Improvement of the Transient Levitation Response of a Magnetic Levitation System Using Hybrid Fuzzy and Artificial Neural Network Control

The magnetic levitation system (MLS) is a versatile technology with wide-ranging applications, offering benefits like contactless operation, high precision, energy efficiency, and reduced maintenance. However, achieving precise and stable levitation, especially during transient states, remains a critical challenge. This paper presents a novel method to improve the transient levitation response of MLS by integrating fuzzy logic with artificial neural network (ANN) control. By leveraging the strengths of both methodologies, the proposed hybrid control aims to improve the performance of MLS during transient operations such as initial levitation. The hybrid control consists of conventional control (PID position and PI current controls), off-line ANN identification, ANN control and fuzzy logic. Experimental comparisons with PID and disturbance observer show that the proposed hybrid ANN control improves not only transient response during the initial levitation (the rise and settling times by 92.7% and 85.0%, respectively) but also the sinusoidal command following by 57.8%. The performance improvement and stability of the proposed control were discussed by measuring the closed-loop frequency responses.