Scholarlink

<p>This paper analyzes the optimization of electricity consumption in the residential sector through improved energy management and planning, combining probabilistic load estimation techniques with demand management policies. First, considering that load forecasting becomes more complex as the level of aggregation decreases, probabilistic demand forecasting methods may be particularly relevant for the participation of small customer segments in energy markets. Second, in order to mitigate the uncertainty associated with forecasts, several demand response and energy storage strategies are proposed and simulated hour-ahead to bound such uncertainty by providing demand flexibility without being limited to load shedding; the drawbacks of the proposed methodology, from the perspectives of control and energy efficiency, are also analyzed for the flexibility deployed in the hour-ahead timeframe. The proposed methodology explores synergies among Probabilistic Prediction, Load Modeling, Demand Response, Energy Storage, and Energy Efficiency. Schematically, a machine learning technique is first employed to estimate the quantiles of the conditional distribution function, generating a set of probabilistic demand predictions for both the day-ahead and hour-ahead horizons. Based on these demand predictions, Demand Response and Storage Strategies are simulated to reduce prediction uncertainty between the day-ahead and the hour-ahead scenarios. In addition, different demand management options are analyzed within the hour-ahead window from an Energy Efficiency perspective, considering the use of grey-box load models for two main residential end uses: Domestic Water Heating and Electric Space Heating. As mentioned above, the ability of Demand Flexibility to compensate for imbalances between actual load and forecasts requires both increases and decreases in demand during the control period; however, such policies may affect the overall efficiency of the appliances involved in flexibility programs. For these reasons, this work proposes the joint evaluation of both flexibility and energy efficiency, resulting in an hour-ahead scenario in which each end use is assigned a specific role to better achieve efficiency and flexibility objectives. The proposed approach is applied to consumption data from a substation with a high share of residential customers.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Physical-based models and load forecasting to improve flexibility and energy efficiency in residential segments

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月18日

查看全文
  • Antonio Gabaldon,
  • Maria Carmen Ruiz-Abellón,
  • Luis Alfredo Fernández-Jiménez,
  • Antonio Guillamón

摘要

This paper analyzes the optimization of electricity consumption in the residential sector through improved energy management and planning, combining probabilistic load estimation techniques with demand management policies. First, considering that load forecasting becomes more complex as the level of aggregation decreases, probabilistic demand forecasting methods may be particularly relevant for the participation of small customer segments in energy markets. Second, in order to mitigate the uncertainty associated with forecasts, several demand response and energy storage strategies are proposed and simulated hour-ahead to bound such uncertainty by providing demand flexibility without being limited to load shedding; the drawbacks of the proposed methodology, from the perspectives of control and energy efficiency, are also analyzed for the flexibility deployed in the hour-ahead timeframe. The proposed methodology explores synergies among Probabilistic Prediction, Load Modeling, Demand Response, Energy Storage, and Energy Efficiency. Schematically, a machine learning technique is first employed to estimate the quantiles of the conditional distribution function, generating a set of probabilistic demand predictions for both the day-ahead and hour-ahead horizons. Based on these demand predictions, Demand Response and Storage Strategies are simulated to reduce prediction uncertainty between the day-ahead and the hour-ahead scenarios. In addition, different demand management options are analyzed within the hour-ahead window from an Energy Efficiency perspective, considering the use of grey-box load models for two main residential end uses: Domestic Water Heating and Electric Space Heating. As mentioned above, the ability of Demand Flexibility to compensate for imbalances between actual load and forecasts requires both increases and decreases in demand during the control period; however, such policies may affect the overall efficiency of the appliances involved in flexibility programs. For these reasons, this work proposes the joint evaluation of both flexibility and energy efficiency, resulting in an hour-ahead scenario in which each end use is assigned a specific role to better achieve efficiency and flexibility objectives. The proposed approach is applied to consumption data from a substation with a high share of residential customers.