الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 16 |  |  |
| | | from | 16 | | |
Abstract
Feeder reconfiguration allows the transfer of load from heavily-loaded portions of the system to locations that are relatively lightly loaded in order to obtain the optimal configuration. That configuration achieves the maximum saving by reducing the active power loss, improves the operating conditions of the system and enables the full utilization of system equipment capabilities.
Loads in the distribution system of Alexandria area, for example vary between three levels per day basis. Therefore, the configuration of the network needs to be changed accordingly, aiming to obtain the optimal configuration that achieves the maximum reduction in the power loss.
A genetic based algorithm (GA) is proposed in this thesis for solving the complicated combinatorial optimization problem for distribution system load balancing reconfiguration in order to determine the optimal topology, that economically reduces the active power loss and improves the system performance, according to the variation of load pattern. Keeping in mind that the practical and recommended system operating conditions and constraints: ( maximum voltage drop, maximum feeder load and maximum number of switching ) are satisfied. In fact, GA has gained a high attention in the power area. They are widely applied in different topics like: energy loss minimization, economic load dispatch, unit-commitment, load forecasting, fault diagnosis, distribution system restoration, power system planning, power system stability, security assessment, reactive power, voltage control, scheduling problem, ...etc.
The developed research work provides the application of the proposed GA for multi-objective programming to solve the reconfiguration in a distribution system. Six different objectives are considered here in conjunction with network constraints.
Concluding, the GA-based method is capable of solving non smooth, non-continuous, and non-differentiable problem and so it can be used effectively to solve such type of % optimization problems.