Using Evolutionary Computation to explore geometry and topology without ground structures

Abstract

Over the past two decades there has been an increasing interest in using what has come to be called Evolutionary Computation (EC) in the analysis and optimization of structural systems. These methods include Genetic Algorithms (GA), Evolution Strategies (ES), Simulated Annealing and other stochastic based numerical methods. Each of these methods shares the drawback that they are very computationally intensive compared to deterministic methods. Furthermore, the computational burden can rapidly increase as the size of the analyzed structure increases. This paper suggests that the level of computation can be significantly reduced by avoiding the common practice of using ground structures in coding the topology. Additionally, comparative examples show that a broader range of good solutions can be reached when the use of ground structures is avoided.