Loading...

Energy Management of a Residential Prosumer Employing Machine Learning Techniques

Derakhshan Mahboob, Fatemeh | 2023

60 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 56313 (46)
  4. University: Sharif University of Technology
  5. Department: Energy Engineering
  6. Advisor(s): Moeini Eghtaei, Moein
  7. Abstract:
  8. This thesis proposes an energy management model for a residential prosumer building which is a multi-carrier energy hub in a grid-connected mode. Power may be bought and sold between the building energy system and the power grid. Solar panels and solar water heaters are integrated into the building's energy system to create electricity and heat, which may be used in a variety of ways: used immediately to meet the home's demands, stored in batteries and heat banks for later use, or sold back to the power grid to earn income. The energy dispatch is managed such that the overall energy consumption cost is minimized, considering the variation in electricity price (grid tariff), renewable energy generation (by PV and SWH), and load demand (heat and electricity). The system is modeled as an economic load dispatch optimization problem over a 24 h horizon (day ahead), and solved using mixed integer linear programming (MILP). This economic load dispatch problem model, therefore, requires knowledge of the expected electricity price, renewable energy production, and load demand (heat and electricity) over the next 24 h as the economic load dispatch model. As accurate forecasting of mentioned requirements has a significant impact on economic dispatch, which ensures the efficiency and smooth operation of the energy system, a long short-term memory (LSTM) network (hyperparameters tuned with Bayesian optimization) is proposed in this thesis. At each hour, the LSTM predicts electricity price, renewable energy generation, and load demand data for the next 24 h, the dispatch problem is then solved. Real data are then used to update the LSTM input, and the process is repeated. Learning how to build and assess long short-term memory based recurrent neural networks for multivariate multi-step time series forecasting concerning weather features and comparing its results with 5 other DRNNs(all combined with LSTM), is the main subject of this thesis
  9. Keywords:
  10. Linear Optimization ; Machine Learning ; Energy and Environmental Management ; Economic Dispatch ; Long Short Term Memory (LSTM) ; Building Energy Management

 Digital Object List

 Bookmark