Nonlinear model identification for Artemia populationmotion

In this paper, two different nonlinear models for Artemia swarming are derived. In order to generate the data suitable for identification, a robot driving the Artemia population has been built. The obtained data have been then used to identify the parameters of a model based on Newton’s equations and a black-box NARX model implemented by neural networks. The performances obtained validate the physical hypotheses underlying the gray-box model