Citation
Abstract
One common way of estimating the failure probability p of a system is to simulate the system until a fixed number, say r, failures are observed and then set p = r/N, where N is the number of trials required. In this paper this technique is analyzed, and, in particular, the relationship between r and the reliability of the estimator p is studied. Pip<p<p)=1-y (2)
Details
- Volume
- 42-26
- Published
- April 15, 1975
- Pages
- 155–167
- File Size
- 981.1 KB