As a key component of a maglev train, the magnetic suspension control system ensures that the airgap between the train and its guideway is stable. Although current suspension controllers meet basic engineering requirements, problems related to their performance often arise during long-term operations. These problems can seriously affect the stability and reliability of a maglev system and may lead to partial suspension-point failure. Hence, this paper uses transfer learning-based deep reinforcement learning to develop an adaptive nonlinear suspension system controller that enables automatic adjustment of control strategies. First, suspension control based on deep reinforcement learning is mathematically modeled using Markov decision processes, and the nonlinear state space of a single electromagnet suspension control system is established as an agent–environment interaction with the developed deep reinforcement learning controller. Then a twin delayed deep deterministic policy gradient algorithm in an actor–critic framework is adopted to solve the Markov decision processes. To address the dispersion caused by nonlinear suspension control, a transfer learning-based two-stage training process is devised that first trains the twin delayed deep deterministic policy gradient networks on a linearized model and then transfers the networks to a nonlinear model. The effectiveness of the new controller is verified by comparing it with a conventional proportional–integral–derivative controller and a neural network based adaptive super-twisting sliding mode controller. The robustness of the transfer learning–deep reinforcement learning controller is examined in the presence of uncertainty, such as train load changes and disturbance forces in the suspension system. Additionally, experiments have also been conducted to confirm the effectiveness of the TL–DRL controller on a magnetic levitation system.

The suspension system in electromagnetic suspension maglev trains is highly nonlinear and sensitive to uncertainties, noise, and disturbances, making it quite challenging to design the proper electromagnetic voltages to control the suspension gap. In this paper, a self-tuning dual-layer sliding mode control system (SD-SMC) is developed to control the electromagnetic suspension system where voltage saturation, sensor noise, and a wide range of parametric uncertainties/variations exist. SD-SMC consists of dual-layer sliding mode controllers, a delay-compensated low-pass filter, and a forgetting least-squares estimator. Compared with existing control systems, SD-SMC can achieve high performance under large model uncertainties and disturbances with relatively low control voltages, and meanwhile, the control chattering and overshooting issues can be well mitigated. Both numerical and experimental examples are investigated to validate the proposed control system.

Advanced rail transit systems (ARTS), including high-speed rail and maglev trains, provide enhanced transportation options to meet the growing demand for efficient transportation systems. However, they present unique challenges in maintaining the safety and performance of their infrastructures. Structural health monitoring (SHM) has emerged as an essential practice to forestall the potential consequences of structural defects in ARTS. Recently, digital twins and digital shadows have been successfully employed in various industries to monitor the state of physical systems. However, their application for structural health monitoring in ARTS remains largely unexplored. Hence, this article explores the potential of digital twins and digital shadows, in improving structural health monitoring in ARTS. Due to the digital twins’ ability to bi-directional communication between a real system and its virtual replica, this article presents a comprehensive literature survey on their enablers and capabilities. Meanwhile, a framework for digital twins-based monitoring in ARTS is also proposed. The key distinctions and benefits of digital twins over other Industry 4.0 digital representation concepts, such as real-time monitoring, optimization, prediction, simulation, and decision-making, are identified. The paper highlights the significant opportunities that digital twins, especially, can offer to improve health monitoring. Similarly, limitations and bottlenecks that must be tackled in future research for implementations are also acknowledged. Finally, harnessing the power of digital twins can catalyze a transformative shift in ARTS, leading to more effective monitoring, enhanced safety, and improved performance.