Flat plates and cylindrical shells are commonly used in large equipment. To locate the low-velocity impact points in these structures, this study proposes an error-index-based algorithm for impact localization. The time of arrival of an impact-generated A0 Lamb waves was first estimated based on the energy of the signal. Equations for calculating the error indices were proposed for flat-plate and cylindrical shell structures, and the probability distribution functions of the impact points are constructed for visual localization. The impact test results on a flat plate and cylindrical shell indicated that, compared to the Morlet wavelet method, the proposed algorithm improved the mean relative error of impact point localization on the flat plate by 0.22%, 15.64%, and 15.26% under three different noise conditions, respectively (i.e., no noise, and SNR = 5 and 0 dB). For the cylindrical shell, the mean relative error of impact localization improved by 1.8%, 3.97%, and 28.12% under the three conditions, respectively. The results indicated that the proposed localization algorithm can accurately locate the impact points on a flat plate and cylindrical shell, even under strong background noise conditions, providing a reference for future research on locating low-velocity impact points in large equipment.

To solve problems related to much calculation to adapt to complex scenes in traditional structural sound source localization, this paper proposes a method based on neural network. The structural sound source at other positions was stimulated by successively striking 36 grid centers on the surface of the plate. The time delay between different accelerometer signals was considered as the input, and the location of the predicted sound source was considered as the output. The influence of the number of test sets and epoch training times on sound source localization accuracy was discussed. These results show that with the increase in the epoch training times, the number of test set decreases, and the number of training set increases, increasing the sound source localization accuracy of backpropagation neural network. However, these error conditions will frequently appear due to the overfitting phenomenon. When the epoch is trained to 50,000 times, and the quantity of the test set is 4, the backpropagation neural network has the best localization accuracy with an order of magnitude of 10−3 in error, and the localization error scope of the plate is between 0.01 and 0.1 m.