Active Vibration Control Of Marine Floating Raft Vibration Isolation System

Security of the ship structure and reliable operation of machinery as well as the comfort on board can be seriously influenced by their vibration and noise. For the warship, vibration and noise of the power plant seriously weaken the battle effectiveness and hiding capability, therefore passive vibration isolation is usually used to reduct the vibration and noise, while limited by its structure and characteristics, the passive vibration isolation device can’t work in which condition vibration frequency lower than natural frequency. However, active vibration control can solve the problem well, effectively reduce the low frequency and high frequency vibration, therefore related research become the hotspot in recent years.This thesis aims at the study of active vibration control on floating raft isolation system. Starting with the fundamental theory of vibration isolation, this thesis analyzes the transmission characteristics of the single stage, two stage, floating raft isolation system, which provide guidance for setting up the test-bed and selecting vibration control frequency band. Then a test-bed of floating raft isolation system is established to measure experimental data as training samples using LMS Test.Lab. And then identify the inherent characteristics of the floating raft isolation device on the basis of experimental data with series-parallel neural network structure, training the neural network offline, establishing a neural network NARMA model, providing error back propagation channel for the neural network controller. Finally contrasting vibration response on middle raft in control system and no control system with MATLAB/Simulink. Due to the insufficient precision of offline identification control system inverse model, the effect of direct self-tuning control based on neural network is not ideal enough, only shows the feasibility of this method. However, the effect of indirect model reference adaptive control based on neural network is remarkable, the amplitude of middle raft decreases, what’s more the robustness of the control system is good, the simulation results turn out that this method is effective and practical.