Numerical Verification on Optimal Sensor Placements for Building Structure Before and After the Stiffness Decrease

Abstract

Japan and Taiwan are located in an area where earthquakes and typhoons occur very frequently. To ensure the safety of building structures, it is necessary to proceed with structural health identification before and after a disaster. This study proposes the optimal placement of sensors to reduce the number of sensors installed in the mid/high-story building for monitoring and inspection purposes, while simultaneously enhancing the accuracy of the determination of modal frequencies for the building structure. To achieve this goal, the first step is to obtain the reference modal frequencies of the structure. That is to install reference sensors on the structure for recording actual time-histories, then to obtain simulated time-histories using the cubic spline interpolation method. After the actual and simulated time-histories are collected from each structure floor, the Frequency Domain Decomposition (FDD) method is used to enhance the accuracy of reference modal frequencies, and then the Stochastic Subspace Identification (SSI) method is utilized to plot out the stabilization diagram. Several reference modal frequencies shown in the stabilization diagram will be determined. Finally, the Optimal Sensor Placement (OSP) locations for placing acceleration sensors can be obtained using a genetic algorithm. Structural models with different story numbers will then be simulated to obtain OSP locations. The condition of stiffness decrease will also be discussed. A comparison is made before and after stiffness decrease to discover the modal frequencies of the structures under the case of both optimal placement and original placement, ultimately to verify the feasibility of the methods proposed in this study.