Optimal Sensor Placement using Genetic Algorithm and Hybrid Crossover Operator

Document Type : Research Article


1 Associate Professor, Amirkabir University of Technology, Tehran, Iran

2 Ph.D. Student, Amirkabir University of Technology, Tehran, Iran


In this study, optimal sensor placement (OSP) which plays a key role in the health monitoring of large-scale structures, is investigated using the genetic algorithm. The OSP problem is among permutation problems that it's challenging to define the crossover operator in this kind of problems. In this study, a new hybrid crossover operator is proposed to find the optimal location for sensors, and two different strategies are investigated for selecting members to form the next generation population. Also, the two-structure coding method has been used instead of the typical binary coding method to create the chromosomes of the population members. The objective function and fitness is defined based on the modal assurance criterion (MAC) matrix that is calculated with identified mode shapes and analytical mode shapes. The efficiency of the proposed method was investigated on a high-rise structure. The results show that the mode shapes identified by the optimal placement obtained from the proposed method are identical to the analytical mode shapes of the finite element model. Also, the comparison between the sensor locations obtained by conventional operators and the proposed operator shows that the proposed hybrid crossover operator outperforms other operators in terms of accuracy and convergence speed.


Main Subjects