An efficient combination of multi-objective evolutionary optimization and reliability analysis for reliability-based design optimization of truss structures. (15th July 2018)

• Record Type: Journal Article
• Authors: Ho-Huu, V.; Duong-Gia, D.; Vo-Duy, T.; Le-Duc, T.; Nguyen-Thoi, T.
• Is Part Of: Expert systems with applications. Volume 102(2018)
• Page Start: 262
• Page End: 272
• Publication Date: 2018-07-15
• Subjects: Multi-objective evolutionary optimization -- Reliability-based design optimization (RBDO) -- Structural optimization -- Non-dominated sorting genetic algorithm II (NSGA-II) -- First order reliability analysis (FORM) -- Truss structures; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2018.02.040
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: Multi-objective evolutionary optimization and reliability analysis are combined. The proposed approach is applied to the optimal design of truss structures. A set of optimal solutions with different levels of reliability can be obtained. The algorithm is simple for engineering designers to understand and implement. The proposed approach is reliable and more competitive compared to other methods. Abstract: Over the past decades, the reliability-based design optimization of truss structures has been still a major challenge for engineering designers and even for researchers because of its complexities and high computational cost. In this paper, a new approach based on a novel combination of multi-objective evolutionary optimization and reliability analysis is proposed to deal with such kinds of problems. The proposed method consists of two separate steps. First, a multi-objective design optimization problem is formulated and solved by a multi-objective evolutionary optimization algorithm. Secondly, reliability analysis problems are formed by taking into account the uncertainty of input data of the problem, and a reliability analysis method is used to evaluate the reliability of all solutions obtained at the first step. Based on the obtained reliability, the suitable optimal solutions with a certain level of reliability can be readily identified. The proposed approach has some advantages such as: 1) it can give a set of many optimal solutions with different levels of reliability by only one run; 2) it can easily handle the design optimization problems of truss structures with continuous, discrete or mixed continuous-discrete design variables; 3) it is quite simple for engineering designers to understand and implement. To demonstrate the efficiency and applicability of the proposed method, three examples of truss structures are carried out, and the obtained results are compared to those available in the literature. The acquired results reveal that the proposed approach is reliable and more competitive compared to other methods.