Application of Artificial Intelligence in Urban Rail Transit Operation and Maintenance

With the continuous expansion of urban rail transit and passenger flows, traditional manual operation and maintenance faces challenges in efficiency and safety. Artificial intelligence technology is driving a shift in operation and maintenance from periodic maintenance to predictive maintenance. This paper focuses on three key artificial intelligence technologies and their applications in rail transit operation and maintenance including Prognostics and Health Management (PHM), Computer Vision (CV), and Large Language Model (LLM). The research demonstrates that implementing PHM model for fault diagnosis of the train, utilizing CV model for visual inspection, and deploying LLM model to construct intelligent question-answering systems can significantly enhance the accuracy of fault diagnosis, optimize maintenance workflows, and effectively reduce lifecycle operational and maintenance costs. The conclusion highlights the practical value of artificial intelligence technologies in advancing intelligent operation and maintenance for rail transit.