Multiobjective optimization of aerospace structures /

"This report presents the multiobjective optimization methodology suitable to the preliminary design of large scale aircraft structures. The algorithm is based on generalized compound scaling techniques to reach the intersection of multiple functions. Multiobjective functions are treated similar to the behavior constraints by generating pseudo targets on objectives. This algorithm is very efficient due to the fact that it does not solve many single objective optimization problems in reaching the Pareto set. The Pareto set gives several optimum solutions, and the designer has to use additional decision criteria in selecting the best design. In this work a reliability index based concept is used as the decision criterion. System reliability of the Pareto designs is calculated due to the random nature of material properties, fabrication tolerances, and service loads, where the random parameters influencing the design appear through their mean and covariance values."--Report documentation page.Cover title.Final report for period: January 1990 - December 1991.""July 1992.""WL-TR-92-3052."Includes bibliographical references (pages 114-120).Ch. 1. Introduction. -- Ch. 2. Finite element formulation. -- Ch. 3. Sensitivity analysis. -- Ch. 4. Optimization algorithm. -- Ch. 5. Numerical results and discussions. --Ch. 6. Conclusions."This report presents the multiobjective optimization methodology suitable to the preliminary design of large scale aircraft structures. The algorithm is based on generalized compound scaling techniques to reach the intersection of multiple functions. Multiobjective functions are treated similar to the behavior constraints by generating pseudo targets on objectives. This algorithm is very efficient due to the fact that it does not solve many single objective optimization problems in reaching the Pareto set. The Pareto set gives several optimum solutions, and the designer has to use additional decision criteria in selecting the best design. In this work a reliability index based concept is used as the decision criterion. System reliability of the Pareto designs is calculated due to the random nature of material properties, fabrication tolerances, and service loads, where the random parameters influencing the design appear through their mean and covariance values."--Report documentation page.Research supported by the United States Air Force Flight Dynamics Directorate, and performed by the Department of Mechanical and Materials Engineering, Wright State University, Dayton,Ohio, under contact no.Mode of access: Internet