Numerical solution of the equations of dynamics
of particles based on neuro accelerator
and neural network selection of potential
of a pair interaction in virtual reality
Abstract:
The paper proposes a method for sampling of a continuous medium interacting dynamic particle, the
solution of equations of particle dynamics on the basis of a series-parallel architecture based on hardware
neuroaccelerator by a neural network selection, the proposed law of interaction between the particles, taking into account the elastic and inelastic properties of solids.
Keywords:dynamic particles, a series-parallel architecture, neuroaccelerator, neural network, adaptive critic.