A Multi-Modeling Strategy based on Belief Function Theory

In this paper, a multi-modeling strategy based on belief function theory is developed. The basic idea is to consider a fuzzy rule based system with a belief structure (BS) as output. The focal elements of each rule are formed by a subset of a collection of functional models. A particular attention is paid to the topic of combining global and local fuzzy models of Takagi-Sugeno type by a specific formulation of the proposed model. Some examples are given to show the validity of the approach.