Abstract
An innovative neural fuzzy system is considered for possibilistic reliability using a neural tree structure with nodes of neuronal type. The total tree structure works effectively as a fuzzy logic system where the possibility theory plays important role with Gaussian possibility distribution at the nodes. The structure of the tree is determined by domain knowledge and each node represents a component of the system of concern. The reliabilities of the nodes are dependent on the reliabilities of the preceding nodes. The relationships among the nodes form the core of possibilistic reliability. For each input reliability composition the status of the system is known and interpreted not only at the system output but also at the granulated level at the system sub-domains which are represented by node outputs. The research is described in detail and a demonstrative computer experiment is reported.
