Dp 586 en
Distributed building identification
Author: Bäumelt Tomáš
This master thesis deals with building identification. At first, two different building thermal continuous-time LTI models are created using a technique of RC thermal circuits. These models are then used for simulations and data generation foridentification process. The main objective of this thesis is to propose and implement an algorithm for distributed building identification. The identification problem of the overall building is decomposed using so called decompositions methods, which results in solving small local identification problems with additional constraints on parameters of mutual walls. The local identification processes measured data (temperatures and heat flows of heat exchangers) and utilizes grey-box modelling and the Least Squares or Maximum Likelihood Estimation. Finally, the proposed algorithm is tested on a 4-zone building benchmark model for various configurations.
- Tomáš Bäumelt, mailto:baumetom@fel.cvut.cz
