SPARSE ARRAY IMAGING OF CHANGE‐DETECTED ULTRASONIC SIGNALS BY MINIMUM VARIANCE PROCESSING

Spatially distributed arrays of permanently attached ultrasonic sensors are being considered for structural health monitoring systems. Most algorithms for analyzing the received signals are based upon change detection whereby baselines from the undamaged structure are subtracted from current signals of interest, and the residual signals are analyzed. In particular, delay‐and‐sum algorithms applied to the residual signals have been shown to be effective for imaging damage in plate‐like structures that support propagation of guided waves. Here we consider minimum variance processing of the residual signals, which is an adaptive beamforming method in common use for processing of radar signals where the weights are adjusted at each pixel location prior to summation based upon actual and expected signal amplitudes. Experimental results from a sparse sensor array show that this processing method can provide a significantly improved signal‐to‐noise ratio by suppressing unwanted sidelobes in the image.