Improved genetic algorithm for minimum thickness composite laminate design

Abstract

The use of a genetic algorithm for the minimum thickness design of composite laminated plates is explored. A previously developed genetic algorithm for laminate design is thoroughly revised and improved, by incorporating knowledge of the physics of the problem into the genetic algorithm. Constraints are accounted for by combining fixed and progressive penalty functions. Improved selection, mutation, and permutation operators are proposed. The use of an operator called scaling mutation that projects designs toward the feasible domain is investigated. The improvements in the genetic algorithm are shown to reduce the average price of a genetic search by more than 50%.