Skip to main navigation Skip to search Skip to main content

High-fidelity random micro-vibration reproduction via a stinger-free magneto-pneumatic platform and multi-resolution adaptive FD-ILC

  • Harbin Institute of Technology

Research output: Contribution to journalArticlepeer-review

Abstract

High-fidelity broadband random micro-vibration reproduction is a fundamental prerequisite for vibration-isolation performance evaluation and metrological calibration of ultra-precision equipment. Unfortunately, conventional ground excitation systems generally rely on mechanical stingers for force transmission, which tend to introduce parasitic stiffness and nonlinear friction. Moreover, under random vibration conditions, standard control algorithms inherently suffer from noise amplification and steady-state overlearning, severely constraining the signal-to-noise ratio and metrological reliability of micro-vibration testing. To address the above issues, this paper proposes a non-contact, direct-drive ground excitation platform based on a stinger-free magneto-pneumatic hybrid suspension, combining contactless actuation via a voice-coil-motor array with air-spring pre-isolation and low-stiffness load-bearing support to establish a low-background dynamic environment with reduced parasitic stiffness and mechanical coupling. A multi-resolution adaptive frequency-domain iterative learning control (MRA-FD-ILC) algorithm is further developed. This algorithm integrates one-third-octave-band error evaluation with narrowband high-resolution spectral updating to effectively suppress stochastic variance disturbances and overlearning, thereby enabling stable convergence in random vibration spectral-shaping control. Experimental results demonstrate that the proposed system reproduces an adjustable random vibration spectrum from VC-C to VC-A over the 0.5–100 Hz frequency band. Across the VC-A–VC-C tests, the mean maximum one-third-octave band-wise absolute error did not exceed 0.78 dB, while the corresponding mean overall RMSE did not exceed 0.68 dB. The system also maintains stable spectral-shaping capability in the ultra-low-frequency range. The proposed method can provide important technical support for the performance evaluation of ultra-precision vibration-isolation systems.

Original languageEnglish
Article number114782
JournalMechanical Systems and Signal Processing
Volume259
DOIs
StatePublished - 1 Sep 2026

Keywords

  • Excitation platform
  • Iterative learning control
  • Micro vibration
  • Multi-resolution
  • One-third-octave band

Fingerprint

Dive into the research topics of 'High-fidelity random micro-vibration reproduction via a stinger-free magneto-pneumatic platform and multi-resolution adaptive FD-ILC'. Together they form a unique fingerprint.

Cite this