Performance Analysis of DE-Sigma Optimization Method
Open Access
Author:
Quinn, John
Area of Honors:
Aerospace Engineering
Degree:
Bachelor of Science
Document Type:
Thesis
Thesis Supervisors:
Robert G. Melton, Thesis Supervisor Mark David Maughmer, Thesis Honors Advisor
Keywords:
Optimization Particle Swarm Optimization PSO Differential Evolution DE Differential Evolution Sigma DE Sigma
Abstract:
The modified heuristic method Differential Evolution Sigma (DE-Σ) could be a powerful tool for mission optimization but requires more research in algorithm parameter permutations and problem applications. This paper will analyze DE-Sigma's performance in two problems: Finding a Lyapunov orbit in the Earth Moon System (Lyapunov Problem) and finite thrust between coplanar circular orbits (FTX Problem). Within these problems, the parameters of the algorithm (weighting function exponent 'm', trial population fraction 'phi', and crossover value 'CR') are varied to see how the final solution is affected. Particle Swarm Optimization (PSO) is used as a control value for comparison to DE-Sigma. Performance is measured by successful convergence.
