Review of modern optimization methods for bearing systems of buildings аnd structures

• Main Authors: Ashot G. Tamrazyan, Anatoly V. Alekseytsev
• Format: Article
• Language: English
• Published: Moscow State University of Civil Engineering (MGSU) 2020-01-01
• Series: Vestnik MGSU
• Subjects: optimization methods; evolutionary modeling; metaheuristic search; mathematical programming; survivability optimization; risk; reliability; safety; emergencies; life cycle
• Online Access: https://doi.org/10.22227/1997-0935.2020.1.12-30

Introduction. A review of modern approaches to the search for design solutions, optimal by various criteria, for load-bearing construction structures is given. The paper discloses problems of finding solutions on discrete sets of variable parameters reflecting modern approaches to design considering the available range of materials and products. It also formulated the goals of the literary review, its relevance to modern socio-economic conditions, and a brief historical digest reflecting the stages of development of search methods. The focus of the review is on new metaheuristic approaches that are used in world practice to optimize the load-bearing systems. Materials and methods. The methods of searching for relevant literature in international scientific databases are described. The selection of sources was based on the international Web of Science databases by publishing house Thomson Reuters, Scopus, EI Compendex by Elsevier publishing house which, had well-developed tools for thematic search as well as guaranteeing the legitimacy and quality of the literature. Results. The article classified the modern methods of searching for optimal design solutions by several criteria. Those included the metaheuristic approach type, number of optimization goals, methods of considering constraints, presentation of varying parameters, consideration of life cycle stages, survivability, reliability, facility safety levels, and accident risks. A series of the most promising optimization methods based on the mathematical formalization of physical processes, as well as combined iterative strategies, are described. Attention is also paid to some gradient approaches and methods of mathematical programming. Conclusions. The paper has shown that to meet modern requirements for building load-bearing structures, including safety criteria, minimizing costs during the life cycle, reliability, survivability, and other properties, the methods similar to biological processes in wildlife are most widely spread. These methods have a relatively fast adaptability to solve a broad class of problems in construction. Analysis of the reviewed sources showed that the far-reaching methods for solving problems of construction design optimization are stochastic methods using discrete sets of design parameters that reflect features of building structure design.