Quantum-Driven Optimization for Current Control of a Single-Coil Active Magnetic Bearing

Active magnetic bearings, inherently open-loop unstable and nonlinear, are essential for non-contact, and high-speed processes facilitated by electromagnetic levitation, but have considerable difficulties in controlling nonlinearities at wide air gaps due to operational factors or dynamic effects. Existing research establishes the framework, providing opportunities for innovative control approaches for enhanced industrial use. This work addresses the need to establish an effective controller by integrating empirical findings from a single-axis active magnetic bearing system with linear superposition principles of quantum mechanics using semiclassical estimates. The functioning of the system is mathematically represented by employing a density matrix framework with quantum observables such as coherence, von Neumann entropy, Hurst exponent, purity, Detrended fluctuation analysis, and fidelity, which guide the quasi-Newton Limited-memory Broyden-Fletcher-Goldfarb-Shanno Bound (L-BFGS-B) optimization process in designing a fast-settling, overshoot-free current controller that stabilizes the electromagnetic coil current while ensuring fewer iterations, rapid convergence, and less computing power. Experimental results inform the controller design, and frequency response and eigenvalues show that the controller can maintain asymptotic stability in the presence of considerable noise and nonlinear dynamics at a high air gap. The work demonstrates the effectiveness of combining theoretical quantum mechanics with classical empirical investigation to provide sophisticated and flexible control methods using quantum computing, demonstrating its major prospects for Industry 4.0 and beyond.