Evolutionary Structural Optimization—A Trial Review


Özkal F. M.

Nature-Inspired Metaheuristic Algorithms for Engineering Optimization Applications, Carbas Serdar,Toktas Abdurrahim,Ustun Deniz, Editör, Springer, London/Berlin , Singapore, ss.277-308, 2021

  • Yayın Türü: Kitapta Bölüm / Araştırma Kitabı
  • Basım Tarihi: 2021
  • Yayınevi: Springer, London/Berlin 
  • Basıldığı Şehir: Singapore
  • Sayfa Sayıları: ss.277-308
  • Editörler: Carbas Serdar,Toktas Abdurrahim,Ustun Deniz, Editör
  • Atatürk Üniversitesi Adresli: Evet

Özet

Since efficient material usage in a structure is an essential intention in engineering design, objective of a structural optimization algorithm should be obtaining the lightest and stiffest structure while considering specific design constraints. Evolutionary structural optimization (ESO) method has been introduced by Xie and Steven in 1993 to overcome structural optimization problems by improving and optimizing the design of structures. Although the algorithm depends on a plain inspiration, performance level of the ESO algorithm has increased owing to the contributions of wide-ranging studies in the subsequent years. This heuristic method has appeared on a simple—sometimes defined as intuitive—idea that is eliminating (or removing) the inefficient elements from the initial design region and attaining a fully stressed structure with an updated topology. Such a process is performed iteratively until an optimum layout is achieved, or until a specified volume is acquired; and the decision of process termination is executed over the feasibility of the designs achieved. This optimization algorithm is based on finite element method and has showed its superiorities by subsequent studies in various disciplines. Within this chapter, a brief on the basis and the main improvement steps of evolutionary structural optimization is presented. Different approaches on objectives, constraints, processing and terminating will be summarized. Thereafter, efficiency and capability of the method will be demonstrated over some examples.