An Integer-Constrained Genetic Algorithm for Sensor Placement in Structural Health Monitoring

Abstract

Application of Genetic Algorithms (GAs) for allotting sensors in Structural Health Monitoring (SHM) has received wide attention during the last three decades because of their potential as global search technique. However, from computational perspective, sensor allocation is a complex combinatorial optimization problem and can lead to some constraints that reduces the efficiency of simple GAs. To eradicate such dilemma, an Integer Constrained Genetic Algorithm (ICGA) is introduced for finding the optimal placement of sensors. Integer coded string and modal assurance criteria oriented objective function are adopted respectively to represent and measure the utility of a sensor configuration. A benchmark bridge structure is studied to demonstrate the feasibility and effectiveness of ICGA. Later, the simulation results obtained by the ICGA are compared to the conventional GA. The result shows that ICGA can satisfactorily identify the number of sensors along with their locations and enhances the convergence of the algorithm. More apparently, proposed algorithm can reduce the dissipative storage space generated by conventional methods, removes any redundancy of sensor and improves the balance between exploitation and exploration of the search space.