Model order reduction in state space based on optimization for sensor placements of frame structures considering damping

Abstract

In Structural Health Monitoring (SHM), determining the best locations for putting measurement sensors to record the effective modal data is extremely important to serve the damage detection process. However, this
implementation is often based on engineers’ experience without using a specific strategy constructed on mathematical bases. Therefore, this work aims to propose an effective technique for optimization of sensor locations for frames considering damping based on a model order reduction (MOR) in state space. By maximizing all the coefficients on the main diagonal line of the modal assurance criterion (MAC) matrix, sensor positions can be optimized by the particle swarm optimization (PSO) with several modifications compared to its original form to meet the current optimization requirement. In which, design variables are integer numbers assigned to the labels of degrees of freedom (DOFs) of a structural system. Two steel frame structures containing one for 2D one-storey frame and one for 2D four-storey frame are illustrated with different sensor quantities to
demonstrate the ability of the present approach. Acquired results show the potential applications to the SHM
area.