Vibration-based damage detection, a nondestructive method, is based on the fact that vibration characteristics such as natural\nfrequencies andmode shapes of structures are changed when the damage happens. This paper presents cooperative coevolutionary\ngenetic algorithm(CCGA),which is capable for an optimization problem with a large number of decision variables, as the optimizer\nfor the vibration-based damage detection in beams. In the CCGA, a minimized objective function is a numerical indicator of\ndifferences between vibration characteristics of the actual damage and those of the anticipated damage. The damage detection\nin a uniform cross-section cantilever beam, a uniform strength cantilever beam, and a uniform cross-section simply supported\nbeam is used as the test problems. Random noise in the vibration characteristics is also considered in the damage detection. In the\nsimulation analysis, the CCGA provides the superior solutions to those that use standard genetic algorithms presented in previous\nworks, although it uses less numbers of the generated solutions in solution search. The simulation results reveal that the CCGA\ncan efficiently identify the occurred damage in beams for all test problems including the damage detection in a beam with a large\nnumber of divided elements such as 300 elements.
Loading....