Improving multi-objective genetic algorithms with adaptive design of experiments and online metamodeling

Applications of multi-objective genetic algorithms (MOGAs) in engineering optimization problems often require numerous function calls. One way to reduce the number of function calls is to use an approximation in lieu of function calls. An approximation involves two steps: design of experiments (DOE) and metamodeling. This paper presents a new approach where both DOE and metamodeling are integrated with a MOGA. In particular, the DOE method reduces the number of generations in a MOGA, while the metamodeling reduces the number of function calls in each generation. In the present approach, the DOE locates a subset of design points that is estimated to better sample the design space, while the metamodeling assists in estimating the fitness of design points. Several numerical and engineering examples are used to demonstrate the applicability of this new approach. The results from these examples show that the proposed improved approach requires significantly fewer function calls and obtains similar solutions compared to a conventional MOGA and a recently developed metamodeling-assisted MOGA.

[1]  Rajkumar Roy,et al.  Multi-objective Optimisation Of Rolling Rod Product Design Using Meta-modelling Approach , 2002, GECCO.

[2]  Kisuk Cheung.  Engineering and Design. Practical Aspects of Applying Geostatistics at Hazardous, Toxic, and Radioactive Waste Sites , 1997 .

[3]  T. Simpson,et al.  Efficient Pareto Frontier Exploration using Surrogate Approximations , 2000 .

[4]  Søren Nymand Lophaven,et al.  DACE - A Matlab Kriging Toolbox , 2002 .

[5]  Riccardo Poli,et al.  Particle swarm optimization , 1995, Swarm Intelligence.

[6]  Raphael T. Haftka,et al.  Design of stiffened composite panels by genetic algorithm and response surface approximations , 1995 .

[7]  James M. Douglas,et al.  Conceptual Design of Chemical Processes , 1988 .

[8]  K. S. Anderson,et al.  Genetic crossover strategy using an approximation concept , 1999, Proceedings of the 1999 Congress on Evolutionary Computation-CEC99 (Cat. No. 99TH8406).

[9]  Søren Nymand Lophaven,et al.  DACE - A Matlab Kriging Toolbox, Version 2.0 , 2002 .

[10]  James Kennedy,et al.  Particle swarm optimization , 1995, Proceedings of ICNN'95 - International Conference on Neural Networks.

[11]  Mian Li,et al.  Robust Optimization and Sensitivity Analysis with Multi-Objective Genetic Algorithms: Single- and Multi-Disciplinary Applications , 2007 .

[12]  Khaled Rasheed,et al.  Comparison of methods for developing dynamic reduced models for design optimization , 2002, Soft Comput..

[13]  Juan J. Alonso,et al.  Mutiobjective Optimization Using Approximation Model-Based Genetic Algorithms , 2004 .

[14]  Yaochu Jin,et al.  A comprehensive survey of fitness approximation in evolutionary computation , 2005, Soft Comput..

[15]  Thomas J. Santner,et al.  The Design and Analysis of Computer Experiments , 2003, Springer Series in Statistics.

[16]  J. -F. M. Barthelemy,et al.  Approximation concepts for optimum structural design — a review , 1993 .

[17]  Marc Schoenauer,et al.  Surrogate Deterministic Mutation: Preliminary Results , 2001, Artificial Evolution.

[18]  Ernest J. Henley,et al.  Separation Process Principles , 1998 .

[19]  Shapour Azarm,et al.  Metrics for Quality Assessment of a Multiobjective Design Optimization Solution Set , 2001 .

[20]  T. J. Mitchell,et al.  Bayesian Prediction of Deterministic Functions, with Applications to the Design and Analysis of Computer Experiments , 1991 .

[21]  G. Gary Wang,et al.  An Efficient Pareto Set Identification Approach for Multiobjective Optimization on Black-Box Functions , 2005 .

[22]  R. Haftka,et al.  Response surface approximations for structural optimization , 1996 .

[23]  S. Ruzika,et al.  Approximation Methods in Multiobjective Programming , 2005 .

[24]  C. Currin,et al.  A Bayesian Approach to the Design and Analysis of Computer Experiments , 1988 .

[25]  S. Azarm,et al.  Approximation of multiresponse deterministic engineering simulations: a dependent metamodeling approach , 2006 .

[26]  Kalyanmoy Deb,et al.  Computationally effective search and optimization procedure using coarse to fine approximations , 2003, The 2003 Congress on Evolutionary Computation, 2003. CEC '03..

[27]  Ivo F. Sbalzariniy,et al.  Multiobjective optimization using evolutionary algorithms , 2000 .

[28]  D. Lindley.  On a Measure of the Information Provided by an Experiment , 1956 .

[29]  Shapour Azarm,et al.  Optimizing thermal design of data center cabinets with a new multi-objective genetic algorithm , 2007, Distributed and Parallel Databases.

[30]  Jack P. C. Kleijnen,et al.  The correct Kriging variance estimated by bootstrapping , 2006, J. Oper. Res. Soc..

[31]  Margaret M. Wiecek,et al.  A Survey of Approximation Methods in Multiobjective Programming , 2003 .

[32]  Shapour Azarm,et al.  A Kriging Metamodel Assisted Multi-Objective Genetic Algorithm for Design Optimization , 2008 .

[33]  Donald R. Jones,et al.  A Taxonomy of Global Optimization Methods Based on Response Surfaces , 2001, J. Glob. Optim..

[34]  C. Geankoplis.  Transport processes and unit operations , 1978 .

[35]  Shapour Azarm,et al.  Constraint handling improvements for multiobjective genetic algorithms , 2002 .

[36]  Ren-Jye Yang,et al.  Approximation methods in multidisciplinary analysis and optimization: a panel discussion , 2004 .

[37]  Haym Hirsh,et al.  Informed operators: Speeding up genetic-algorithm-based design optimization using reduced models , 2000, GECCO.

[38]  Henry P. Wynn,et al.  Maximum entropy sampling , 1987 .