Using fuzzy bases to resolve nonlinear programming problems (Q1595197): Difference between revisions

The concept of a fuzzy basis is introduced as a special fuzzy granulation of a given real interval. Theoretical properties of fuzzy bases are derived and the concept of a fuzzy linear basis is defined. It is shown how to use fuzzy linear bases for solving nonconvex optimization problems. In the further part of the article, fuzzy bases are used in fuzzy interpolation. It is shown that any continuous differentiable function defined on the one-dimensional real space can be approximated by fuzzy linear interpolation to any given accuracy. It is shown how to transform general nonconvex programming problems to a mixed integer bilinear form using fuzzy interpolation. If the objective function is convex, then the problem can be relaxed by making use of the fuzzy bases to an ordinary linear programming problem. Simplifying cutting methods for solving the nonlinear problems are suggested.