Loading...
| Friend's email | |
| Your name | |
| Your email | |
| enter code | |
This page was sent successfuly
- Type of Document: Ph.D. Dissertation
- Language: English
- Document No: 52119 (52)
- University: Sharif University of Technology, International Campus, Kish Island
- Department: Science and Engineering
- Advisor(s): Hemmatyar, Ali Mohammad Afshin
- Abstract:
- Moving towards a sustainable energy system in the evolving society requires the utilization of new applications like Electrical Vehicles (EVs) and integration of Renewable Energy Sources (RESs) which will make new challenges for today’s power grids. To address these challenges, it is essential for emerging demand control paradigm to not only adapt the required demand with available supply but also consider the comfort and the level of lifestyle each customer may have in practice. In this research, due to the ever-growing interconnectivity of the grids, a distributed Commodity Market (CM) framework is proposed in which intelligent agents embedded in Demand Control Entity (DCE) units in customers want to maximize their preferred welfare through real-time demand of power from an energy market. Since there is not a comprehensive model for the grids, utilizing Reinforcement Learning (RL) technique proves that the global optimal performance is achieved at the Nash Equilibrium (NE) of the proposed framework. This solution not only meets the resource utilization of the market, but also allocates strategically optimal demands to the customers who have budget constraints. The resulted strategic capability means that the DCE agents seek to optimize long-term welfare instead of instantaneous welfare. Considerable buying power and diverse assumption-free interests of the customers are also other novelties offered. In next step, to consider social objectives for the communities that may give priority to social values, a socially intelligent ability is added to the proposed framework as an option. In this case, social concepts such as social welfare and social fairness will be met among the customers. In this new framework, participating customers are not merely competitive and they will consider welfare objectives of other customers which will consequently cause the customers to be satisfied according to their contributions in achieving the overall system’s welfare. Finally, a framework is developed for the customers who have the aim to consider economic goals like cost minimization in addition to welfare maximization. Considering simultaneous quantitative and qualitative goals in a joint optimization form for the budget-constrained customers without any restrictions on the customers’ preferences or the supply side is another novelty offered. The simulation results confirm that not only can the developed frameworks significantly improve the welfare in a stable manner, but they are more successful in obtaining the demand during peak hours than just the economical frameworks proposed in literature. In addition, they confirm that the proposed framework has s scalability power due to the independent operation of the customers and the market
- Keywords:
- Intelligent Control ; Reinforcement Learning ; Intelligent Agent ; Smart Power Grid ; Social Welfare ; Sustainable Development ; Budget Constraint
Digital Object List-
محتواي کتاب - view
Bookmark