A Hybrid Coordinated Decision-Making Method for CAVs at Unsignalized Intersection

He, Shan; Adouane, Lounis · 2024 · Crossref

DOI: 10.23919/ecc64448.2024.10591063

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the challenge of coordinating Connected Autonomous Vehicles (CAVs) at unsignalized intersections, aiming to reduce urban traffic congestion and improve safety without relying on traditional traffic signals. The authors propose a hybrid coordinated decision-making method that combines two algorithms: Multi-Risk Management Cooperative Optimization based on Predicted Inter-Distance Profile (MRMCO-PIDP) and the Epsilon Probability Collective algorithm (Epsilon-PC). The motivation stems from the limitations of existing methods; while probabilistic collective algorithms like Epsilon-PC can find high-quality solutions, they suffer from long computation times due to inefficient exploration. Conversely, local search methods are fast but may yield only locally optimal results. The proposed hybrid approach leverages the speed of MRMCO-PIDP to quickly generate a feasible, collision-free solution, followed by Epsilon-PC to refine this solution into a potentially better one within the remaining negotiation time. The methodology utilizes the Predicted Inter-Distance Profile (PIDP) to assess collision risks by projecting the distance between vehicles over a future time horizon. The MRMCO-PIDP algorithm evaluates potential speed profiles (acceleration or deceleration) by calculating the difference between the minimum predicted distance and a safety threshold ($ePIDP$). It employs a multi-risk management strategy to resolve conflicting acceleration/deceleration requirements from multiple interacting vehicles. The optimization process involves iterative negotiation where CAVs propose joint strategies, selecting the best combination that minimizes an objective function balancing safety, crossing speed, and collision penalties. The hybrid architecture prioritizes MRMCO-PIDP for rapid initial solution generation, using its output as the baseline for the subsequent, more computationally intensive Epsilon-PC search. Simulations were conducted in MATLAB using randomly generated scenarios involving 3, 4, and 5 vehicles. The results demonstrate that MRMCO-PIDP significantly outperforms Epsilon-PC in terms of computation speed, requiring only about 4% of the processing time in some cases. While Epsilon-PC occasionally achieves slightly lower average crossing times due to its ability to find better global solutions, it requires substantially more iterations and time to converge. MRMCO-PIDP proved more stable and efficient, particularly as the number of vehicles increased. The study concludes that the hybrid approach is optimal for real-time applications: MRMCO-PIDP ensures a safe, feasible solution is found quickly, while Epsilon-PC improves the solution quality if sufficient negotiation time remains. This method effectively balances the trade-off between computational efficiency and solution optimality for complex intersection scenarios.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• gap acceptance