A5025873277- Building Energy and Comfort Optimization
- Smart Grid Energy Management
- Energy Efficiency and Management
- Energy Load and Power Forecasting
- Integrated Energy Systems Optimization
- Advanced Control Systems Optimization
- Wind and Air Flow Studies
- Microgrid Control and Optimization
- Air Quality Monitoring and Forecasting
- Time Series Analysis and Forecasting
- BIM and Construction Integration
- Advanced Neural Network Applications
- Control Systems and Identification
- Thermodynamic and Exergetic Analyses of Power and Cooling Systems
- Traffic control and management
University College Dublin
2019-2022
Abstract This paper describes an open data set of 3,053 energy meters from 1,636 non-residential buildings with a range two full years (2016 and 2017) at hourly frequency (17,544 measurements per meter resulting in approximately 53.6 million measurements). These were collected 19 sites across North America Europe, one or more building measuring whole electrical, heating cooling water, steam, solar as well water irrigation meters. Part these was used the Great Energy Predictor III (GEPIII)...
In late 2019, ASHRAE hosted the Great Energy Predictor III (GEPIII) machine learning competition on Kaggle platform. This launch marked third energy prediction from and first since mid-1990s. this updated version, competitors were provided with over 20 million points of training data 2,380 meters collected for 1,448 buildings 16 sources. competition's overall objective was to find most accurate modeling solutions 41 private public test points. The had 4,370 participants, split across 3,614...
The increasing penetration of renewable energy sources has the potential to contribute towards decarbonisation building sector. However, this transition brings its own challenges including that integration and grid instability issues arising due stochastic nature variable sources. One approach address these is demand side management, which increasingly seen as a promising solution improve stability. This achieved by exploiting flexibility shifting peak periods generation. single needs be...
Buildings are increasingly being seen as a potential source of energy flexibility to the smart grid form demand side management. Indicators required quantify available from buildings, enabling basis for contractual framework between relevant stakeholders such end users, aggregators and operators. In literature, there is lack consensus standardisation in terms approaches indicators quantifying flexibility. present paper, current reviewed most recent market independent compared through...
This research is concerned with the novel application and investigation of 'Soft Actor Critic' based deep reinforcement learning to control cooling setpoint (and hence loads) a large commercial building harness energy flexibility. The motivated by challenge associated development conventional model-based approaches at scale wider stock. Soft Critic model-free technique that able handle continuous action spaces which has seen limited real-life or high-fidelity simulation implementations in...
Reinforcement learning is a promising model-free and adaptive controller for demand side management, as part of the future smart grid, at district level. This paper presents results algorithm that was submitted CityLearn Challenge, which hosted in early 2020 with aim designing tuning reinforcement agent to flatten smooth aggregated curve electrical diverse buildings. The proposed solution secured second place challenge using centralised 'Soft Actor Critic' deep able handle continuous action...
The increasing share of renewable energy sources in the power industry poses challenges for grid management due to stochastic nature their production. Besides traditional supplyside regulation, flexibility can also be provided by demand side. Demand-Response is an attractive approach based on adapting user profiles match supply constraints. Nevertheless, defining potential related buildings not straightforward and continues pose challenges. Commonly accepted standardized indicators...
Data-driven approaches are playing an increased role in building automation. This can, part, be attributed to operation and energy management system data becoming more readily accessible. A particular application is models allow predictive control harnessing flexibility, which of interest different stakeholders including; utilities, aggregators end-users. Given the possibility thousands features, feature selection becomes a critical part model development process. paper considers various...
Commercial buildings are significant end-use energy consumers and given their inherent thermal mass use of advanced control infrastructure together with heating, ventilation air conditioning systems, have the potential to offer load shifting opportunities. Data-driven frameworks show promising results as a robust scalable technique for heterogeneous building stock that will enable automated demand response whilst ensuring occupant comfort is maintained. This research compares two different...