Nowadays, more energy related data are created than ever before resulting large datasets, due to advances in technology, markets, and policies. In order to benefit from the large datasets, data must be managed and analyzed intelligently. Data analytics means procedures and processes for the systematic analysis (collection, analysis and presentation) of data in electronic form. The aim is to gain insight from data to support strategic, tactical and operational tasks. The large datasets have to be managed through a big data architecture which consists of storage, processing, analytics components.
The main objective of this bachelor thesis is to develop a concept of a big data architecture for building energy management by performing experiment, benchmarking, and finally selecting the technologies for each component of the architecture.
This thesis consists of the following works:
- Analysis of related fundamental theories, general requirements of the system related to big data architecture for building energy management
- Evaluation of technologies correspond to each big data architecture component and their compatibility with Java technology by performing experiments and benchmarking
- Evaluation of existing data analytics/machine learning methods and tools
- Design of the big data architecture based on evaluation result
- Validation using real data
- Documentation and presentation