Structural Parameter Identification Using Robust Genetic Algorithm Optimization Method

Abstract

Parameter identification is the process of obtaining the structural parameters of a system based on the measured input dynamic force and output responses. Most of the classical identification methods reformulate the equation of motion to the state space equation and use the gradient-based search method. Hence, they are usually effective only for a system with a small number of unknown parameters, due to its numerically ill conditioned nature. This study uses a non-classical search method, namely Genetic Algorithm (GA), which has several advantages over classical search method. The search capability is further improved with the introduction of an embedded Local Search (LS) method, which will benefit the fine-tuning of the search process. Examples of a cantilever beam and a 'free-free' beam are presented to show the effectiveness of the combined GA and LS method (GA-LS). The GA-LS method is finally applied to identify the mass and stiffness parameters of a model helicopter blade using measured excitations and accelerations.