Machine Learning Approach for Particle Kinetics Modeling

In this work is presented numerical modeling of magnetic particle kinetics in outer magnetic fieldby FDM method and machine learning approach. Particle dynamics method is performed by nodal solver, based on first order finite difference method, with step-wise transient collision estimator and static outer magnetic field gradient intensity force correction. Magnetic field distribution is modeled independently of particle dynamics by Biot-Savart method. Further, a machine learning approach based on LSTM network is trained for particle trajectory modeling. Consideration on network parameters is made for the particular modeling task for Fe<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf>O<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</inf> particles convective kinetics.