An adaptive response surface method for engineering analysis
Başlık çevirisi mevcut değil.
- Tez No: 400341
- Danışmanlar: CHRIS MCMAHON
- Tez Türü: Doktora
- Konular: Makine Mühendisliği, Mechanical Engineering
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 2000
- Dil: İngilizce
- Üniversite: University of Bristol
- Enstitü: Yurtdışı Enstitü
- Ana Bilim Dalı: Makine Mühendisliği Ana Bilim Dalı
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 232
Özet
Özet yok.
Özet (Çeviri)
Uncertainty evaluation in structural analysis has been gaining more interest recently, and several methods have been developed to carry out the evaluation. Although these methods have been applied to various problems in different engineering fields varying from aerospace to coastal engineering, the conventional methods for structural reliability analysis have the limitation that they are computationally expensive, to the extent that substantial simplifications have to be applied, and for some problems it is impossible to apply the methods.Approximation methods such as the Response Surface Method (RSM),in which an approximation function is used to model the performance of the structure under study, can be used to alleviate performance limitations. For structural reliability analysis, the common approach in the RSM is to use a second-degree polynomial as a response surface function, but in many reliability problems this may not be the best choice. In the present thesis an adaptive response surface method, called ADAPRES, is developed, in which the formation of the response surface function starts with a linear function and higher terms are added according to an assessment of modelling error.Uncertainty evaluation of a real engineering system often involves many different analyses with the output from one analysis forming the input for another, and interaction and communication between analyses is crucial to the overall evaluation. For each analysis, a response surface function can be used in place of the actual performance function. In order to carry out different analyses efficiently, formation of the response surface function for each analysis step can be carried out concurrently as well as being distributed to computing workstations in a network provided that the problem permits. Therefore, a network tool, called ADAPRES_NET, has been developed, by which the time consuming evaluation of the performance function can be distributed on computing workstations over a network as well as handling communication and data transformation between necessary applications. ADAPRES_NET models the dynamic relationship between computing processes using Petri nets.The applicability and efficiency of ADAPRES and ADAPRES_NET have been demonstrated for example analyses taken from the literature, and for a case problem which includes the optimisation of a connecting rod subjected to uncertain problem parameters. The example applications show that ADAPRES can give a reduction in the computational effort required while leading to equally or more satisfactory results compared with standard techniques. The suitability of the approach for engineering structural reliability assessment is evaluated, and suggestions are made for its future development.
Benzer Tezler
- Araç çatısının ezilmesi durumunda B-sütunu tasarım optimizasyonu
B-pillar design optimization in case of crushing of vehicle roof
TAYFUN BAŞKARA
Yüksek Lisans
Türkçe
2022
Mühendislik BilimleriPamukkale ÜniversitesiOtomotiv Mühendisliği Ana Bilim Dalı
DR. ÖĞR. ÜYESİ İSMAİL ÖZTÜRK
- Asenkron motor için adaptif kesirli kayan kipli gözlemci tasarımı
Adaptive fractional order sliding mode observer design for induction motor
ERDEM İLTEN
Doktora
Türkçe
2019
Elektrik ve Elektronik MühendisliğiBalıkesir ÜniversitesiElektrik-Elektronik Mühendisliği Ana Bilim Dalı
PROF. DR. METİN DEMİRTAŞ
- Bir binek araca ait fren diskindeki hataların titreşim parametreleri ile tespiti
Determination of defects on the passenger car brake disc using vibration parameters
ERTUĞRUL İLHAN
Yüksek Lisans
Türkçe
2019
Makine MühendisliğiBursa Teknik ÜniversitesiMakine Mühendisliği Ana Bilim Dalı
DOÇ. DR. HAKAN GÖKDAĞ
- Uyarlamalı süzgeçler
Adaptive filters
RIDVAN AYSEL
Yüksek Lisans
Türkçe
1994
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiPROF.DR. AHMET H. KAYRAN
- Enabling dynamics in face analysis
Başlık çevirisi yok
HAMDİ DİBEKLİOĞLU
Doktora
İngilizce
2014
Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve KontrolUniversiteit van AmsterdamPROF. DR. THEO GEVERS
PROF. DR. A. W. M. SMEULDERS