MEKF-Tuned Full-State Feedback for Two-Mass Resonance Suppression

To address the core challenge of unknown parameters and unmeasurable states in mechanical resonance suppression for two-mass systems, this paper proposes a self-tuning control scheme integrating a Multilayer Extended Kalman Filter (MEKF) with full-state feedback. A parallel dual-layer EKF architecture identifies load inertia in real-time while concurrently achieving high-precision estimation of load speed, shaft torque, and disturbance torque. By dynamically configuring full-state feedback gains based on the optimal inertia ratio strategy, optimal system damping characteristics are preserved. This method overcomes the limitations of slow convergence in single EKF implementations and observer mismatch, enabling dynamic resonance suppression. Simulation results verify its effectiveness in resonance mitigation.