| Popis: |
Ill-posed problems are usually solved by the transformation to minimization problems. They are ill-conditioned, then additional techniques, i.e. regularizations, are adopted to avoid the oscillation. In practical problems they are usually so complicated that it is not easy to adopt such effective techniques. Here two points should be focused on for practical problems. First, high accuracy is not necessary. Second, engineers who have much experience about the problems know how to deal with them qualitatively. These points suggest the validity of flexible minimizers. In this paper the fuzzy theory is introduced to construct such minimizers and it is applied to an ill-posed shape design problem. Numerical results are satisfactory. |
