Li, Caiyu;
Geng, Yang;
Tang, Hao;
Mumovic, Dejan;
Korolija, Ivan;
Lv, Zihui;
Cao, Tianxiang;
... Lin, Borong; + view all
(2023)
Load prediction with an improved feature selection method for building energy management of an office park.
In:
11th International Conference on Indoor Air Quality, Ventilation and Energy Conservation in Buildings: IAQVEC2023.
: Tokyo, Japan.
Preview |
Text
Load prediction with an improved feature selection method for building energy management of an office park.pdf - Published Version Download (798kB) | Preview |
Abstract
Load prediction plays a significant role in building energy management. An accurate HVAC load prediction model highly depends on the feature selection and the quality of training data. In previous work on load prediction, the input features are majorly manually selected by expertise, which is relatively subjective and lacks theoretical supports. Using the real building operational data collected from an office park located in Hangzhou, this paper developed a short-term cooling load prediction model, in which the input features are selected based on an analysis on the heat transfer process. Combined with qualitative analysis of the real data, several features such as outdoor air enthalpy and indoor black-bulb temperatures from different orientations are introduced into the model. The proposed model was then applied to the HVAC control system of the office park. Compared to the load prediction model with commonly used features, the proposed model reduced CRVMSE by 21% and MAPE by 30% during the operation period of the system. Furthermore, the impacts of training dataset size and prediction time range on model’s accuracy and training time were discussed.
| Type: | Proceedings paper |
|---|---|
| Title: | Load prediction with an improved feature selection method for building energy management of an office park |
| Event: | IAQVEC2023 - 11th International Conference on Indoor Air Quality, Ventilation and Energy Conservation in Buildings |
| Location: | Tokyo, Japan |
| Dates: | 20 May 2023 - 23 May 2023 |
| Open access status: | An open access version is available from UCL Discovery |
| Publisher version: | https://www.iaqvec2023.org/index.html |
| Language: | English |
| Additional information: | This version is the author accepted manuscript. For information on re-use, please refer to the publisher’s terms and conditions. |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of the Built Environment UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of the Built Environment > Bartlett School Env, Energy and Resources |
| URI: | https://discovery-pp.ucl.ac.uk/id/eprint/10171652 |
Archive Staff Only
 |
View Item |
