Optimal Trajectory Planning Algorithm for Connected and Autonomous Vehicles towards Uncertainty of Actuated Traffic Signals

Shafik, Amr; Eteifa, Seifeldeen; Rakha, Hesham A. · 2023 · ROSA P / Morgan State University. Urban Mobility & Equity Center

archive: archived pipeline: cataloged verified

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

Summary

This study addresses the challenge of optimizing vehicle trajectories for Connected and Autonomous Vehicles (CAVs) approaching actuated traffic signals, where signal switching times are uncertain. While eco-driving systems can significantly reduce fuel consumption by smoothing speed profiles, existing methods often assume deterministic signal timing or rely on computationally complex models impractical for real-time application. The authors aim to develop a robust Green Light Optimal Speed Advisory (GLOSA) system that minimizes fuel consumption despite the stochastic nature of Signal Phasing and Timing (SPaT) data from actuated signals. The research specifically investigates how uncertainty in SPaT predictions impacts achievable fuel savings and establishes benchmarks for the reliability required in prediction models to ensure effective eco-driving. The methodology extends a deterministic vehicle control algorithm into a stochastic optimal control framework. The system utilizes Dynamic Programming (DP) incorporated with an A* algorithm to efficiently compute the least-cost trajectory, defined as minimum fuel consumption. The model accounts for vehicle dynamics, including tractive and resistance forces, and imposes constraints to prevent collisions, red-light violations, and excessive jerk (limited to 1.3 m/s³ for passenger comfort). A risk assessment procedure ensures safety by enforcing maximum deceleration when the vehicle enters a critical stopping distance. The study evaluates three scenarios: a baseline of uninformed drivers using empirical field data, a deterministic scenario with exact signal timing, and a stochastic scenario where switching times are sampled from normal distributions with varying bias and standard deviation. Simulations were conducted on a single-lane approach with a 3% grade, comparing the optimized trajectories against the baseline. The results demonstrate that the proposed stochastic GLOSA system yields significant fuel savings compared to uninformed drivers. On average, the system achieved 37% fuel savings under deterministic SPaT conditions and 30% under stochastic conditions. A sensitivity analysis revealed that the system remains effective even with prediction errors; specifically, the algorithm can achieve 85% of the maximum possible fuel savings if the timing error is within ±3.3 seconds at a 95% confidence level. The study further highlights that the reliability of SPaT predictions becomes increasingly critical as the time to green decreases relative to the vehicle's time to reach the intersection. The findings provide clear benchmarks for practitioners, indicating the necessary accuracy levels for statistical or machine learning models predicting signal switching times to realize substantial energy benefits.

Key finding

The proposed stochastic trajectory planning algorithm achieved average fuel savings of 37% for deterministic signal information and 30% for stochastic signal information compared to uninformed drivers.

Methodology

simulator

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. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

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

• situational awareness

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Theoretical Contribution: computational model