Parametric Fuzzy Model Identification Based on a Hybrid Ant Colony Algorithm
I. A. Khodashinsky, P. A. Dudin
Pages: 24-35
Abstract
Applying the ant colony algorithm for solving the problem of parametric fuzzy model identification is presented. Transition from continuous optimization to discrete one via constructing a complete oriented decision-search graph is determined. A gradient algorithm is considered as the second optimization step. Experiments for analyzing the performance of the algorithms for optimization and fuzzy system are described.
