Optimization techniques based on swarm theory (particle swarm optimization, PSO), originally developed to simulate the social behavior of animals, are non linear methods belonging to the class of evolutionary computational techniques, like genetic algorithm (GA). Both of them are population-based optimizers that find solution through a probabilistic search process only guided by fitness value. In the particle swarm approach, however, neither individuals (or particles) are replaced nor new offspring are generated during evolution: individuals stay alive and simply change their position within the search space during the optimization process. Such position change (named velocity) 9is guided by personal experience of each particle and by swarm experience shared among all the individuals. To this end, each particle keeps a memory of its own best position as well as of the best position reached so far by all the particles. Following the PSO approach we can get population evolution trough cooperation rather than competition among individuals. The main advantage of the PSO algorithm is its simple implementation compared to GA: besides velocity, no additional operator is requested in PSO, whereas GA needs three different operators (Selection, cross-over, mutation). Many reports compare particle swarm to genetic algorithm showing, in some cases, that PSO can give better results [2,3]. For this reason, PSO applicability to an aircraft conceptual design as been deemed an interesting problem to be investigated. The definition of a preliminary aircraft configuration is, in fact, a critical task for deterministic approaches, whereas probabilistic-type methods (e.g. GA) have already proved their effectiveness [4,5].

Particle Swarm approach in finding optimum aircraft configuration

DEL CORE Giuseppe;
2007

Abstract

Optimization techniques based on swarm theory (particle swarm optimization, PSO), originally developed to simulate the social behavior of animals, are non linear methods belonging to the class of evolutionary computational techniques, like genetic algorithm (GA). Both of them are population-based optimizers that find solution through a probabilistic search process only guided by fitness value. In the particle swarm approach, however, neither individuals (or particles) are replaced nor new offspring are generated during evolution: individuals stay alive and simply change their position within the search space during the optimization process. Such position change (named velocity) 9is guided by personal experience of each particle and by swarm experience shared among all the individuals. To this end, each particle keeps a memory of its own best position as well as of the best position reached so far by all the particles. Following the PSO approach we can get population evolution trough cooperation rather than competition among individuals. The main advantage of the PSO algorithm is its simple implementation compared to GA: besides velocity, no additional operator is requested in PSO, whereas GA needs three different operators (Selection, cross-over, mutation). Many reports compare particle swarm to genetic algorithm showing, in some cases, that PSO can give better results [2,3]. For this reason, PSO applicability to an aircraft conceptual design as been deemed an interesting problem to be investigated. The definition of a preliminary aircraft configuration is, in fact, a critical task for deterministic approaches, whereas probabilistic-type methods (e.g. GA) have already proved their effectiveness [4,5].
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11367/16632
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 16
  • ???jsp.display-item.citation.isi??? 6
social impact