Self-Tuning Parameters of a Maglev Control System Based on Q-Learning
Journal Title: Mechatronics and Intelligent Transportation Systems - Year 2024, Vol 3, Issue 2
Abstract
Maglev transportation, as an innovative mode of rail transit, is regarded as an ideal future transportation system due to its wide speed range, low noise, and strong climbing ability. However, the maglev control system faces challenges such as significant nonlinearity, open-loop instability, and multi-state coupling, leading to issues like insufficient tuning and susceptibility to environmental influences. This paper addresses these problems by investigating the self-tuning parameters of a maglev control system using Q-learning to achieve optimal parameter tuning and enhanced dynamic system performance. The study focuses on a basic levitation unit modeled after the simplified control system of an electromagnetic suspension (EMS) train. A Q-learning reinforcement learning environment and Q-learning agent were established for the levitation system, with a forward "anti-deadlock" reward function and discretization of the action space designed to facilitate reinforcement learning model training. Finally, a Q-learning-based method for self-tuning the parameters of the maglev control system is proposed. Simulation results in the Python environment demonstrate that this method outperforms the Linear Quadratic Regulator (LQR) control method, offering better control effects, improved robustness, and higher tracking accuracy under system parameter perturbations.
Authors and Affiliations
Yang Wang, Yougang Sun, Wen Ji, Junqi Xu
Keywords
Related Articles
- EP ID EP742914
- DOI -
- Views 65
- Downloads 0
Performance Evaluation of a Four-Legged Signalized Intersection with Variable Traffic Flow Dynamics
The rapid growth of population and vehicular traffic has necessitated effective urban planning strategies to mitigate traffic congestion and enhance roadway efficiency. This study focuses on a critical signalized interse...
Enhancing Urban Traffic Management through YOLOv5 and DeepSORT Algorithms within Digital Twin Frameworks
The acceleration of urbanization and the consequent increase in population have exacerbated urban road traffic issues, such as congestion, frequent accidents, and vehicle violations, posing significant challenges to urba...
Advanced Vehicle Detection and License Plate Recognition via the Kanade-Lucas-Tomasi Technique
The optimization of traffic flow, enhancement of safety measures, and minimization of emissions in intelligent transportation systems (ITS) pivotally depend on the Vehicle License Plate Recognition (VLPR) technology. Cha...
Enhancing Cold Chain Logistics: A Framework for Advanced Temperature Monitoring in Transportation and Storage
In the face of the increasingly demanding of goods transportation and storage, the orchestration of cold chain logistics emerges as a critical and multifaceted endeavor. This study, addressing a notable gap in literature...
Integrating Cultural Norms and Behavioral Risk Factors into Traffic Accident Mitigation: A Hybrid MCDM Approach for Libya
The mitigation of road traffic accidents remains a critical global challenge, particularly in regions where cultural norms and behavioral risk factors significantly influence driving practices. This study employs a hybri...