Towards digital twin trains: implementing a cloud-based framework for railway vehicle dynamics simulation

Abstract

ABSTRACT The digital twin technology holds great promise in driving the railway industry into a new era of digital intelligence. By creating a dynamic, personalized digital replica of a physical train, it has the potential to enable a range of advanced applications, including predictive maintenance, adaptive control, and early fault detection. However, substantial challenges exist that hinder the effective implementation of the digital twin concept for trains. To resolve this bottleneck issue, in this paper, these challenges are first identified and thoroughly examined, with a specific focus on those within the realm of railway vehicle dynamics simulation. A novel cloud-based simulation framework for the development of railway vehicle dynamics simulation software is then proposed to tackle the identified challenges. The framework makes use of cloud computing and cloud-based visualization technology to offer a scalable and flexible solution for railway vehicle dynamics simulation. Using the framework as its foundation, a cloud platform for constructing digital twin trains is also proposed and presented, with a detailed explanation of its architecture and functionalities. In addition, CTTSIM, a cloud computing-aided train-track system dynamics real-time simulation software, is developed as an illustrative application case of the proposed framework. An evaluation is undertaken to demonstrate the feasibility and utility of both CTTSIM and the proposed framework. This work provides a practical and effective solution for using vehicle dynamics simulation for constructing digital twin trains, representing a step towards the full implementation of the digital twin concept in the railway industry.