Building energy optimisation under uncertainty using ACOMV algorithm
- Publication Type:
- Journal Article
- Citation:
- Energy and Buildings, 2018, 167 pp. 322 - 333
- Issue Date:
- 2018-05-15
Closed Access
| Filename | Description | Size | |||
|---|---|---|---|---|---|
| 1-s2.0-S0378778817315943-main.pdf | Published Version | 1.36 MB |
Copyright Clearance Process
- Recently Added
- In Progress
- Closed Access
This item is closed access and not available.
© 2018 Elsevier B.V. In building optimisation many parameters are uncertain due to their dependence on the building operation and environment (e.g. internal loads). This uncertainty implies that the “optimised” building is likely sub-optimal for the actual parameters. This study develops a new scenario-based optimisation methodology to address building parameter uncertainty. A multi-objective optimisation problem based on three objective functions (“low” “base” and “high” simulation scenarios) is developed and scalarised using the weighted sum method to find the optimised compromise between energy use for different scenarios. Necessitated by the increased computational demand of multi-objective problems, a modified version of the Ant Colony Optimisation algorithm for Mixed Variables (ACOMV-M) is developed. A comparison between ACOMV-M and a benchmark algorithm showed that ACOMV-M converged to solutions of similar quality with approximately 50% fewer simulations. The results on an Australian office building showed that the energy-optimised building parameters can vary significantly for different assumptions. Furthermore, inaccurate assumptions on internal loads and infiltration rate can reduce energy savings achieved by optimisation up to 4.8 percentage points. The proposed methodology is used to identify parameters that are sensitive to different scenarios and demonstrated that more robust solutions can be achieved through modest sacrifices in optimality to any one scenario.
Please use this identifier to cite or link to this item: