cooperative merging; delay compensation; model predictive control; model-free control