DISTRIBUTION OF ESCAPE TIMES IN A DRIVEN STOCHASTIC-MODEL

We analyze the probability distribution of escape times out of a metas table well for a stochastic model driven by a large-amplitude sinusoid al time-dependent field. The system obeys a Langevin equation which is solved numerically by generating stochastic trajectories, both for a white and for an Ornstein. Uhlenbeck noise. The probability distributi on changes from monomodal to multimodal as the noise strength is incre ased. The average escape time shows a nonmonotonic behavior with the n oise intensity which is associated with the change in the structure of the probability distribution. For a noise with a correlation time muc h longer than the period of the driving field, the resonance effects a re enhanced with respect to the white-noise case.