我國人工智慧車聯網之號誌控制(1/2)-都會區幹道實作與交流道區域模式發展探討

本計畫延續111年「我國人工智慧車聯網之號誌控制模式探討」之研究計畫，進行與縣市政府合作之現場實驗測試，實作測試過程中視需求持續精進人工智慧號誌控制模式，透過強化學習通用環境之探討與建置、導入多目標強化學習演算法、以及探討車流模擬測試模型於各種車流場景之泛用性，期望能提升人工智慧號誌控制模式之運作效率。

This project continues the research project entitled *Discussion on Signal Control Modes for Artificial Intelligence-Enabled Internet of Vehicles* from the 111th year of the ROC calendar. It carries out on-site experimental tests in collaboration with county and municipal governments. During the implementation and testing stages, the artificial intelligence signal control mode is continuously refined in line with practical requirements. By exploring and constructing general reinforcement learning environments, introducing multi-objective reinforcement learning algorithms, and investigating the generalizability of traffic flow simulation test models across diverse traffic scenarios, this project seeks to enhance the operational efficiency of the artificial intelligence signal control mode.