ИСТИНА |
Войти в систему Регистрация |
|
ИСТИНА ИНХС РАН |
||
Numerical solution of the multicomponent Smoluchowski coagulation equation is a complex problem of numerical mathematics. The most spread technique for the mathematical modelling of the coagulation process is Monte Carlo methodology. The classical and highly-accurate finite-difference schemes are useless in this application due to their extremely high algorithmic complexity. In this work we applied the low-rank approximations in TT-format of both the solution and the coagulation kernel and the fast TT-arithmetics to the implementation of the time-steps of the classical Runge-Kutta method. The obtained acceleration allows us to claim that we have constructed a novel numerical method which is more efficient than the known Monte Carlo technique.