Research on the construction of the knowledge bases of urban rail transit rail infrastructure maintenance
In response to the challenges posed by the diverse technical standards,abundant data,and various document structures in the field of operation and maintenance(O&M)of urban rail transit,as well as the difficulty in querying and retrieving maintenance data due to the wide variety of on-site defects,a solution for constructing an infrastructure maintenance knowledge base is proposed for the first time in the domain of urban rail transit.By utilizing OCR technology and deep learning algorithms,image data and video data were textualized,and intelligent cleansing and organization of data was achieved with the assistance of a GDR model.Natural language processing techniques such as word segmentation and distributed storage technology were also employed to standardize the management of data materials.Indexing technology was implemented to significantly improve the speed of content retrieval.By establishing index technology,content could be found quickly.Finally,by establishing the knowledge base on an intelligent maintenance platform and using rail tracks as an example,possible defect issues and root cause analyses were demonstrated,along with corresponding maintenance solutions.This solution not only addresses the difficulty of maintenance personnel in accessing information,but also significantly reduces maintenance costs,effectively improving the efficiency and quality of maintenance work.
urban rail transitinfrastructuremaintenance knowledge base
NETL
NSTL
万方数据
