抄録
Many techniques for searching the optimum of unimodal unknown function are discussed in nonlinear optimization problems and the mathematical programmings. On the other hand, though the search method for multimodal unknown function is an important technique in the optimization of systems, it is not discussed so sufficiently as the unimodal unknown function.
This paper describes a new method for searching the global maximum of multimodal unknown function. The method is composed of three kinds of searches. They are called G (Grasping)-mode search, F (Finding)-mode search and C (Confirming)-mode search. In the G-mode search and the C-mode search, a heuristic method is used which was extracted from search behaviors of human subjects. In F-mode search, the simplex method is used which is well-known as a search method for unimodal unknown function. Each mode search and its transitions are shown in the form of flowchart. The numerical results for one-dimensional through six-dimensional multimodal functions prove the proposed search method to be an effective one.
