Structural system modelling from base excitation measurements using swarm intelligence

Abstract

Structural system identification is a fundamental task in mechanical engineering since it enables the characterization of the dynamic behaviour of structures from experimental data. The identification of the mass, stiffness, and damping matrices of a structure is essential for the design, analysis, and control of mechanical systems. This paper proposes a novel method for identifying the spatial properties of a structure using base excitation measurements and swarm intelligence. Maintaining relevant physical properties is ensured by placing algebraic constraints on factors of the mass and stiffness matrices. This formulation circumvents the complexities of additional constraints in metaheuristic techniques. Consistency of the eigenvalues with respect to the test data is assured through an integrated refinement process during the stiffness matrix estimation. A healing process enables the rescaling of the mass matrix, guaranteeing an accurate representation of the structure’s total mass. The search space definition only requires prior knowledge of the total mass of the structure and an estimate of anticipated maximum natural frequencies within the targeted frequency range. The efficacy of this method is tested on a numerical example.