الفهرس 
الغلافOnly 14 pages are availabe for public view
 |  | from | 141 |  |  |
| | | from | 141 | | |
Abstract
Nowadays, wind energy is the most popular method of generating electricity from renewable sources and installed wind farms are continuously and rapidly increasing in many power systems in developed countries. But the penetration numbers of the wind farms in the power system will have an adverse impact on transmission lines protection while the characteristics of the wind farms are different from the conventional sources when a fault
occurs. Different factors such as the distributed generation concept, the own behavior of the induction generator, variation of wind farm impedance during faults inside the farm, and varying wind speed arise different challenges regarding the behavior of the protection of the
transmission lines connected to the wind farm, so a detailed dynamic simulation of large wind farms must be taken into consideration for rotection applications.
In this thesis, a new protection scheme for transmission lines connected to the wind farm and power grid was presented using Support Vector Machine (SVM) classification tool.
The classifier can differentiate between the normal and abnormal conditions. The classifier is based on the positive and negative component of the current. The developed model represents the real fifth stage of Zafarana wind farm in Egypt. This system consists of 100x850 kW GAMESA wind turbine type connected to the power grid through two parallel transmission lines. The wind turbine is modeled using MATLAB/SIMULINK. The following points are considered; each turbine is simulated in the detailed mode to conduct WF dynamic behavior,
large-scale wind power plants are simulated to conduct high penetration, and all turbines, collecting feeders, transmission lines, and transformers are simulated with typical data of fifth stage Zafarana wind farm (Case study).
The performance of the developed algorithm is tested on a simulated data set of 1890 cases and overcomes many problems such as penetration numbers of the wind farms, wind
speed variation, variation of wind farm impedance during faults inside the farm, fault location, fault type, moreover; the proposed scheme remains stable with high fault resistance.