All Information of Global Properties in Dusty Plasmas from Individual Particle Dynamics Using Machine Learning

Currently, it is a great challenge to accurately diagnose global properties of dusty plasmas from limited data. Based on machine learning, a novel diagnostic of various global properties in dusty plasma experiments is developed from individual particle dynamics. It is found that, for both two-dimensional (2D) dusty plasma simulations and experiments, the global properties of the screening parameters κ and the coupling parameter Γ can be accurately determined purely from the position fluctuations of individual particles. Hundreds of independent Langevin dynamical simulations are performed with various specified κ and Γ values, leading to abundant individual particle position fluctuation data, which can be used for training, validating, and testing various convolutional neural network (CNN) models. To confirm the feasibility of this diagnostic, three different CNN models are designed to determine the κ value. For the simulation data, all these CNN models have excellent performance in determining the κ value, i.e., the averaged determined κ value is nearly the same as the specified κ value. For the experiment data, the distributions of the determined κ values always exhibit one prominent peak, whose locations well agree with the determined κ value from the widely accepted phonon spectra fitting method. Furthermore, this diagnostic method is further developed to determine the κ and Γ values simultaneously, with the satisfactory outcome using the input 2D dusty plasma data from both simulations and experiments. The excellent performance of the CNN models developed here clearly indicates that, using machine learning, all information of global properties of 2D dusty plasmas can be obtained purely from individual particle dynamics.