New Approaches For Out of Step Detection

Abstract

A reliable, continuous supply of electrical energy is very essential for the functioning of today's modern complex and advanced society. Electricity is one of the prime factors of the growth. It determines the value of the society. The energy balance between generation and consumption is very important for power system operation. Disturbance such as faults, line outages, and load changes …etc cause changes for system variables and deviations from the normal values. An abnormal condition may lead to unstable power swing situation for a synchronous generator that is called out of step. Therefore, a suitable system of protection is necessary for a secure power system operation. Several researches discussed the different detection methods for the out of step condition. Some of the researches depended on classical methods of parameters calculations. Others depended on artificial intelligence tools. This thesis introduces an analytical technique for such detection that is based on the frequency deviation of the generator voltage waveform. Also, it is based on the plane trajectory of the speed and acceleration of the rotor. A single machine to infinite bus (SMIB) model is simulated to verify the validity of the technique. Moreover, the thesis presents a new approach for out of step detection based on usage of wavelet transform and the calculation of wavelet energy. The (SMIB) model is used again in the simulations that are carried out using MATLAB/Simulink package. The simulation includes different cases at different initial rotor angles and different fault clearing times. In addition, a simulation of multi machines system is applied to extend the applicability of the proposed technique.

Keywords: out of step, frequency deviation, wavelet transform and Multi machines system.