engleski

Benchmark tests of a new fault tree engine BDDCUT

During the development of the BDD package our implementations have been tested thoroughly to ensure both correctness and efficiency. Many of the proposed data structures and algorithms were discarded as they turned out to be either inefficient or insufficient. In this paper we will discuss the efficiency of some of the implementations in the BDD package. Firstly we briefly describe the problems we have used in the test with the BDD package. In the next section we present the results of the tests we have performed, and in the last section we analyse the results. We will use fault trees, which come from the nuclear industry to examine the efficiency of the BDD package we have implemented. We will study the running time for calculation of the probability of top event from the BDD representing the fault tree we test, and after that the running time to do calculations of different configurations.

fault tree; probability; nuclear industry

nije evidentirano