This article provides two robust and effective inverse design procedures based on fuzzy logic. The procedures benefits from the fuzzy logic non-analytical structure, which allows one to handle numerical map and data measurements, exploiting the designers' knowledge distributed in already existing systems. Based on such data, the fuzzy logic algorithm is able to promptly provide a first glance design which represents a good starting solution. The two proposed design algorithms differ in the optimisation process which improves the first glance design. The first one considers the optimisation as a false control problem, solved by,in adaptive control strategy which, treating the design error its the control error, tries to annihilate it by acting on the design variables. The second algorithm is based on it multiagent structure that, starting from the first glance design, performs the subsequent optimisation step by mapping, through it fuzzy logic structure, the relationship between the variation of the design variables and the design error. The effectiveness of the proposed methods is verified performing the design of it nozzle in different flow regimes, using both 1- and 2D flow solvers to perform the analyses.
Aerodynamic inverse design using fuzzy logic / Dambrosio, Lorenzo; Pascazio, Giuseppe; Semeraro, S.. - In: INVERSE PROBLEMS IN SCIENCE & ENGINEERING. - ISSN 1741-5977. - 16:2(2008), pp. 249-268. [10.1080/17415970701434083]
Aerodynamic inverse design using fuzzy logic
DAMBROSIO, Lorenzo;PASCAZIO, Giuseppe;
2008-01-01
Abstract
This article provides two robust and effective inverse design procedures based on fuzzy logic. The procedures benefits from the fuzzy logic non-analytical structure, which allows one to handle numerical map and data measurements, exploiting the designers' knowledge distributed in already existing systems. Based on such data, the fuzzy logic algorithm is able to promptly provide a first glance design which represents a good starting solution. The two proposed design algorithms differ in the optimisation process which improves the first glance design. The first one considers the optimisation as a false control problem, solved by,in adaptive control strategy which, treating the design error its the control error, tries to annihilate it by acting on the design variables. The second algorithm is based on it multiagent structure that, starting from the first glance design, performs the subsequent optimisation step by mapping, through it fuzzy logic structure, the relationship between the variation of the design variables and the design error. The effectiveness of the proposed methods is verified performing the design of it nozzle in different flow regimes, using both 1- and 2D flow solvers to perform the analyses.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.