Abstract:
The power system continuously experiences various disturbances, which lead to lowfrequency oscillations. It takes place among the rotors of the synchronous generator, which
is connected to the power system. Low frequency oscillation (LFO) may sustain and grow
in magnitude to cause system interruption if no adequate damping is provided. Power
system stabilizer is one of the signals generating control devices to solve this oscillation
problem. In this research, the damping of the LFO with an oscillation frequency of
1.1512Hz in the Tis Abay II hydroelectric power plant was presented by designing a
mathematically model-free adaptive network fuzzy inference system-based power system
stabilizer (ANFIS-based PSS) to improve the transfer capability and small-signal stability.
To verify the capability of suggested stabilizer and system stability enhancement
comparative analysis of existing system without controller, with conventional power
system stabilizer (CPSS) and with proposed controller is performed using the time domain
simulation of linearize model of Tis Abay II Hydropower plant in MATLAB SIMULINK
software. The time domain simulation results show that the steady state response of the
system settling time improved. The CPSS improved the settling time from 9.9377 to
4.5989sec, 9.9777 to 4.48448sec,9.9777 to 5.4164 sec, and 9.9744 to 4.7151sec for rotor
speed, power angle, electrical, and accelerating torque parameters respectively. In this
approach, the hybrid-learning can tune the fuzzy rules and membership function of the
fuzzy logic in the ANFIS system. The dynamic performance of the proposed system was
investigated for a given operating conditions for system parameters using eigenvalue and
time response analysis. The ANFIS-based PSS improved the settling time from 9.9377 to
3.9807sec, 9.9777 to 4.78220sec,9.97776 to 5.0088sec and 9.9744 to 2.0480sec for rotor
speed, power angle and electrical accelerating torque respectively. The neuro fuzzy system
combines the characteristics of fuzzy logic systems with their ability to deal with imprecise
knowledge, and neural networks, with their advantages of establishing a relationship
between the system's inputs and outputs, are represented as qualified tools for systems of
unknown plan.
Keywords: Adaptive Network Fuzzy Inference System (ANFIS), Low Frequency
Oscillation (LFO), and Conventional Power System Stabilizer (CPSS).