الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 16 |  |  |
| | | from | 16 | | |
Abstract
The concept of a distributed utility is instaIlation of smaIler modular resources in a distributed manner closer to the point of end use. Such 10caIly distributed generation has several merits ITom the viewpoint of environmental restriction and location limitations, as well as transient
and voltage stability in t~e power systems.
The objective of this thesis is to obtain the solution for minimized network power loss by optimalIy locating the dispersed wind generation under the constraint of the total injection of instalIed dispersed generation. This problem has severe non-linearity due to the physical constraints such as balance between power supply and demand, limitation for the total dispersed generation injection capacity and other atTecting factors. The larger system for the analysis the more memory required for traditional methods. Genetic Algorithm (SGA: Simple Genetic Algorithm) can iichieve global optimum solution (through the directional irregular searching). But it has several drawbacks such as excessive convergence time and premature
convergence.
This thesis, tried to overcome the defects of Genetic Algorithm, by Modified Genetic Algorithm (MGA) to solve the optimal distribution of dispersed generation problem. By using the concept of sexual ditTerentiation and a particular type of selective breeding in choosing parents for each generation of the genetic population. Modified Genetic Algorithm tried to reduce diversity loss rate and increase the robustness of Genetic Algorithm. We tried to make convergence time and stability better by improving the operators of Genetic
Algorithm.
A new approach to dispersed wind generation planning based on modified Genetic Algorithm ’in a sub-transmission system is presented. A method to optimally locate such generation in a meshed network, and maximizing the potential benefits is outlined using Genetic Algorithm and its improvement. The benefit, expressed as a performance index, is minimization of losses. The proposed method is tested by simple power system with 14-bus type