Fei Ni
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Fei Ni has a total of 53 co-authors. Here are the top co-authors by number of shared publications:
- Junqi Xu
Tongji University
11 shared publications - Guobin Lin
Tongji University
5 shared publications Wantao Jia
3 shared publications- Wen Ji
Maglev Transportation Engineering R&D Center, Tongji University, Shanghai 201804, China
3 shared publications - Yifan Luo
College of Transportation, Tongji University, Shanghai, China
3 shared publications - Boyang Shen
School of Materials & Energy, Lanzhou University, Lanzhou, China
2 shared publications
Latest Publications
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Electronics
- Vol. 73 (10) , 14978-14992
Intelligent Explicit Model Predictive Control of Planar Motors for High-Precision Position Tracking Applications
To address the challenges of achieving high-precision motion control under physical constraints in motion systems, an intelligent explicit model predictive control (IEMPC) method for planar motors is proposed in this article. Unlike conventional explicit model predictive control (EMPC) approaches that suffer from excessive memory requirements, limited model flexibility, and dependence on commercial solvers, the proposed method integrates intelligent algorithms with an error state feedback solution space to construct a novel EMPC framework. First, a prediction model of error state is established using an augmented state-space model. Based on this model, a multi-objective optimal position control problem under control input constraints is formulated by incorporating a defined cost function. An offline solving strategy is then developed, in which the error state space is partitioned into multiple convex elliptical regions and explicit error state feedback control laws are derived for each subregion. To solve the constrained multi-objective optimal position control problem, an improved nondominated sorting genetic algorithm II (NSGA-II) is designed, generating a set of Pareto-optimal parameters for the control laws. These parameters are subsequently mapped to corresponding elliptical subregions to determine the control law in each subregion. Finally, the effectiveness and superiority of the proposed method are verified through systematic comparative experiments.
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Electronics
- Vol. 73 (10) , 14978-14992
Intelligent Explicit Model Predictive Control of Planar Motors for High-Precision Position Tracking Applications
To address the challenges of achieving high-precision motion control under physical constraints in motion systems, an intelligent explicit model predictive control (IEMPC) method for planar motors is proposed in this article. Unlike conventional explicit model predictive control (EMPC) approaches that suffer from excessive memory requirements, limited model flexibility, and dependence on commercial solvers, the proposed method integrates intelligent algorithms with an error state feedback solution space to construct a novel EMPC framework. First, a prediction model of error state is established using an augmented state-space model. Based on this model, a multi-objective optimal position control problem under control input constraints is formulated by incorporating a defined cost function. An offline solving strategy is then developed, in which the error state space is partitioned into multiple convex elliptical regions and explicit error state feedback control laws are derived for each subregion. To solve the constrained multi-objective optimal position control problem, an improved nondominated sorting genetic algorithm II (NSGA-II) is designed, generating a set of Pareto-optimal parameters for the control laws. These parameters are subsequently mapped to corresponding elliptical subregions to determine the control law in each subregion. Finally, the effectiveness and superiority of the proposed method are verified through systematic comparative experiments.
- 5 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Electronics
- Vol. PP (02780046) , 1-13
Composite Control With Prescribed Performance for High-Speed Maglev Train Levitation Systems
This article presents an improved composite control framework for high-performance levitation gap regulation in electromagnetic suspension (EMS) type high-speed maglev trains. The control architecture incorporates three strategically coordinated components: a prescribed performance function (PPF) to enforce transient performance by confining the tracking error within prescribed bounds, a cascaded finite-time extended state observer (CFTESO) for rapid and accurate estimation of lumped disturbances, and a global fast nonsingular terminal sliding mode controller (GFNTSMC) to achieve high-precision reference tracking with strong disturbance rejection. The stability of the closed-loop system and finite-time convergence are rigorously established via Lyapunov analysis. Real-time hardware-in-the-loop (HiL) validation under high-speed operation with real-world disturbances verifies the superiority of the proposed control framework. Compared with conventional control strategies, it performs coordinated optimization of the transient response and steady-state performance, along with higher control accuracy and more effective disturbance rejection.
- 5 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Electronics
- Vol. PP (02780046) , 1-13
Composite Control With Prescribed Performance for High-Speed Maglev Train Levitation Systems
This article presents an improved composite control framework for high-performance levitation gap regulation in electromagnetic suspension (EMS) type high-speed maglev trains. The control architecture incorporates three strategically coordinated components: a prescribed performance function (PPF) to enforce transient performance by confining the tracking error within prescribed bounds, a cascaded finite-time extended state observer (CFTESO) for rapid and accurate estimation of lumped disturbances, and a global fast nonsingular terminal sliding mode controller (GFNTSMC) to achieve high-precision reference tracking with strong disturbance rejection. The stability of the closed-loop system and finite-time convergence are rigorously established via Lyapunov analysis. Real-time hardware-in-the-loop (HiL) validation under high-speed operation with real-world disturbances verifies the superiority of the proposed control framework. Compared with conventional control strategies, it performs coordinated optimization of the transient response and steady-state performance, along with higher control accuracy and more effective disturbance rejection.
- 28 February 2026
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 249
GC-TimeGAN: Time–frequency domain model for high-fidelity reconstruction of Maglev track irregularity
- 28 February 2026
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 249
GC-TimeGAN: Time–frequency domain model for high-fidelity reconstruction of Maglev track irregularity
- Open Access
Machine Learning in Maglev Transportation Systems: Review and Prospects
Magnetic levitation (Maglev) technology has long garnered significant attention in the engineering community due to its inherent advantages, such as contactless operation, minimal friction losses, low noise, and high precision. Based on electromagnetic suspension (EMS) and electrodynamic principles, these systems are primarily developed for advanced transportation, while also inspiring emerging applications such as vibration isolation and flywheel energy storage. Despite progress, practical deployment faces critical challenges, including accurate modeling, robustness against nonlinear and uncertain dynamics, and control stability under complex conditions. Artificial intelligence (AI), particularly machine learning (ML) offers promising solutions. Studies show ML-based methods, i.e., improved particle swarm optimization (PSO) optimize proportional-integral-derivative (PID) to reduce controller output overshoot, deep reinforcement learning (DRL) reduces levitation gap fluctuation under complex conditions, ensemble learning achieves high fault diagnosis accuracy, and convolutional neural network-long short-term memory (CNN-LSTM) predictive maintenance cuts costs. This review summarizes recent AI-enabled advances in Maglev transportation system modeling, control, and optimization, highlighting representative algorithms, performance comparisons, technical challenges, and future directions toward intelligent, reliable, and energy-efficient transportation systems.
Open AccessMachine Learning in Maglev Transportation Systems: Review and Prospects
Magnetic levitation (Maglev) technology has long garnered significant attention in the engineering community due to its inherent advantages, such as contactless operation, minimal friction losses, low noise, and high precision. Based on electromagnetic suspension (EMS) and electrodynamic principles, these systems are primarily developed for advanced transportation, while also inspiring emerging applications such as vibration isolation and flywheel energy storage. Despite progress, practical deployment faces critical challenges, including accurate modeling, robustness against nonlinear and uncertain dynamics, and control stability under complex conditions. Artificial intelligence (AI), particularly machine learning (ML) offers promising solutions. Studies show ML-based methods, i.e., improved particle swarm optimization (PSO) optimize proportional-integral-derivative (PID) to reduce controller output overshoot, deep reinforcement learning (DRL) reduces levitation gap fluctuation under complex conditions, ensemble learning achieves high fault diagnosis accuracy, and convolutional neural network-long short-term memory (CNN-LSTM) predictive maintenance cuts costs. This review summarizes recent AI-enabled advances in Maglev transportation system modeling, control, and optimization, highlighting representative algorithms, performance comparisons, technical challenges, and future directions toward intelligent, reliable, and energy-efficient transportation systems.
- 25 November 2025
- Published by Elsevier in Probabilistic Engineering Mechanics
- Vol. 82
Neural network-based probabilistic tracking control for levitation systems under stochastic track irregularities
- 25 November 2025
- Published by Elsevier in Probabilistic Engineering Mechanics
- Vol. 82
Neural network-based probabilistic tracking control for levitation systems under stochastic track irregularities
- 1 July 2025
- Published by Elsevier in Probabilistic Engineering Mechanics
- Vol. 81
Data-driven modeling of high-speed maglev track irregularity
- 1 July 2025
- Published by Elsevier in Probabilistic Engineering Mechanics
- Vol. 81
Data-driven modeling of high-speed maglev track irregularity
- 16 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Transportation Electrification
- Vol. 11 (5) , 11582-11592
Global Fast Nonsingular Terminal Sliding Mode Control With Finite-Time Extended State Observer for Levitation Systems of High-Speed Maglev Train
High-speed maglev trains are attracting attention for their contactless operation and enhanced comfort. However, high-speed operation makes these systems particularly susceptible to significant disturbances that can impair levitation control performance. To address this challenge, this study proposes an advanced control strategy that integrates a global fast nonsingular terminal sliding mode control (GFNTSMC) with a finite-time extended state observer (FTESO). Based on the nonlinear dynamic characteristic, the proposed FTESO enables real-time disturbance estimation while the GFNTSMC provides effective compensation. Theoretical validation of finite-time stability is proven using Lyapunov stability theory. Numerical simulations and hardware-in-the-loop (HiL) experiments consistently show that the proposed FTESO-GFNTSMC method significantly outperforms the conventional linear extended state observer-based nonsingular terminal sliding mode control (LESO-NTSMC), achieving a 94.1% faster convergence rate and a 99.3% reduction in maximum tracking error under equivalent disturbance conditions. This demonstrates its effectiveness as a high-performance control strategy for maglev train operations.
- 16 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Transportation Electrification
- Vol. 11 (5) , 11582-11592
Global Fast Nonsingular Terminal Sliding Mode Control With Finite-Time Extended State Observer for Levitation Systems of High-Speed Maglev Train
High-speed maglev trains are attracting attention for their contactless operation and enhanced comfort. However, high-speed operation makes these systems particularly susceptible to significant disturbances that can impair levitation control performance. To address this challenge, this study proposes an advanced control strategy that integrates a global fast nonsingular terminal sliding mode control (GFNTSMC) with a finite-time extended state observer (FTESO). Based on the nonlinear dynamic characteristic, the proposed FTESO enables real-time disturbance estimation while the GFNTSMC provides effective compensation. Theoretical validation of finite-time stability is proven using Lyapunov stability theory. Numerical simulations and hardware-in-the-loop (HiL) experiments consistently show that the proposed FTESO-GFNTSMC method significantly outperforms the conventional linear extended state observer-based nonsingular terminal sliding mode control (LESO-NTSMC), achieving a 94.1% faster convergence rate and a 99.3% reduction in maximum tracking error under equivalent disturbance conditions. This demonstrates its effectiveness as a high-performance control strategy for maglev train operations.