Yuce, Baris ORCID: https://orcid.org/0000-0002-9937-1535, Rezgui, Yacine ORCID: https://orcid.org/0000-0002-5711-8400 and Mourshed, Monjur ORCID: https://orcid.org/0000-0001-8347-1366 2016. ANN-GA smart appliance scheduling for optimized energy management in the domestic sector. Energy and Buildings 111 , pp. 311-325. 10.1016/j.enbuild.2015.11.017 |
Preview |
PDF
- Accepted Post-Print Version
Download (2MB) | Preview |
Abstract
Smart scheduling of energy consuming devices in the domestic sector should factor in clean energy generation potential, electricity tariffs, and occupants’ behaviour (i.e. interactions with their appliances). The paper presents an ANN–GA (Artificial Neural Network / Genetic Algorithm) smart appliance scheduling approach for optimized energy management in the domestic sector. The proposed approach reduces energy demand in “peak” periods, maximizes use of renewable sources (PV and wind turbine), while reducing reliance on grid energy. Comprehensive parameter optimization has been carried out for both ANN and GA to find the best combinations, resulting in optimum weekly schedules. The proposed artificial intelligence techniques involve a holistic understanding of (near) real-time energy demand and supply within a domestic context to deliver optimized energy usage with minimum computational needs. The solution is stress-tested and demonstrated in a four bedroom house with grid energy usage reduction by 10%, 25%, and 40%, respectively.
Item Type: | Article |
---|---|
Date Type: | Publication |
Status: | Published |
Schools: | Engineering |
Subjects: | T Technology > TA Engineering (General). Civil engineering (General) |
Uncontrolled Keywords: | ANN; Optimisation; Genetic Algorithm; Scheduling; Energy Management; Parameter Tuning; PMV; Domestic Building. |
Publisher: | Elsevier |
ISSN: | 0378-7788 |
Funders: | BRE |
Date of First Compliant Deposit: | 30 March 2016 |
Date of Acceptance: | 6 November 2015 |
Last Modified: | 06 Nov 2023 23:44 |
URI: | https://orca.cardiff.ac.uk/id/eprint/82228 |
Citation Data
Cited 109 times in Scopus. View in Scopus. Powered By Scopus® Data
Actions (repository staff only)
Edit Item |