A NONLINEAR ADAPTIVE GENETIC ALGORITHM FOR PARAMETER IDENTIFICATION OF THE BOUC-WEN MODEL AND ITS EXPERIMENTAL VERIFICATION

The Bouc-Wen model is a versatile nonlinear smooth hysteretic model that can characterize the stiffness degradation and strength degradation of structures and components under cyclic loads. It can be widely applied to describe the hysteretic behavior of various structures. The parameters of the Bouc-Wen model are critical to determining the mechanical characteristics of the hysteretic behavior of structural components. However, because there are many parameters of unclear physical meaning, the parameters are generally determined by approaching a group approximate solution from the hysteretic data. This paper aims at efficient identification of the parameters of the Bouc-Wen model and proposes a nonlinear adaptive genetic algorithm (NAGA). The efficiency is verified through the parameter identification of four RC shear walls with different reinforcement tested under low-cycle cyclic loading by using this algorithm. The hysteretic curves and algorithm efficiency obtained from parameter identification by NAGA and by the Standard Genetic Algorithm (SGA) are compared with the experimental data. The results show that the proposed method significantly improves the identification accuracy and efficiency. The proposed method can be used in hysterical model identification and corresponding nonlinear behavior simulation of structures.