| In this dissertation, we address the problem of optimal resource allocation through distributed market-based techniques in future power systems. Several connected problems are considered in different stages of this research project: 1) optimal decision making of generation companies in wholesale markets, 2) demand response and energy pricing in retail markets, 3) single and multiple microgrid power management within the retail sector, and 4) introducing a resiliency-aware power management for microgrid-based distribution systems. Hence, this work addresses the challenges that are connected to the economics of energy.;All of these problems are concerned with optimizing the behavior of the participants in electricity markets (in wholesale and retail sectors) to achieve certain objectives in power system operation (e.g., profit maximization, peak shaving, resiliency improvement, extreme event awareness and preparedness, etc.). Solution concepts from the fields of machine learning and game theory have been employed to address these problems. Moreover, distributed agent-based frameworks are designed for modeling the interactive decision making processes and implementing control laws in electricity markets. The goal is to introduce automated "intelligent" decision making capabilities into power systems to improve the efficiency of grid operation. This is closely related to the concept of "smart grids".;The proposed solution strategies are verified through numerical simulations in MATLAB environment. The results demonstrate that using the proposed distributed decision tools and machine-learning-based techniques we can improve the performance of power systems to achieve higher levels of controllability and efficiency, while enhancing system resiliency. |
