MIMO active vibration reduction control method and system with deep learning automatic diagnosis mechanism

Abstract

The invention discloses an MIMO active vibration reduction control method and system with a deep learning automatic diagnosis mechanism. The method comprises the steps that a vibration acceleration signal of a measuring point and an excitation source vibration signal are collected through a sensor; the excitation source vibration signals are input into a multi-channel coupling type Fxlms self-adaptive algorithm for active vibration reduction control calculation, and control signals of all channels for counteracting the vibration excitation effect are obtained; establishing a stability automatic diagnosis mechanism based on a convolutional neural network according to the vibration acceleration signal; judging whether the current active vibration reduction control is control divergence or control convergence according to a stability automatic diagnosis mechanism; if the control divergence is the control divergence, giving an alarm, automatically resetting the control coefficient of the multi-channel coupling type Fxlms adaptive algorithm, and adaptively adjusting the control coefficient until the control coefficient is converged; and if control convergence occurs, the alarm is released. According to the method, adaptive vibration reduction of multiple coupling channels can be realized, automatic diagnosis of algorithm stability is realized by using a deep learning technology, and the robustness of a control system is guaranteed.

Description

technical field [0001] The invention belongs to the technical field of vibration reduction control, and in particular relates to a MIMO (Multiple Input Multiple Output, Multiple Input Multiple Output) active vibration reduction control method and system with a deep learning automatic diagnosis mechanism. Background technique [0002] Due to the complex vibration environment of systems such as buildings, ships, and railways in practical engineering, active control algorithms are required to simultaneously suppress periodic vibrations with multiple line spectra and be robust to changes in the external environment. Therefore, the multi-channel adaptive active vibration reduction control algorithm with high reliability is the basis to ensure the effect of active vibration reduction. However, the current active vibration reduction algorithm lacks an automatic stability diagnosis mechanism, and cannot effectively prevent the control divergence caused by changes in the external env...

AI Technical Summary

Problems solved by technology

[0003] In view of this, the present invention proposes a MIMO active vibration reduction control method and system with a deep learning automatic diagnosis mechanism, which is used to solve the problem that the active vibration reduction algorithm lacks a stable automatic diagnosis mechanism

Images