3 présentations

• 
  14h45 - 15h07

  Enhancing Energy Management for Electric Vehicles through Private Charging Pile Sharing

  • Mohammad Reza Khodoomi, prés., School of Engineering, University of British Columbia, Kelowna, BC, Canada
  • Babak Mohamadpour Tosarkani, School of Engineering, University of British Columbia, Kelowna, BC, Canada
  • Eric Ping Hung Li, Faculty of Management, University of British Columbia, Kelowna, BC, Canada

  The rapid adoption of electric vehicles (EVs) has introduced complexities in designing efficient battery charging strategies. One promising approach is to share private charging piles (PCPs) in smart homes powered by renewable energy sources like solar panels and wind turbines. By enabling the sharing of PCPs, this approach not only reduces costs for both EV owners and households but also contributes to grid stability through vehicle-to-grid services. This research proposes a two-stage stochastic optimization framework for managing the energy systems of smart homes equipped with distributed generation units and PCPs. To address uncertainties in demand and renewable energy generation, the model employs a sample average approximation technique. Scenarios are generated using an autoregressive moving average model, while a fuzzy c-means clustering algorithm is applied to mitigate computational challenges. Finally, the model is validated using data from Calgary, Canada, demonstrating its effectiveness in a practical urban setting.

• 
  15h07 - 15h29

  Stochastic Optimization for Demand Response Scheduling of Green Hydrogen Production

  • Gabriel Patron, prés., Imperial College London
  • Nilay Shah, Imperial College London
  • Calvin Tsay, Imperial College London

  Hydrogen is increasingly being incorporated into global net zero plans with a particular emphasis on electrolysis-based “green” hydrogen. When integrated with the power grid, the demand response scheduling of a hydrogen production plant can be posed as a cost minimization problem that relies on having predicted electricity price signals as inputs. In reality, these price signals are often subject to considerable uncertainty and, especially when participating in several power markets (e.g., day-ahead, intraday), the performance of a deterministic scheduling solution can be significantly suboptimal. In this presentation, we propose a stochastic risk-aware formulation to determine the optimal hydrogen production schedule in a multi-market context with uncertain price signals. A multi-stage formulation allows for recourse actions upon price perturbations and prediction errors, while risk-aware objective functions such as conditional value-at-risk (CVaR) can minimize shortfall. Further, the proposed approach determines the proportion of participation in each power market.

• 
  15h29 - 15h51

  Deep Reinforcement Learning Based Dynamic Management for Peer-to-Peer Energy-Sharing Networks

  • Vahid Gholamzadeh, prés., University of Victoria
  • Adel Guitouni, University of Victoria
  • Haworth Brandon, University of Victoria
  • Curran Crawford, University of Victoria

  This paper explores dynamic management of peer-to-peer energy-sharing networks in the face of uncertainty to achieve optimal energy production, trading and storage decisions. Prosumers, who both produce and consume energy, face challenges in balancing supply and demand due to renewable energy variability, changing consumption patterns, and market conditions. We developed a centralized management framework that integrates agent-based modeling and deep reinforcement learning to tackle these issues. The framework dynamically adjusts decision-making policies to optimize energy allocation and trading, while prioritizing social welfare across interconnected prosumer networks. The model is validated by IEEE 14-bus system case study, which incorporates scenarios with renewable energy variability and demand fluctuations. The framework effectively achieves high demand satisfaction, boosts renewable energy use, and reduces reliance on non-renewable resources. It emphasizes the importance of grid integration and adaptive pricing for enhancing network resilience, contributing to centralized management solutions for scalable and flexible P2P energy-sharing networks.