Optimization of a hybrid solar air conditioner using efficient approximation assisted optimization technique.

Numéro : pap. n. 68

Auteurs : ALILI A. al-, SALEH K.

Résumé

Various optimization approaches have been used in the design of solar air conditioners (A/C)s. The solar A/C under investigation is a hybrid solar A/C which uses a hybrid photovoltaic/thermal collector to drive a hybrid A/C. The hybrid A/C consists of a solid desiccant wheel cycle and a conventional vapor compression cycle. The solar hybrid A/C is modeled using Transient Systems Simulation (TRNSYS) program which is coupled to MATLAB optimization toolbox in order to carry out the optimization study. However, one common problem faced in optimization is the high computational cost of the simulation. In the simulation of the hybrid solar A/C, each simulation run requires about half an hour of actual time and the global optimization algorithms usually requires hundreds of runs. This paper discusses one approach that can be used to deal with this issue by using approximation assisted optimization (AAO). This approach is found to use less number of function evaluations, hence less computational time, in order to find the Pareto solutions of the multi-objective optimization problem in hand. The paper discusses the steps needed to carry out the AAO namely: design variables’ space sampling, building the meta-models, meta-models’ verification, running the optimizer based on meta-models, and finally optimum solutions’ verification. In addition, the results of the AAO are compared to the Multi-Objective Genetic Algorithm (MOGA) solution. The AAO is able to find comparable results to MOGA while significantly reduce the computational time. It has been shown that this approach could reduce the computational time by more than 80% compared with conventional optimization approaches such as MOGA.

Documents disponibles

Format PDF

Pages : 9 p.

Disponible

  • Prix public

    20 €

  • Prix membre*

    15 €

* meilleur tarif applicable selon le type d'adhésion (voir le détail des avantages des adhésions individuelles et collectives)

Détails

  • Titre original : Optimization of a hybrid solar air conditioner using efficient approximation assisted optimization technique.
  • Identifiant de la fiche : 30014354
  • Langues : Anglais
  • Source : International sorption heat pump conference (ISHPC2014), College Park, United States, March 31-April 2, 2014.
  • Date d'édition : 02/03/2014

Liens


Voir d'autres communications du même compte rendu (125)
Voir le compte rendu de la conférence