الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 131 |  |  |
| | | from | 131 | | |
Abstract
With the improved optimized operation and increase in loading of a power system, the phenomenon of voltage instability and hence the voltage collapse attracts more and more attention due to this reason. The phenomenon of possible steady state voltage instability is creasing serious concern among operation of large interconnected power systems.
Methods for predicting voltage instability in power system can categorized either into steady state or dynamic methods. The methods of dynamic ones are very consuming in computation time. The static methods with their much less computing time provide sensitive information of the system. Steady state methods are widely used to provide a close observation of the voltage instability phenomenon.
There are different methods to study this phenomnon. Some of these methods are the jacobain method, the simplified jacobain method, the multiple load flow method and the voltage collapse proximity indicator VCPI method.
The methods of steady state voltage instability assessments are reviewed and discussed and a power flow MATLAB program was built using gauss method. In this thesis, it is suggested to use the voltage collapse proximity indicator method VCPI to detect the weakest buses as a traditional method. The load shedding at the weakest buses are applied to two different power systems. These studies illustrate that the indicator can provide useful and necessary information about the system state at any operating point.
In this thesis a load shedding under frequency have to be used on IEEE 30bus using PSS/E power system simulation for engineers program.
The traditional methods can not provide a fast to the frequency stability problem while, the time is very critical to take fast action to save the system from collapse.
Artifical intelligent can overcome the lone time needed to study system frequency stability and provide fast solution online.
AI has found four branches of application, which are Genetic algorithms, Fuzzy sysytems, Expert sysytems, Neural networks.
The expert sysytem gas to be used as a proposed method. The simulation program will be developed to intergrate the expert system module. Load shedding with frequency response program is used to simulate the power system behavior. Hence the expert system module exchange information with this program. The interface module takes care of the presentation of the solution proposed by the expert system with the sequence of applied rules.
Expert system programs are usually written in language like LISP and PROLOG. In this thesis the PROLOG language is used as a language for designing the suggested expert system.