الفهرس 
INTRODUCTIONOnly 14 pages are availabe for public view
 |  | from | 115 |  |  |
| | | from | 115 | | |
Abstract
Traditional power grids use one-way communication for power distribution from central generators to a large number of users. It is not an interconnected network, so a lot of procedures and processes are required to determine and fix any arising problem. In contrast to the conventional power grid, smart grid utilizes two-way communications to make an automatic and innovative power delivery network. Smart grid defines a self-healing network built with dynamic optimization methods that use real-time measurements to reduce network losses. The Smart grid has a lot of research topics from the power aspects and communications aspects. This research started with gathering the information for both power and communications points of view and a survey for the smart grid has been conducted to capture the whole view. This thesis discusses one of the problems in the electrical distribution systems which is the load balancing between phases. The unbalance of phases may result in neutral conductor damage. Furthermore, the unbalancing in the LV network causes high power losses and increases in energy demand. The generation station will not be able to support the increase in demand and high-power losses, which may lead to a blackout. In order to overcome this problem, a control system is formulated in this research to properly design and optimize the load balance distribution system. This thesis presents a control system that uses an optimization technique to minimize the neutral current. The control system depends on rearranging the loads in such a way that the system becomes more balanced. Applying an optimization technique to reorganize the loads connected to each phase will result in minimizing the neutral current which will lead to minimizing power losses. Different methods of optimization are used to find the best performance of the control system. Genetic algorithm and Ant colony optimization techniques are presented and simulation study for both of them is conducted. In addition, a comparison between GA and ACO is discussed. The unique advantage of each technique inspired us to present a hybrid technique that merges the advantages of each GA and ACO while avoiding the limitation of each technique. The hybrid technique started with ACO approach as an initial solution then continues the optimization process with the GA technique to ensure the accuracy and the speed of conversion. A simulation study is conducted among those techniques according to the performance of each one.