Now showing 1 - 4 of 4
  • Publication
    Evaluation of Machine Learning Algorithms for Demand Response Potential Forecasting
    This paper focuses on the ability of machine learning algorithms to capture the demand response (DR) potential when forecasting the electrical demand of a commercial building. An actual sports-entertainment centre is utilised as a testbed, simulated with EnergyPlus, and the strategy followed during the DR event is the modification of the chiller water temperature of the cooling system. An artificial neural network (ANN) and a support vector machine (SVM) predictive model, are utilised to predict the DR potential of the building, due to the significant amount of execution time of the EnergyPlus model. The data-driven models are trained and tested based on synthetic databases. Results demonstrate that both ANN and SVM models can accurately predict the building electrical power demand for the scenarios without or with daily DR events, whereas both predictive models are not accurate in forecasting the electrical demand during the rebound effect.
      74
  • Publication
    A Study on the Trade-off between Energy Forecasting Accuracy and Computational Complexity in Lumped Parameter Building Energy Models
    The development of urban scale cost-optimal retrofit decision making requires the development of simplified building energy models which provide satisfactory energy prediction accuracy while remaining tractable when implemented at scale. Lumped parameter building energy models are computationally efficient representations of building thermal performance. The current paper introduces a user-led iterative model reduction methodology which identifies potential trade-offs between model complexity (thus computational requirements) and energy estimation accuracy. Model complexity is progressively reduced using an energy performance criterion prior to model trimming. The methodology is applied to a building energy model of a mixed-use building, which is developed in the EnergyPlus Building Energy Model Simulation (BEMS) environment. The energy performance of the building is evaluated using a linear energy minimisation problem. The proposed methodology shows a potential reduction by half of the model complexity is possible, while retaining annual energy estimation errors below 10% for the target building.
      314
  • Publication
    Input Variable Selection for Thermal Load Predictive Models of Commercial Buildings
    Forecasting of commercial building thermal loads can be achieved using data from Building Energy Management (BEM) systems. Experience in building load prediction using historical data has shown that data analysis is a key factor in order to produce accurate results. This paper examines the selection of appropriate input variables, for data-driven predictive models, from wider datasets obtained from BEM systems sensors, as well as from weather data. To address the lack of available complete datasets from actual commercial buildings BEM systems, detailed representation of reference buildings using EnergyPlus were implemented. Different types of commercial buildings in various climates are examined to investigate the existence of patterns in the selection of input variables. Data analysis of the simulated results is used to detect the correlation between thermal loads and possible input variables. The selection process is validated by comparing the performance of predictive models when the full or the pre-selected set of variables is introduced as inputs.
      217Scopus© Citations 65
  • Publication
    Prediction of Residential Building Demand Response Potential Using Data-Driven Techniques
    This paper is concerned with the evaluation of the ability of data-driven predictive models to capture the demand response potential in residential buildings. A mid-floor apartment with an air to water heat pump for space heating, utilised as an archetype dwelling, is simulated using EnergyPlus. The research is focused on forecasting the electrical demand from the heating load for the coldest month of the year, considering two types of DR events, load reduction and load increase. After the generation of the synthetic database, an artificial neural network model and a support vector machine model are examined regarding their ability to predict the electrical demand from heating loads.
      54