Akademik Veri Yönetim Sistemi
Tezin Türü: Doktora
Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, İnşaat Mühendisliği Bölümü, Türkiye
Tezin Onay Tarihi: 2015
Öğrenci: GENÇER GENÇOĞLU
Danışman: NURİ MERZİ
Özet:From design stage to operation, water distribution networks hold the vital part of the infrastructure systems. Pumps contain significant operational consumptions by pumping and maintenance costs. Wise scheduling is an important tool that may decrease these costs. Leakage, defined as a portion of treated and pumped water; and through a typical water distribution network, it may possess a large percentage of the total water supplied. Pressure heads within the water distribution network are one of the major causes of leakage. This study is a combination of pump scheduling and excess pressure minimization of the same network. The methodology is divided into (i) pump schedule optimization and (ii) valve locating and opening determination. For pump scheduling studies, objective is to reduce operational costs of pumps. On the otherhand, for valve locating and opening determination studies, objective is minimizing excess pressures. Genetic algorithms are applied on both optimization problems, with specific objective and penalty functions. Through pump scheduling studies, the schedules casting the energy usage, tank volume periodicity and pump switches are obtained. Real time daily demand measurements and predictions are integrated to the developed program to model unexpected cases. Kalman Filter is applied on the improved demand prediction model. The results indicate that the developed program is applicable to the network for real time scenarios. Throughout valve location and opening determination studies, the groups of open and closed valves are found considering different simulation durations. By determining the openings of the valves; almost 60 % excess pressure decrement is obtained.