Design optimization of artificial lateral line system under uncertain conditions

An artificial lateral line consists of a set of flow sensors arranged around a fish-like body which aims at localizing the surrounding moving objects, a common example of which is a vibrating sphere, called a dipole. The presence of diverse sources of uncertainty in the flow environment and flow sensors leads to an error in localization and thus challenges practicability of the underlying idea, especially considering that localization accuracy significantly declines when uncertainties intensify. Accuracy of localization depends on selection of the parameters of the artificial lateral line including the number and the location of the sensors as well as the shape and the size of the lateral line. In this study, different sources of uncertainties are identified and modeled in the problem formulation. A parametric fitness function is defined that addresses computational and practical goals and encompasses the effect of different sources of uncertainties. A bi-level optimization tool is formed to find the optimum artificial lateral. Comparison of the optimized designs in different cases reveals that, the optimized design highly depends on the amount of uncertainties in the problem as well as the number of available sensors. The proposed methodology for handling noisy and nested optimization task can be extended to solve other similar problems.