Optimal placement of sensors for structural system identification and health monitoringusing a hybrid swarm intelligence technique

Abstract

Setting up a health monitoring system for large-scale civil engineering structures requires alarge number of sensors and the placement of these sensors is of great significance for suchspatially separated large structures. In this paper, we present an optimal sensorplacement (OSP) algorithm by treating OSP as a combinatorial optimizationproblem which is solved using a swarm intelligence technique called particle swarmoptimization (PSO). We propose a new hybrid PSO algorithm by combininga self-configurable PSO with the Nelder–Mead algorithm to solve this ratherdifficult combinatorial problem of OSP. The proposed algorithm aims precisely toachieve the best identification of modal frequencies and mode shapes. Numericalexperiments have been carried out by considering civil engineering structures toevaluate the performance of the proposed swarm-intelligence-based OSP algorithm.Numerical studies indicate that the proposed hybrid PSO algorithm generates sensorconfigurations superior to the conventional iterative information-based approaches whichhave been popularly used for large structures. Further, the proposed hybrid PSOalgorithm exhibits superior convergence characteristics when compared to other PSOcounterparts.