A Hybrid Multi-Objective Genetic Algorithm for Topology Optimization

A new tool is developed in order to solve computationally expensive multi-objective topology optimization problems related to the design for flexible active and passive skins for morphing aircraft. The approach used is based on a multiobjective genetic algorithm coupled with a local search algorithm to create a hybrid multi-objective algorithm. The ability of the developed algorithm to find efficiently Pareto fronts of problems with two and three objectives is evaluated using three test problems. A multi-objective topology optimization problem related to the design of flexible skins is then solved as a proof of concept for more advanced models of flexible skins. It is shown that the hybrid multiobjective algorithm performs significantly better than the multi-objective algorithm it is based on. Additionally, the algorithm is able to solve a larger number of problems. The topology optimization results gives some very promising solution and the approach used lays the ground-work for more advanced topology optimizations.