Active Vibration Control Based on BP Neural Network PID Algorithm

In order to solve the problem of poor adaptability of traditional methods to broadband line spectrum in ship lowfrequency vibration control, a double-layer active vibration isolation control method based on Sigmoid function BP neural network PID (BP-PID) is proposed. Verified by the rigid-flexible coupling simulation platform, the BP-PID achieves a peak attenuation rate of $\mathbf{4 6. 9 7 \% - 4 9. 4 0 \%}$ of the vibration acceleration of the middle layer under a single frequency excitation of 20 50 Hz. At the same time, a double-layer active vibration isolation experimental platform was built to verify the effectiveness of the algorithm, and the results showed that in the face of the typical “fundamental frequency dominant, multi-harmonic coupling” broadband line spectrum vibration of the ship, the control effect of 47.48% attenuation of the peak acceleration of the fundamental frequency ($\mathbf{2 0 H z}$) and the synchronous suppression of the frequency doubling peak $(40 / 60 / 80 \mathrm{~Hz})$ of more than $42 \%$ were realized. The research results reveal the technical advantages of neural network PID in broadband line spectrum signal control, and provide a highly adaptable solution for the active control of vibration of ship power plant.