.png){kind=icon}
.png){kind=icon}
Return
Feasibility-ratio based sequencing for computationally efficient constrained optimization
.png){kind=icon}
OA
AI
DOI：10.1016/j.swevo.2021.100850.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Real world optimization problems typically involve constraints that must be satisfied for a design to be viable. Such constraints could represent physical limitations on induced behavior such as stress, displacement, temperature etc., or other practical limitations such as cost, geometry, manufacturability, etc. Constraint handling is therefore an active area of research, and a number of approaches have been proposed in the literature to be used with evolutionary algorithms to solve constrained optimization problems. Most approaches assume that the objective and constraints are simultaneously evaluated, and counted together as one evaluation as far as computation budget is concerned. However, in practical applications, there could be situations where the constraints and the objective can be independently evaluated (e.g., using different analyses). Therefore, it is worthwhile exploring if one can forego evaluation of certain constraints or the objective to save computational cost without misleading the search owing to insufficient information. Towards this end, the main focus of this study is to deal with a partial evaluation paradigm. We propose a feasibility-ratio based sequencing for partial evaluation of solutions during the evolutionary search. A lexicographic ranking approach is used for ordering the partially evaluated solutions. To balance the evaluation cost and information gain near constraint boundaries, a dynamic feasibility control technique is incorporated. A number of variants are systematically constructed to demonstrate the efficacy of partial evaluation and proposed sequencing approach. An extensive numerical study is conducted to evaluate the approach on a range of constraint problems, which confirms its competence.
Keywords:
Constrained optimization
Partial evaluation
Feasibility ratio based sequencing
Single-objective optimization
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
8.5
Papers:
2.1K
Citations:
1.0W
Organization
No organization information available