Study on Active Control in Centrifuge-Vibration Compound Environment by Genetic Algorithm-Neural Network System

摘要

Working under the centrifuge-vibration compound environment, the vibrator is unbalanced because of the centrifugal force, which prevents the vibrator from working properly. So the active control on the vibrator is necessary. Aiming at controlling the countertop vibration triggered by the sinusoidal stimulus under the compound environment, a new active control algorithm, based on the governing equation of the vibrator, is proposed by means of the genetic algorithm and the neural network. The genetic algorithm is used to determine the optimal control force and a trained BP network is used to simulate the dynamic characteristic of the vibrator. The numerical examples demonstrate the effectivity of the proposed method.