Control of static levitation attitude based on CNF-LADRC for distributed electric-drive maglev car

This paper proposes a CNF-LADRC strategy for static levitation attitude control of a distributed electric-drive maglev car actuated by permanent-magnet electrodynamic wheels, targeting simultaneous disturbance rejection and transient shaping in the pitch, roll, and yaw channels. The problem is challenging because the platform combines nonlinear electromagnetic force characteristics, cross-axis coupling, model uncertainty, external disturbances, and a weakly damped yaw mode that conventional ADRC/LADRC designs struggle to regulate without compromising response speed. A three-DOF pitch-roll-yaw model is formulated in state-space form, and a Composite Nonlinear Extended State Observer (CNESO) is integrated with LADRC and CNF-type nonlinear feedback to compensate for lumped disturbances while enhancing near-equilibrium damping. Observer-gain feasibility conditions, quantitative boundedness estimates, and a tuning procedure are established through Lyapunov analysis, and the resulting controller is evaluated in simulations and prototype experiments against PID, SMC, robust control, and ADRC under constant-force disturbances, composite sinusoidal/noise disturbances, and reference-tracking tasks. The results show that CNF-LADRC suppresses divergent yaw oscillations, reduces overshoot and amplitude variations, smooths speed-differential commands, and improves disturbance-immunity and tracking indices, demonstrating a practical route toward stable and safer static levitation of distributed electric-drive maglev cars.