Design and Development of an Experimental Device for Underwater Target Recognition Based on Bionic Lateral Line

As a critical organ perceiving surrounding fluid environment, the lateral line endows fish with the ability of navigating, feeding and avoiding predators and obstacles. Its specialized structure and function derive from adapting to natural environment and thousands of years' evolution, which inspire researchers to explore a brand-new submarine perception technology. In this paper, an experimental device based on the bionic lateral line is designed and developed to recognize underwater target in three-dimensional space, coalescent with the Generalized Regression Neural Network. Results show that the experimental device can localize the underwater target at a distance of two-body-length away, when the training data are collected from a cuboid sampling space at the size of $12 \text{cm} \times 12\text{cm} \times 24 \text{cm}$.