%0 DATA
%A Lisa, Jackson
%A J.D., Andrews
%D 2008
%T Choosing a heuristic for the “fault tree to binary decision diagram” conversion, using neural networks
%U https://repository.lboro.ac.uk/articles/Choosing_a_heuristic_for_the_fault_tree_to_binary_decision_diagram_conversion_using_neural_networks/9224951
%2 https://repository.lboro.ac.uk/ndownloader/files/16804442
%K Binary decision diagrams
%K Fault-tree analysis
%K Neural networks
%K Variable ordering heuristics
%X Fault-tree analysis is commonly used for risk assessment
of industrial systems. Several computer packages are
available to carry out the analysis. Despite its common usage there
are associated limitations of the technique in terms of accuracy
and efficiency when dealing with large fault-tree structures. The
most recent approach to aid the analysis of the fault-tree diagram
is the BDD (binary decision diagram). To use the BDD, the
fault-tree structure needs to be converted into the BDD format.
Converting the fault tree is relatively straightforward but requires
that the basic events of the tree be ordered. This ordering is
critical to the resulting size of the BDD, and ultimately affects
the qualitative and quantitative performance and benefits of
this technique. Several heuristic approaches were developed to
produce an optimal ordering permutation for a specific tree. These
heuristic approaches do not always yield a minimal BDD structure
for all trees. There is no single heuristic that guarantees a minimal
BDD for any fault-tree structure. This paper looks at a selection
approach using a neural network to choose the best heuristic from
a set of alternatives that will yield the smallest BDD and promote
an efficient analysis. The set of possible selection choices are 6
alternative heuristics, and the prediction capacity produced was
a 70% chance of the neural network choosing the best ordering
heuristic from the set of 6 for the test set of given fault trees.