Multi-Objective Evolutionary Optimization Algorithms for Machine Learning: A Recent Survey

2019 ◽  
pp. 35-55 ◽  
Author(s):  
Stamatios-Aggelos N. Alexandropoulos ◽  
Christos K. Aridas ◽  
Sotiris B. Kotsiantis ◽  
Michael N. Vrahatis
Keyword(s):  
Machine Learning ◽  
Optimization Algorithms ◽  
Evolutionary Optimization ◽  
Recent Survey ◽  
Multi Objective ◽  
Evolutionary Optimization Algorithms

Discovering Sequenced Origami Folding Through Nonlinear Mechanics and Topology Optimization

2019 ◽  
Vol 141 (4) ◽  
Author(s):  
Andrew S. Gillman ◽  
Kazuko Fuchi ◽  
Philip R. Buskohl
Keyword(s):  
Topology Optimization ◽  
Design Space ◽  
Optimization Algorithms ◽  
Element Model ◽  
Evolutionary Optimization ◽  
Nonlinear Mechanics ◽  
Design Problems ◽  
Mechanics Model ◽  
Highly Nonlinear ◽  
Evolutionary Optimization Algorithms

Origami folding provides a novel method to transform two-dimensional (2D) sheets into complex functional structures. However, the enormity of the foldable design space necessitates development of algorithms to efficiently discover new origami fold patterns with specific performance objectives. To address this challenge, this work combines a recently developed efficient modified truss finite element model with a ground structure-based topology optimization framework. A nonlinear mechanics model is required to model the sequenced motion and large folding common in the actuation of origami structures. These highly nonlinear motions limit the ability to define convex objective functions, and parallelizable evolutionary optimization algorithms for traversing nonconvex origami design problems are developed and considered. The ability of this framework to discover fold topologies that maximize targeted actuation is verified for the well-known “Chomper” and “Square Twist” patterns. A simple twist-based design is also discovered using the verified framework. Through these case studies, the role of critical points and bifurcations emanating from sequenced deformation mechanisms (including interplay of folding, facet bending, and stretching) on design optimization is analyzed. In addition, the performance of both gradient and evolutionary optimization algorithms are explored, and genetic algorithms (GAs) consistently yield solutions with better performance given the apparent nonconvexity of the response-design space.

Evolutionary Optimization Algorithms

2001 ◽  
pp. 97-116
Author(s):  
Luigi Fortuna ◽  
Gianguido Rizzotto ◽  
Mario Lavorgna ◽  
Giuseppe Nunnari ◽  
M. Gabriella Xibilia ◽  
...  
Keyword(s):  
Optimization Algorithms ◽  
Evolutionary Optimization ◽  
Evolutionary Optimization Algorithms
Sign in / Sign up

Export Citation Format

Share Document