Repository of DUTH

Distributed optimization for intelligent demand response in smart grids (Doctoral thesis)

Κόρκας, Χρήστος/ Korkas, Christos

Over the recent past years, big research effort has been dedicated towards addressing a generic solution in Smart-Grid operation control problems. The integration of smart-metering, renewable energy resources, electric vehicle charging and energy storage have recently emerged in the electrical system scene and have caused the Demand Response problem to gain even more interest. Throughout the literature, various optimization techniques have been used to minimize the operation cost on Smart-Grid nodes, like buildings, microgrids and charging stations by reducing energy consumption, emissions and improving the quality of life and service of consumers. However, such an operation, is by no means an easy task. In this work, we present a series of novel algorithms which are aiming to optimize the operation of Smart-Grid nodes. More specifically, through different test cases of smart buildings, microgrids and electric vehicle charging stations, we validate the performance of these algorithms in minimizing the energy consumption, the financial cost, the emissions and the grid dependency of these test cases without neglecting the user comfort and needs. Furthermore, these operations take place under dynamically changing conditions like weather conditions, renewable energy resources, fluctuating energy prices and heterogeneous occupancy schedules. This work can be divided into the following main subject sections: _ Introduction: in this section the main definitions of Smart-Grid and Demand Response are presented. Moreover, a literature review of optimization techniques in Demand Response programs is presented, while discussing advantages and drawbacks of each method. _ Centralized Intelligent Management for Grid-Connected Microgrids: in this section, a microgrid test case is presented and used to validate a novel control algorithm for centralized joint demand response management and thermal comfort optimization. _ Decentralized Intelligent Management in Microgrids with Renewable Energy Sources and Energy Storage: in this section, a decentralized version of the previous presented algorithm is established, through a bigger microgrid test case, with the integration of various renewable energy sources and energy storage. _ Distributed and human-in-the-loop Optimization in Large Scale Networks: in this section, we move into a Large-Scale-System consisting of 100 buildings, in an attempt to validate the previous established algorithms, and their performance in distributed optimization. _ Nearly-Optimal Dynamic Charging of Electric Vehicle Fleets: in this section, a novel adaptive algorithm is presented, in an attempt to optimize the operation of a charging station by minimizing the operation cost under dynamically changing conditions. Each chapter, introduces each new test case, the proposed solution, the related state-of-theart and the results that validate the performance of the presented strategies.