Graduate Thesis Or Dissertation
Bayesian Based Parameter Identification for Building Energy Models Public Deposited
https://scholar.colorado.edu/concern/graduate_thesis_or_dissertations/vt150j50t
- Abstract
- In this research work, a series of sensitivity analyses were performed to validate the proposed Bayesian approach to identify unknown parameters in building energy models. The proposed Bayesian approach mainly consisted of creating a Gaussian process emulator to sample the posterior distribution. Sensitivity case studies were carried out to investigate followings: appropriate sampling numbers, size of Gaussian process, observation noise, continuous/discrete variables situation. Validation on the proposed approach was done with closed loop results (one RC model and two DOE2.2 models) as well as three actual buildings (two commercial buildings and one residential building). The result showed success of identifying unknown parameters by higher occurrences on target values. Moreover, the proposed approach was tested in actual buildings and shown to calibrate the building energy models with unknown parameters still inside. As an application of the proposed Bayesian approach, development of identification of Energy Conservative Measures (ECMs) were carried out. The proposed approach succeeded in identifying the appropriate ECMs with uncertainties in budgets, initial costs, and actual performance of the ECMs. Furthermore, comparison studies between other linear models and traditional Bayesian approach have been carried out to demonstrate the characteristic of the proposed approach to other methods. Also, this study has validated the possibility of utilizing the simplified approach in a future study.
- Creator
- Date Issued
- 2014
- Academic Affiliation
- Advisor
- Committee Member
- Degree Grantor
- Commencement Year
- Subject
- Last Modified
- 2019-11-14
- Resource Type
- Rights Statement
- Language
Relationships
Items
Thumbnail | Title | Date Uploaded | Visibility | Actions |
---|---|---|---|---|
bayesianBasedParameterIdentificationForBuildingEnergyModel.pdf | 2019-11-14 | Public | Download |