Modeling Route Choice Behavior using Multi-Agent Reinforcement Learning

 Suggested by: Johannes Scholz 

Keywords: Geo-Semantic Search, Natural Language Processing, Spatial Data, Semantic Indexing, GIS, Information Retrieval, GeoAI, Knowledge Graphs Route Choice Modeling, Multi-Agent Systems, Reinforcement Learning, Traffic Simulation, Decision-Making, Intelligent Transportation, Agent-Based Modeling, Dynamic Routing 

 

Objective: Develop a multi-agent reinforcement learning model to simulate route choice behavior, where agents make real-time link-by-link navigation decisions based on perceived traffic conditions and past interactions with the environment and other agents. 

 

Short Description: This thesis explores the application of multi-agent reinforcement learning to model route choice behavior in dynamic traffic environments. Unlike traditional route assignment models, this approach enables agents to adapt their paths in real time based on personal experiences and observed congestion patterns. The research aims to develop and evaluate a reinforcement learning framework that improves traffic flow prediction and decision-making in intelligent transportation systems. 

 

Start: 

As soon as possible 

 

Prerequisites/qualification: 

• Knowledge in GIS and spatial data processing 

• Interest/Knowledge in Agent-based Modelling