Publication: Metasezgisel Yöntemlerle Proje Çizelgeleme Optimizasyonu
Abstract
Günümüzün refahını oluşturan tüm sanayi ürün ve üretim araçları, çeşitli nitelikteki başarılı projelerin sonuçlarıdır. Başarılı bir projenin en temel özelliği, iyi bir planlama sürecinin ardından mükemmel bir icra ve kontrolle sonuçlandırılmasıdır. Planlama sürecini oluşturan en temel unsur ise, projedeki faaliyetlerin zaman bakımından çizelgelenmesi ve kaynak dağıtımıdır. Bu çalışma ile planlama sürecine katkıda bulunacağı düşünülen proje çizelgeleme ve kısıtlı kaynakların faaliyetlere dağıtımı için, yeni yöntemler olarak bilinen metasezgisel yöntemler incelenerek bir uygulama geliştirilmiştir. NP-tam bir problem tipi olan kaynak kısıtlı proje çizelgeleme problemleri dağıtık arama ve genetik algoritmalarla çözülmüştür. Çözüm aşamasında olurlu çözüm yapısını bozmayan mutasyon ve çaprazlama operatörleri tasarlanarak genetik algoritmada kullanılmıştır. Olurlu bir populasyon üzerinde arama yapan benzer tabanlı iki algoritma gerek çalışma prensibi, gerekse sonuçları bakımından kıyaslanmıştır. Aynı konfigürasyona sahip bir bilgisayarla yapılan çözümler, iki algoritmanın objektif olarak kıyaslanabilme imkanını sunmuştur.
PROJECT SCHEDULING OPTIMIZATION WITH METAHEURISTIC METHODS The welfare of present-day is constituted by the outcome of the succesfull projects, in which entire products and production instruments are produced. A good planning, which should be followed by a complete execusion and control is the basic characteristic of the succesfull projects. The basis of the planning process includes, scheduling of the Project activities in a time horizon and resource allocation. With this study a survey of the metaheuristics and an application with metaheuristics is presented, which are considered as new methods to contribute project scheduling and resource allocation. Project scheduling problems, known as NP-hard problems, are solved by scatter search algorithms and genetic algorithms. In solution phase a crossover and mutation operator is developed, which does not degenerates the feasible structure of the solution. Two algorithms working with the similar substructures are compared according to their working principles and results. This study also provided us an objective comparison since both of the algorithms are performed on the same computers with the same configuration.
PROJECT SCHEDULING OPTIMIZATION WITH METAHEURISTIC METHODS The welfare of present-day is constituted by the outcome of the succesfull projects, in which entire products and production instruments are produced. A good planning, which should be followed by a complete execusion and control is the basic characteristic of the succesfull projects. The basis of the planning process includes, scheduling of the Project activities in a time horizon and resource allocation. With this study a survey of the metaheuristics and an application with metaheuristics is presented, which are considered as new methods to contribute project scheduling and resource allocation. Project scheduling problems, known as NP-hard problems, are solved by scatter search algorithms and genetic algorithms. In solution phase a crossover and mutation operator is developed, which does not degenerates the feasible structure of the solution. Two algorithms working with the similar substructures are compared according to their working principles and results. This study also provided us an objective comparison since both of the algorithms are performed on the same computers with the same configuration.