Variational bayesian based interactive multiple model adaptive vibration active control method

The variational Bayesian interactive multi-model adaptive active vibration control method solves the problems of slow convergence speed and insufficient robustness of traditional methods under unsteady excitation, and achieves fast response and high-precision vibration control effect, which is suitable for complex multi-channel systems.

CN122363395APending Publication Date: 2026-07-10HARBIN ENG UNIV +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610821184.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-06-09
Publication Date
2026-07-10

AI Technical Summary

Technical Problem

Existing adaptive active vibration control methods based on filter feedforward/feedback structures suffer from slow convergence and insufficient robustness when faced with unsteady excitations common in marine propulsion machinery. In particular, they are difficult to respond quickly after sudden changes in frequency or amplitude, leading to instantaneous failure of vibration suppression or increased residuals.

Method used

An interactive multi-model adaptive vibration active control method based on variational Bayes is adopted. By identifying the multi-channel system, a system model set with multiple potential dynamic modes is established, and online joint estimation and optimization are performed under the variational Bayes framework. Combined with the covariance cross-fusion method, an anti-noise signal is generated to drive the actuator output.

Benefits of technology

It achieves rapid response and high-precision vibration control under unsteady excitation, improves the robustness and adaptability of the system, can effectively respond to sudden changes in frequency or amplitude in a very short time, is suitable for complex multi-channel systems, and can maintain vibration control effect even when the reference signal is inaccurate or missing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122363395A_ABST
    Figure CN122363395A_ABST
Patent Text Reader

Abstract

This invention discloses an interactive multi-model adaptive active vibration control method based on variational Bayesian methods, belonging to the field of vibration control technology. The method includes establishing a system model set containing multiple potential dynamic modes and initializing each model; performing variational Bayesian iteration on each model within each control sampling period; repeating the above process for each channel to obtain the optimal state estimate for each channel; fusing the state estimates of multiple channels using a covariance cross-fusion method to calculate the global optimal state estimate; convolving the global optimal state estimate with a reference signal filtered by the secondary channel model to generate an inverse control signal, which drives an actuator via a power amplifier to produce a secondary force that counteracts the original vibration, thus achieving vibration suppression. This invention achieves high-precision and highly robust vibration suppression by fusing variational Bayesian adaptive Kalman filtering with interactive multi-model estimation.
Need to check novelty before this filing date? Find Prior Art