以AI技術實現適應性號誌控制

本篇文章主要在探討如何運用AI（artificial intelligence）人工智慧技術於適應性號誌控制，並以「高雄港聯外AI智慧號誌控制系統計畫」及「高雄市脆弱路段智慧化號誌交控應用計畫」等二項計畫為案例，分析導入之策略、方法與效益及未來可能之發展方向。AI人工智慧應用於適應性號誌控制可分為不固定週期與固定週期動態調整時比兩種方式，不固定週期主要應用於高快速道路匝道鄰近路口，以尋求最大之紓解率；固定週期則應用於路段之號誌控制，以達到車流續進及最大紓解率。導入AI智慧號誌控制後可以降低10%-20%旅行時間，減少10%左右停等延滯，並達到節能減碳效果。

關鍵字

AI人工智慧；適應性號誌控制

並列摘要

This article discusses the application of AI (artificial intelligence) technology in adaptive traffic signal control, using two projects, namely the "Kaohsiung Port AI Intelligent Signal Control System Project" and the "Smart Signal Control Application Project for Vulnerable Road Sections in Kaohsiung City," as case studies. The strategies, methods, benefits, and possible future development directions of these implementations are analyzed. The application of AI in adaptive traffic signal control can be categorized into two approaches: dynamic adjustment with a non-fixed cycle length and dynamic adjustment with a fixed cycle length. The non-fixed cycle length approach is mainly applied to intersections near highway and express-way ramps to achieve the maximum relief rate. The fixed cycle length approach, on the other hand, is used for signal control on road sections to achieve traffic flow progression and maximize relief rates. After implementing AI intelligent signal control, travel time can be reduced by 10% to 20%, stop delays can be reduced by around 10%, and energy conservation and CO2 emissions reduction goals can be achieved.