Active vibration control using nonlinear auto-regressive neural network to identify secondary channel

The power unit on board the ship generates periodic low-frequency vibration that affects normal operation of equipment board, and adaptive feedforward control algorithm can effectively suppress such harmful noise. But needs to obtain identification model secondary channels, frequency domain least squares method based linear Extended auto-regressive (ARX) is difficult with nonlinear characteristics. (NARX) adds mapping layers topology ARX enhance capability NARX for complex systems. In this paper, a block diagram Fx-LMS proposed, then initial parameters neural network are optimized using Quantum Particle Swarm Optimization (QPSO) channel identified, results show accuracy identifying higher than model. simulation experimental damping effect proposed better traditional both single-line spectrum multi-line disturbances, which provides new suppression disturbances.