Stochastic Hill Climbing for damage detection
Description
The dataset ,,Stochastic hill cliimbing _severity" consists of the deflection under its own weight values for a cantilever beam of dimensions 1x0.05x0.005 m both in an undamaged state and affected by different transverse cracks. The dataset ,,Severities obtained with the SHC algorithm" are the severity of different transverse cracks (open and closed) determined at the fixed end by using the SHC method described in the paper ,,Determining the severity of open and closed cracks using the strain energy loss and the hill-climbing method". The dataset ,,Training data for damage detection_RFS" is generated by using the severities calculated eith the SHC method and the procedure described in paper ,,Determining the severity of open and closed cracks using the strain energy loss and the hill-climbing method" and can be used to train an ANN for detecting, locating and evaluating transverse cracks of different depth and width present in cantilever beams. The datasets ,,FEM RFS values for testing damage detection ANN" and "Experimental testing values" are RFS values calculated for the obtained natural frequencies of the same cantilever beam affected by different transverse cracks in known locations. The frequency values are determined using FEM and also from real measurements in the laboratory.
Files
Steps to reproduce
1. generate an initial point 2. evaluate the initial point 3. take a step s 4. evaluate candidate point 5. check if we should keep the new point After generating the objective function, given in relation 9 from paper,, Determining the severity of open and closed cracks using the strain energy loss and the hill-climbing method" and by involving relation 10, the deflection at the fixed end of a beam can be found. As input three points with k=1…3 are needed. These points are the deflections of the beam at the free end when the crack is located at different distances from the fixed end, found involving the FEA.