Year: 2001
Type: article
Abstract: Many real-world problems, including the design problem, involve multiple and often conflicting objectives. Genetic Algorithm (GA) is an effective method to solve such multi-objective optimization problems. The designer usually has some information and knowledge based on his/her experience. This paper describes a new method for multi objective design problems using GA. which can integrate designer's vague preferences into GA search. First, a multi objective optimization problem with constraints is expressed as a satisficing problem of constraints by introducing an aspiration level for each objective. Furthermore, the unsatisfying function is introduced in order to handle fuzziness involved in aspiration level and constraint, and the problem is formulated as a multi objective minimization problem of unsatisfaction ratings. In order to obtain a group of Pareto optimal solutions in which the designer is interested, a new strategy based on tournament selection is proposed. Finally, the proposed method is applied to a simple numerical example and a design problem of a four bar plane truss.
Authors: Yasutaka TSUJI, Takanori KIYOTA, Eiji KONDO
Institution: Kyushu University
Language: English
Subfield: Computational Theory and Mathematics
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W2322339082
