الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 149 |  |  |
| | | from | 149 | | |
Abstract
The smart grid has emerged as a promising paradigm to promote and deliver a clean, modern, and efficient electricity grid to all customers. It allows local distribution companies to integrate renewable sources more reliably, efficiently, safely, and economically. When combined with flexible demand devices (such as electric vehicles or various household appliances), increased distributed generation, and the potential development of small-scale distributed storage, they could allow for energy procurement at the lowest possible cost and environmental impact. However, it presupposes real-time coordination of demand for individual households and electric vehicles at the distribution level, with generation and renewable sources at the transmission level. This implies the need to solve energy management problems on a much larger scale compared to the ones that are currently solved today. This, of course, raises significant computational and communications challenges. The need for an answer to these problems is reflected in todays power systems literature where a significant number of papers cover subjects such as generation and/or demand management at both transmission and/or distribution, electric vehicle charging, voltage control device setting, etc. The methods used are centralized or decentralized, handle continuous and/or discrete controls, approximate or exact, and incorporate a wide range of problem formulations. All these papers focus on one or two of the following power grid issues: total power losses, excessive voltage drops, load/frequency fluctuations, or peak shaving, regardless of the other issues. Moreover, there is no technique that has been designed specifically to address all these power grid issues at medium voltage and low voltage levels, taking into consideration the owners satisfaction factor, charging cost, computational time and the stochastic behaviour of the problem. This thesis proposes a two-level hierarchical controller-based coordination scheme that can be implemented in both medium and low voltage networks to handle the most critical electric vehicle charging impacts on the power grid while also allowing for energy optimization at all voltage levels.The optimal scheduling policy for charging and discharging of electric vehicles is designed to handle real circumstances, such as all the previously stated random and stochastic behaviors, different capacities and charging rates, bi-directional V2G and distribution grid restrictions. The proposed model is carried out on IEEE 31-bus 23 kV system with several low voltage residential feeders. The results proved that the proposed model can significantly reduce the aforementioned issues , namely, total losses are dramatically decreased resulting in a significant cost savings. On the other hand, the voltage remains within acceptable limits, and the demand does not exceed the capacity of the system. Finally, a modified salp swarm algorithm is developed for optimal scheduling applications. These salps belong to the jellyfish family. The results from the modified algorithm shows superior performance in terms of exploration, exploitation, and execution time which is dramatically reduced compared to other algorithms such as pattern search, genetic algorithms.