NONSTATIONARY OPTIMAL PATHS AND TAILS OF PREHISTORY PROBABILITY DENSITY IN MULTISTABLE STOCHASTIC-SYSTEMS

The rails of prehistory probability density in nonlinear multistable s tochastic systems driven by white Gaussian noise, which has been a sub ject of recent study, are analyzed by employing the concepts of nonsta tionary optimal fluctuations. Results of numerical simulations show th at the prehistory probability density is non-Gaussian and highly asymm etrical, and that it is an essential feature of noise driven fluctuati ons in nonlinear systems. We also show that in systems with detailed b alance the prehistory probability density is the conventional transiti on probability that obeys the backward Kolmogorov equation.