IDENTIFICATION OF RESTORING FORCES IN NON-LINEAR VIBRATION SYSTEMS USING FUZZY ADAPTIVE NEURAL NETWORKS

The fuzzy adaptive back-propagation (FABP) algorithm which combines fuzzy theory with artificial neural network techniques is applied to the identification of restoring forces in non-linear vibration systems. Simulated results show that the FABP algorithm is effective for the identification of dynamic systems. The FABP algorithm not only increases the training speed of the network, but also decreases the artificial interference of network parameters to a certain extent. Based upon the FABP algorithm, an improved scheme with a mutation mechanism is presented in this paper. The improved fuzzy adaptive BP (IFABP) algorithm extends the effectiveness and adaptivity of the FABP algorithm still further. The successful estimation of simulated systems show that a feasible method of identification is provided, which can be used to estimate the restoring forces in non-linear vibrating systems quickly and effectively.