Proceedings
home
preface
contents
authors
keywords
copyright
reference ©2012 Civil-Comp Ltd

Paper 40

Local and Global Pareto Dominance in Multi-Populations Evolution applied to Optimization of Composites

C.A.C. António
IDMEC, Faculty of Engineering, University of Porto, Portugal

Keywords: multi-objective, co-evolution, non-domination, self-adaptive, age control, composites.

full paper (pdf) - reference

A multi-objective hierarchical genetic algorithm (MOHGA) with age structure and based on local and global dominance concepts to deal with multi-objective optimization of composite structures is proposed. The approach based on multi-populations evolution uses the species conservation technique to address the optimal stacking sequence and material topology of composite structures in multi-objective optimization problems. Thus, individuals corresponding to the same material selection and topology of the hybrid composite structure belong to the same species. Material distribution of the hybrid composite structures is performed at two levels: laminate level and structural topology level.

The goal of the proposed multi-objective optimization algorithm is to search for the Pareto-optimal front keeping population diversity and using a hierarchical genetic algorithm with co-evolution of multi-populations. A self-adaptive genetic search incorporating Pareto dominance and elitism is presented. The structure in multi-populations enables the concurrent search of an elitist strategy together with species conservation paradigm supported by migration rules between populations. Two concepts of dominance are used: the first one denoted by local non-dominance is performed at the isolation stage of populations and the second one called global non-dominance is performed on the age structured population. The age structured population is used to store the ranked solutions with the aim of obtaining the Pareto front at the end of the evolutionary process.

The application to composite structures deals with the compromise between maximum performance versus minimum strain energy and minimum weight or cost based on a shared fitness calculated on the Pareto front. The trade-off between the performance target, depending on given stress, displacement and buckling constraints imposed on composite structures, against robustness, is searched. The design variables are associated with the sizing and material distribution of composite structures with stiffeners. The constraints are imposed on the critical load factor and on the critical displacement associated with buckling and first ply failure. The results show that MOHGA is promising for multi-objective optimization of composite structures.