.png){kind=icon}
.png){kind=icon}
Return
Network Generation Artificial Intelligence: Multimodal Road Network Generation Based on Large Language Models
.png){kind=icon}
PRE
AI
DOI：10.1177/03611981251357933.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Traffic simulation is essential for traffic-related research, but its use demands significant time and cognitive effort for road network generation. The burgeoning popularity of large language models (LLMs), which highlight powerful communication and logical reasoning capabilities, provides a promising avenue for the intelligentization of road network generation. Despite this, LLMs need help with domain-specific knowledge, particularly in addressing complex issues within transportation. This paper proposes the network generation artificial intelligence (NGAI) framework that integrates the reasoning capabilities of LLMs with traditional road network generation methods, significantly enhancing the intelligent operability of transportation simulation. Based on multiple inquiry experiments, NGAI has been proven capable of selecting suitable traffic network generation models (TNGMs) and generating content based on specified parameters. In the experimental scenarios of this study, TNGMs demonstrated extremely high selection accuracy and a low probability of repeated invocations under detailed prompts. The effective use of NGAI has significantly reduced the cost of road network generation and optimized the steps for users employing simulation software, making the process of transportation simulation simpler and more intelligent.
Keywords:
data and data science
artificial intelligence and advanced computing applications
machine vision
planning and analysis
transportation network modeling
simulation modeling
Journal
T
IF:
1.8
Papers:
854
Citations:
3.4W