Data-driven model discovery using SINDy on particle based simulations of dry active nematics

Active nematic liquid crystals exhibit a wide range of exotic phenomena that cannot be observed in equilibrium systems. However, due to their complex non-equilibrium nature, the theoretical tools available to probe active nematics are quite limited. In order to better understand these systems, we use a tensor form of "Sparse Identification of Nonlinear Dynamics" (SINDy), which is a regression technique that takes raw data as an input and returns a parsimonious governing partial differential equation. Using tensor SINDy greatly reduces the complexity of the input library and makes it simple to include terms up to a desired order in the Q, velocity, and density fields. Furthermore, the output is more easily interpretable to humans. In addition to introducing the tensor SINDy framework, I will discuss results that we have obtained by applying tensor SINDy to particle-based dry active nematic simulations, where we control the microscopic parameters of the system. Using SINDy, we are able to connect the emergent hydrodynamic equations to the microscopic properties of our model system.