학술논문

Hierarchical genetic algorithm for non destructive thermal evaluation of layered composite materials
Document Type
Conference
Source
IEEE International Conference on Systems, Man and Cybernetics Systems, man and cybernetics Systems, Man and Cybernetics, 2002 IEEE International Conference on. 7:6 pp. vol.7 2002
Subject
Robotics and Control Systems
Computing and Processing
Genetic algorithms
Composite materials
Inverse problems
Optimization methods
Heat transfer
Thermal conductivity
Sun
Algorithm design and analysis
Solids
Nonhomogeneous media
Language
ISSN
1062-922X
Abstract
Deals with the estimation of the internal thermal properties distribution in a layered composite material. Our approach, based on a flash method considers only boundaries measurements. The special nature of the diffusion makes this inverse problem particularly difficult, in addition, they are often ill posed. So we present a study based on a global optimisation method from a real coded hierarchical genetic algorithm. Numerical results with the method indicate that close estimates of the unknown distribution can be obtained on highly discontinuous profiles.