COMPARISON OF HEURISTIC CONVERGENCE STRATEGIES FOR MULTIDISCIPLINARY ANALYSIS

This work develops and compares heuristic convergence strategies for complex, coupled, multidisciplinary analysis. A convergence strategy is a method for ordering the execution of the subproblems in a multidisciplinary analysis to produce a convergence. The aim of this investigation is to provide guidance on which strategies to use in particular multidisciplinary design situations to minimize the time or cost of a system-level optimization. The analysis process that would be used in the multidisciplinary-feasible (MDF) optimization process is simulated using systems generated with the CASCADE methodology. Different convergence strategies are tested by collecting time and cost information over large numbers of such systems and using statistical techniques to compare the results. Primary emphasis is given to parallel strategies as might be used in a cooperative design environment. The results have demonstrated a profound impact of strategy choice on the cost and time of the convergence of a multidisciplinary analysis. They have also shown the importance of sequencing as used in some of the strategies. Results for the parallel strategies demonstrate an important time/cost trade-off.

[1]  James L. Rogers.  A knowledge-based tool for multilevel decomposition of a complex design problem , 1989 .

[2]  Christina Bloebaum,et al.  Optimal sequencing for complex engineering systems using genetic algorithms , 1994 .

[3]  Ilan Kroo,et al.  A Genetic Algorithm for Scheduling and Decomposition of Multidisciplinary Design Problems , 1996 .

[4]  John E. Dennis,et al.  Problem Formulation for Multidisciplinary Optimization , 1994, SIAM J. Optim..

[5]  Christina Bloebaum,et al.  A COMPARISON OF SOLUTION STRATEGIES FOR SIMULATION-BASED MULTIDISCIPLINARY DESIGN OPTIMIZATION , 1998 .

[6]  Christina Bloebaum,et al.  Application of global sensitivity equations in multidisciplinary aircraft synthesis , 1990 .

[7]  Christina Bloebaum,et al.  Complex system design task sequencing for cost and time considerations , 1996 .

[8]  Tien-Chien Chang,et al.  An approach to object-oriented discrete-event simulation of manufacturing systems , 1991, 1991 Winter Simulation Conference Proceedings..

[9]  Jaroslaw Sobieszczanski-Sobieski,et al.  Multidisciplinary Design Optimization: An Emerging New Engineering Discipline , 1995 .

[10]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[11]  Christina Bloebaum,et al.  Development of a multidisciplinary design optimization test simulator , 1997 .

[12]  L. Darrell Whitley,et al.  A Comparison of Genetic Sequencing Operators , 1991, ICGA.

[13]  Christina Bloebaum,et al.  Ordering design tasks based on coupling strengths , 1994 .

[14]  Eyal Arian,et al.  Convergence Estimates for Multidisciplinary Analysis and Optimization , 1997 .

[15]  Jaroslaw Sobieszczanskisobieski,et al.  On the sensitivity of complex, internally coupled systems , 1988 .

[16]  Christina Bloebaum,et al.  Coupling strength-based system reduction for complex engineering design , 1995 .

[17]  Christina Bloebaum,et al.  Integrating a Genetic Algorithm Into a Knowledge-Based System for Ordering Complex Design Processes , 1996 .

[18]  Christina Bloebaum,et al.  A genetic tool for optimal design sequencing in complex engineering systems , 1996 .

[19]  Christina Bloebaum,et al.  Total derivative-based coupling suspension for system reduction in complex design , 1996 .