Crack identification in beam-like structures using cuckoo optimization algorithm

To identify cracks in a structure, an inverse analysis has been performed using the static and dynamic response of the structure. The accuracy of the inverse analysis is highly dependent on the form of the objective function and the choice of optimization method. Generally, the optimization method searching the solution by using the gradient of the objective function is adopted. However, for the crack problem interested in this study, it is known to have the drawback that the search solution strongly depends on the initial value setting. The biomimetics-based cuckoo optimization algorithm (COA) is a method that can overcome this drawback. However, few examples of this method for identifying the crack depth and location in a structure are available. In this study, an inverse analysis method for determining crack depth and location in beam-like structures by the COA was developed. The size and location of the crack were identified by the COA. In this study, the transfer matrix method (TMM) was utilized to numerically calculate the deformation and rotation angle of the beams, which are required in the inverse analysis. For this purpose, the fundamental theory was extended to the cracked element. Consequently, it was confirmed that the crack depth, crack location, and magnitude of the applied load can be predicted by the developed method with good accuracy.