Extreme shock models (Q5926465): Difference between revisions

Shock models are systems that at random times are subject to shocks of random magnitudes. One distinguishes between two major types: cumulative shock models and extreme shock models. Systems governed by the former kind break down when the cumulative shock magnitude exceeds some given level -- the effect of the shocks is cumulative -- whereas systems modeled by the latter kind break down as soon as an individual shock exceeds some given level. Both models are in turn divided into two kinds: model I and II, respectively. In the first model a given shock is correlated to the length of the time interval preceding the given shock, and in model II to the time interval following the given shock. The general setup in cumulative shock models is a family \(\{(X_{k}, Y_{k})\), \(k\geq 0\}\) of i.i.d. two-dimensional random vectors, where \(X_{k}\) represents the magnitude of the \(k\)th shock and where \(Y_{k}\) represents the time between the \((k-1)\)st and the \(k\)th shock in model I, and the time between the \(k\)th and the \((k+1)\)st shock in model II. The main object of interest is the lifetime/failure time of the system. In the model I: \(X_{0}=Y_{0}=0,\) and \(T_{n}=\sum_{n}Y_{k}\) and \(S_{n}=\sum_{n}X_{k},\) \(n\geq 1,\) and the first passage time process \(\{\tau(t), t\geq 0\}\) is \(\tau(t)=\min\{n: S_{n}>t\}.\) The failure time is described by the random variable \(T_{\tau(t)}.\) The aim of the present paper is to provide an analogous treatment for the extreme shock model. The model I is considered. The difference is that \(\tau(t)\) is replaced by \(\tau(t)=\min\{n: X_{n}> t\}.\) Moment relations and asymptotic distributions of \(T_{\tau(t)}\) as \(t\) increases in such a way that \(P\{X_{1}> t\}\) tends to \(0\) are derived. A final section discusses some extensions: more general events of failure, the non-i.i.d. case, and point process convergence for a particular case.