Akademik Veri Yönetim Sistemi
Tezin Türü: Yüksek Lisans
Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, Bilgisayar Mühendisliği Bölümü, Türkiye
Tezin Onay Tarihi: 2015
Öğrenci: DOĞAN POYRAZ
Danışman: CEVAT ŞENER
Özet:Energy consumption affects both energy bills of the buildings and environment greatly. Especially HVAC systems are the components that consume the most energy in commercial or residential buildings. HVAC stands for heating, ventilating and air conditioning systems in the buildings. In this thesis, data organization, retrieval and processing needs of a HVAC energy optimization system (EOS) have been analyzed, underlying technologies have been examined and architectural solutions have been proposed in order to solve various problems that a HVAC EOS may encounter. This research defines a HVAC EOS in a formal way, so that computer scientists can understand needs of energy domain, specifically HVAC EOS, easier. In addition, this research describes technologies related to BigData and presents architectural examples so that it gives insight about how to solve BigData related problems in energy domain. In order to build a HVAC EOS, there are several steps throughout the flow of the data. First one is the data stream that is between the sensors in the field and the system. In this thesis, methods to manipulate these data streams are presented so that data can be pre-processed before it is written to a database or a persistent medium. This enables data to be processed much closer to real-time. The second step is continuous processing of the persistent data for forecasting. This operation can be performed in a distributed environment so that when data size is very large, processing power of the several nodes can be used and operation can be completed much faster. The final step is the data visualization. In this step, a user of the system interacts with the HVAC EOS and manually processes some of the data or query data and display it. Again, depending on the data size, this operation can be performed in a distributed environment or data can be stored in a small in-memory medium so that response was returned to the user quickly.