Optimal sensor placement of a high-rise building using the combined genetic algorithm-ensemble Kalman filter framework

Abstract

Structural Health monitoring (SHM) techniques are utilized to discover damage in engineering structures, by reporting the health conditions of such structures using recorded measured data. Due to the improvement of technology in industry field, inventors developed sensors and many other monitoring machines to detect and locate damage in structures. Optimal sensor placement (OSP) is a major part of Structural Health monitoring. OSP methods are used to provide the user with important information and data that requires the lowest computational cost and time to be administered for damage identification. The Genetic Algorithm (GA), the Simulated Annealing (SA), the Particle Swarm Optimization (PSO), the Monkey Algorithm (MA), the Tabu search (TS) and many other approaches, are known under the name of optimization-based methods in OSP for locating and detecting damage in a structure. In this research, a 20-story building subjected to an earthquake load on its based is used as a numerical example. The novelty of this research is to test the efficiency, robustness, and accuracy of the combined Genetic Algorithm – Ensemble Kalman Filter (GA-EnKF) approach on high rise complex buildings. Three available fixed number of sensor cases are taken in consideration: six sensor cases, seven sensor cases and eight sensors case. For the six sensors case, the best sensors locations are found to be at floors 1, 2, 3, 6, 9 and 18. For the seven sensors case, the optimal sensors locations are found to be at floors 1, 2, 4, 5, 7, 12 and 20. For the eight sensors case, the best sensors locations are found to be at floors 1, 3, 4, 6, 8, 12, 14 and 20. Once the available sensors are placed on their optimal locations, found using the GA-EnKF methodology, the difference between the actual displacement and velocity of the different floors of the 20-story building, and the corresponding predicted values, found using the Ensemble Kalman Filter (EnKF), becomes very negligible. This means that the GA-EnKF framework converged successfully to the actual best sensors locations. Therefore, this work proves the efficiency and robustness of this method in determining the optimal sensor configurations for high-rise complex structures.