A hybrid particle swarm-genetic algorithm optimization scheme was combined with Nelder and Mead's simplex method to produce an experimental hybrid method that integrates the global search capabilities of evolutionary heuristics with the highly effective localized search ability of the simplex method. This algorithm was applied to various missile design problems where the goal is to design a missile that can match the given flight trajectory. Three different flight trajectories were analyzed for this work: a 117 km trajectory, a 164 km trajectory, and a 221 km trajectory. The efficiency and effectiveness of the performance of the simplex method was compared with the performance of the same algorithm using the pattern search method as the local search routine in lieu of the simplex method. Optimizer performance was evaluated according to two criteria: (1) “fitness function” accuracy (how closely solutions are able to match a given objective function), and (2) convergence speed (how many calls to the objective function are required in order to reach an acceptable solution). The pattern search proved to be a more robust and reliable algorithm for such a large and complex solution space, as it was always able to avoid getting lost in the solution space. Conversely, the simplex method would occasionally become trapped around a bad solution, typically while traversing regions of the solution space sensitive to discontinuities formed by mixed-variable conversions.
Missile system design using a hybrid Evolving Swarm Algorithm
2014-03-01
1933758 byte
Conference paper
Electronic Resource
English
Missile Guidance Algorithm Design Using Particle Swarm Optimization
Tema Archive | 2013
|Ramjet Powered Missile Design Using I Genetic Algorithm
British Library Conference Proceedings | 2004
|Ramjet Powered Missile Design Using a Genetic Algorithm
AIAA | 2004
|