Doğrusal olmayan amaç programlamanın süreç optimizasyonu olarak stiren ünitesine uygulanması
Başlık çevirisi mevcut değil.
- Tez No: 18291
- Danışmanlar: PROF.DR. GÜNEŞ GENÇYILMAZ
- Tez Türü: Doktora
- Konular: İşletme, Business Administration
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 1992
- Dil: Türkçe
- Üniversite: İstanbul Üniversitesi
- Enstitü: Sosyal Bilimler Enstitüsü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 210
Özet
Bilindiği gibi günümüzde gittikçe karmaşık hale gelen sosyal, ekonomik ve teknolojik problemlerin çözümünde bilgisayarlarla birlikte matematiksel yöntemlerde yoğun bir biçimde kullanılmaktadır. Bu problemler içerisinde özellikle endüstriyel yapıda olanlar ağırlık kazanmaktadır. Bu acıdan, bu tip problemleri matematiksel olarak Sürekli ve Kesikli seklinde iki kısma ayırabiliriz. Bu tasnife göre, genellikle probleme üretim acısından bakıldığında sürekli ve kesikli üretim sistemlerinden bahsedilebilir. Yine bilindiği gibi sürekli üretim sistemlerinin konusunu ağırlıklı olarak kimyasal süreç endüstrileri oluşturmaktadır. Kimyasal süreç endüstrileri hem modeli eme hem optimizasyon ve hem de kontrol yönüyle kullanılan analitik ve sayısal çözüm yöntemleri yoğun bir ilgi alanı oluşturmaktadır. 1960 1 ardan beri bu alanda özellikle bir çok sayısal yöntem geliştirilmiştir. Bunlar matematiksel programlama baslığı altında toplanabilir. Diğer yandan, kimyasal bi r süreç problemini klasik optimizasyon teknikleri seklinde bir amaç fonksiyonu ve çok sayıda kısıtla ifade etmek çoğu kere pratik açıdan yetersiz olmaktadır. Bunun için, amaç fonksiyonunda yer alan kriterin veya ölçütün birden fazla olması problemi hem ekonomik ve hem de teknolojik acıdan incelememizi mümkün kılar. Bu çalışmada, bir petrokimya ünitesinin uyusan ve çatışan bir çok amacını kapsayacak teknolojik hedefleri yanında firmanın kısa ve uzun vadeli hedeflerini de kısıtlar kümesi seklinde gösterebilecek bir modelin kurulması daha uygun olacaktır. Yukarda bahsedilen ihtiyaca cevap verecek bir matematiksel programlama tekniği olarak Amaç Programlama* nm böyle bir üretim sistemine nasıl uygulandığı bu çalışmanın ana konusu olacaktır.Bu amaçla tez 6 ayrı.Bölüme ayrılmıştır. îlk Bölümde giriş düzeyinde SUrec
Özet (Çeviri)
In this Ph.D. Study, nonlinear programming technique has been applied to a petro-cbemical unit, which produces Styrene via Ethylbenzene Dehidregenation in Turkish Petro--Chemical Industry Co. Process optimisation concept or approach has been applied to process operation, which satisfy several absolute requirements* material balances, production capacities etc.) and conflicting objectives. This Ph.D. Thesis has been divied into 6 chapters. In the first chapter,the introductory process optimisation definition has been made with the method of nonlinear programming form and the determination of basic chemical data such as kinetics, material ba lances, energy changes and energy balances, which chemical process industries ( involve chemical conversion and physical operation in any chemical plant) rest. Optimisation techniques in a continious process (e.g. chemical process industries) may be viewed at different hierarchical levels; plant unit design, production planning, prcsdu.ct.ion schedulling and process operation. It has been described applicable optimisation techniques in steady-state dynamic process, which are linear programming, nonlinear programming and constrained optimisation techniques are very important. Another stage, modelling approach in general definition and in a chemical process has been presented lastly by multiple criteria decision making.In second chapter,' the importantce of styrene and its place in petro-chemical materials has been presented. The predominant route of the production of styrene is via ethylbenzene which alkylated from benzene with ethylene and subsequently clehidregenated over an aluminum chloride catalyst, solid phosphoric acid or silica alumina cata lists. This chemical! reaction has been presented scematically with a flowchart. Nonlinear modelling and optimisation articles in chemical processes have been summarized. Especially Griffith and Stewart's Method of Approximation Programming which is a' cliff erantial technique utilizing the linear programming algoritm has been used » Sengupta reported fixing the levels of inputs and process variables in order to meet a required specification of output. However, difficulties can arise when the output has a number of characterics which has to satisfy a specification. Such a problem was met in a. paper manufacturing plant and the problem was solved using Goal Programming. The third chapter was allocated to the Genaral Goal Programming Model, Nonlinear Models, Multiple Criteria Decision Making, Goal Programming Models Studies, Modified Pattern Search Method on Nonlinear Goal Programming. Most real-world decision problems involve multiple conflicting objectives. If we decide to use linear programming to solve decision problems with multiple objectives, we may introduce other objectives (other than the objective function) as model constraints» The linear programming model, however, requires that the optimum solution must satisfy allconstraints. Goal Programming is a linear mathematical model in which the optimum attainment of multiple goals is sought within the given decision environment. To solve nonlinear Goal Programming Problem a Simplex Based Nonlinear Goal Programming Algoritm is used However in this method each iteration employs the use of Modified Simplex Algoritm (and of course, the size of the problem has been larged so as to employ this technique. Another drawback includes that is one must employ either numerical or axact methods of differentiation in a performance of the algoritm. The method to be discussed n e xt, avoids these two drawbacks. This method is based on an extension (the extension itself, being very minor for this case) of the search method of Hooke and Jeeves or, as it often known, the Method of Pattern Search.Pattern Search is one of a class of search techniques as accelerated search method. Such method increase their search step size? if previous searches have been succesful and maintain or decrease the step sise otherwise. Prior to our discussion of the method it is necessary to list and explain the terminology that will be employed. We shall begin our search at our initial base point. This point, designated as X, may be selected by a number of means; arbitrarily, by an educated quess, or by using a present solution (if we B.re concerned about an existing system). The better our chaise of this point(i"e., the closer to the true optimal ), the faster will the search, in general converge.For convenience, this first base point will also be denoted as our first trial point. A search is conducted about, each trial paint», To form this search, we perturb each variable, one at a time, about present trial point. This perturbation is determine by selecting a step sise for each of the n variables? x, x,. n It is important to note that it is not necessary to examine all of the perturbations for. a given trial point. We stop as soon as we obtain a point giving a value of a that is prefered.to our previous best perturbation point, Articles and books based on survey related to goal programming and its extensions have been presented. In chapter 4, modelling and solution of the Styrene Unit as a process control problem. has been studied. In this step, firstly goal programming formulation of the general process control problem has been initiated. Secondly, nonlinear goal programming of this formulation of the Styrene Unit with given past data has been adapted. In the last chapters, five and six, with the use of written program, which is BASIC based, optimal solution has been obtained, Defination of the optimal solution gives minimum cost and satisfies input and output level of process. More obvious explenation is that optimal solution is 1,766.35a- TL/ton. In final chapter, results, solutions, discussions and proposals have been given. Sensitivity analysis of nonlinear goal programming with pre-emption to date have failed with the effects of changes in parameters after the optimalsolution has been obtained. It is known that post-optimal ity analysis nonlinear goal programming involvolves changes in priority factors , changes in w. for non-basic activities, changes w. for basic activities, overall rearrangement of priority structure and changes in resources or goal levels. We have applied changes in priority factors. It has been seen that if we change cost priority and value an aproach in determining limits or constraints of the variable related to working area to decision maker can be obtained >
Benzer Tezler
- Design and implementation of high power GaN amplifiers with nonlinear optimization techniques
Doğrusallaştırılmış yüksek güçlü GaN kuvvetlendiricilerin tasarımı ve gerçeklemesi
LIDA KOUHALVANDI
Doktora
İngilizce
2021
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiElektronik ve Haberleşme Mühendisliği Ana Bilim Dalı
PROF. DR. İSMAİL SERDAR ÖZOĞUZ
- Gün öncesi elektrik piyasası ve yan hizmetlerin birlikte optimizasyonu
Co-optimization of the day ahead electricity market andancillary services
RIDVAN MEMİŞ
Yüksek Lisans
Türkçe
2019
Endüstri ve Endüstri MühendisliğiTOBB Ekonomi ve Teknoloji ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
DOÇ. DR. KÜRŞAD DERİNKUYU
- An application of stochastic programming on robust airline scheduling
Rassal programlamanın dayanıklı havayolu çizelgeleme üzerinde uygulanması
NİL KARACAOĞLU
Yüksek Lisans
İngilizce
2014
Endüstri ve Endüstri Mühendisliğiİhsan Doğramacı Bilkent ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
PROF. DR. MEHMET SELİM AKTÜRK
PROF. DR. HANDE YAMAN PATERNOTTE
- La selection des inducteurs de cout dans la methode ABC avec la methode Zionst-Wallenius et programmation de but
Faaliyet tabanlı maliyetlendirmede Zionst-Wallenius yöntemi ve hedef programlama ile maliyet sürücü seçimi
ABDULLAH ÇAĞRI TOLGA
Yüksek Lisans
Fransızca
2003
Endüstri ve Endüstri MühendisliğiGalatasaray ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
PROF. DR. E. ERTUĞRUL KARSAK
- Gestion de performance avec DEA en utilisant les extensions de controle des poids, d'efficacite croisee et d'analyse de fenetre
Ağırlık kısıtlamaları, çapraz etkinlik ve pencere analizi yöntemleri kullanılarak veri zarflama analizi ile performans yönetimi
FİRÜZAN İŞCAN
Yüksek Lisans
Fransızca
1999
Endüstri ve Endüstri MühendisliğiGalatasaray ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
DOÇ. DR. E. ERTUĞRUL KARSAK