Deployment and Application of On-Board PHM Models Based on Docker Containers for Intelligent Operation and Maintenance

A series of challenges have emerged in traditional prognostics and health management (PHM) systems equipped on rail transit vehicles, particularly under the intelligent operation and maintenance mode, including single-node services with low scalability, low resource utilization rate, and poor portab...

Full description

Saved in:
Bibliographic Details
Main Authors: DU Jiwei, WEN Lin, LYU Yu, JIANG Guotao, CHEN Kai, XIONG Yukai
Format: Article
Language:zho
Published: Editorial Office of Control and Information Technology 2025-06-01
Series:Kongzhi Yu Xinxi Jishu
Subjects:
Online Access:http://ctet.csrzic.com/thesisDetails#10.13889/j.issn.2096-5427.2025.03.015
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A series of challenges have emerged in traditional prognostics and health management (PHM) systems equipped on rail transit vehicles, particularly under the intelligent operation and maintenance mode, including single-node services with low scalability, low resource utilization rate, and poor portability. To address these issues, this paper introduces the current mainstream virtualization technology—Docker containers—to build a virtualization platform suitable for PHM systems on rail transit vehicles. The paper details the overall architecture design of the virtualization platform and the deployment plan for the PHM model, based on the development process of an on-board PHM product for intelligent operation and maintenance. Through experiments and toolkit porting, the functions of the designed product were comparatively analyzed and verified. The results show that the virtualization technology platform based on Docker containers facilitates desired performance gains. Specifically, this platform improves resource utilization and allows for more flexible allocation of hardware resources. Additionally, the parallel computing efficiency of multiple models is improved due to the lack of data interactions among them, facilitating superior portability and scalability.
ISSN:2096-5427