OPTIMIZATION OF EXPERIMENTAL PARAMETERS FOR SPARSE ARRAY IMAGING

Significant research effort has been put into developing permanently attached sparse array networks for structural health monitoring. It is envisaged that such networks will be used to monitor the state of a complete structure over time. When these systems are installed on existing structures, it is possible that there will be errors in the assumed transducer locations due to the complexity of the structure and the large areas that will need to be populated. This has the potential to adversely affect the ability of a sensor network to detect the presence of a defect and also to cause misclassification of detected features. We address this problem, presenting an autofocus algorithm for correcting errors in the experimental parameters, including transducer locations, group velocity, and equipment delay. The improvements on the resulting imagery and defect detection are shown, clearly indicating an enhancement in the point-spread function. The ability of the technique to correct the parameters for complex geometries is illustrated using experimental data.