A fish-shaped minimal prototype of lateral line system based on pressure sensing

Abstract

Lateral line is a unique organ, with which fish could sense surrounding environment by processing hydrodynamic information about local flow. This sensing mechanism provides a new perspective for researchers and engineers to build such a sensing system that could be applied to control and near field navigation for underwater robots and vehicles. In this paper, a pressure-sensing-based and fish-shaped minimal prototype is proposed, with two sensor arrays comprising eight pressure sensors acting as trunk lateral line. The effectiveness of our proposed prototype is validated by localizing a dipole source. Three methods, derivative free, quasi newton and genetic algorithm, are adopted to assess the sensing performance. Preliminary experiments demonstrate that the localization results match well with the actual positions of the dipole source in the direction along the artificial lateral line. While in the perpendicular direction, the localization results show increasing errors with the increase of dipole-sensor distance, but concentrate at a certain value at different distances. An error correction mapping is introduced to modify this deviation and the overall localization error is reduced to less than 0.06 Body Length (BL). In addition, by comparing the results obtained from different methods, we find that a simple algorithm can be effective for our proposed model-based scheme in localizing the dipole source.