Performance assessment of a neuro-fuzzy load frequency controller in the presence of system non-linearities and renewable penetration

Abstract The existing literature demonstrates the application of neuro-fuzzy based control techniques to load frequency control in interconnected power systems. However, their performance has not been evaluated in the combined presence of renewable resources and system non-linearities. The focus of this paper is to present the design and simulation of an Adaptive Neuro Fuzzy Inference System (ANFIS) controller for a power network possessing non-linearities such as boiler dynamics, generation rate constraint, governor dead-band and time delay. A proportional integral (PI) controller was tuned with the Bode plot approach to obtain the training data set for the proposed controller. In the simulation model, multiple scenarios of wind and solar penetration levels in a two-area system were considered. The dynamic performance of the ANFIS controller was found to be superior when compared to a conventional controller in regards to peak overshoot and settling time.