DynamicRouteGPT: A Real-Time Multi-Vehicle Dynamic Navigation Framework Based on Large Language Models
Ziai Zhou, Bin Zhou, Hao Liu
TL;DR
DynamicRouteGPT tackles real-time multi-vehicle dynamic navigation by fusing a global baseline path derived from Dijkstra with a GPT-driven local decision layer guided by Bayesian inference and Markov-chain traffic modeling. The framework comprises Real-Time Information Acquisition, Alternative Path Generation, GPT Decision-Making, and Dynamic Path Adjustment modules, implemented on Llama3 8B with LLAMA-Factory, BAdam optimization, and LoRA fine-tuning, achieving adaptable, globally informed route planning without pre-training. Experimental results in SUMO across multiple networks show state-of-the-art performance in travel time and waiting-time reductions, along with explainable path choices and strong generalization to new road networks and varying penetration rates. The work offers a practical, scalable approach for intelligent transportation systems, providing a new baseline for dynamic path planning under real-time conditions while highlighting computational and data-quality considerations for real-world deployment.
Abstract
Real-time dynamic path planning in complex traffic environments presents challenges, such as varying traffic volumes and signal wait times. Traditional static routing algorithms like Dijkstra and A* compute shortest paths but often fail under dynamic conditions. Recent Reinforcement Learning (RL) approaches offer improvements but tend to focus on local optima, risking dead-ends or boundary issues. This paper proposes a novel approach based on causal inference for real-time dynamic path planning, balancing global and local optimality. We first use the static Dijkstra algorithm to compute a globally optimal baseline path. A distributed control strategy then guides vehicles along this path. At intersections, DynamicRouteGPT performs real-time decision-making for local path selection, considering real-time traffic, driving preferences, and unexpected events. DynamicRouteGPT integrates Markov chains, Bayesian inference, and large-scale pretrained language models like Llama3 8B to provide an efficient path planning solution. It dynamically adjusts to traffic scenarios and driver preferences and requires no pre-training, offering broad applicability across road networks. A key innovation is the construction of causal graphs for counterfactual reasoning, optimizing path decisions. Experimental results show that our method achieves state-of-the-art performance in real-time dynamic path planning for multiple vehicles while providing explainable path selections, offering a novel and efficient solution for complex traffic environments.