An integrated under frequency load shedding protection based on hybrid intelligent system

Abstract

Recent blackouts, which are associated with severe technical and economic damages, show that current protection systems are not reliable enough when power system is in an emergency condition. This research attempts to address the issue by introducing a novel, integrated and optimized frequency modelling approach and Under Frequency Load Shedding (UFLS) protection for electric power systems. This system is capable to consider various aspects of the problem simultaneously in modern power systems. Furthermore, it takes advantage of a new multi-objective decision making approach considering all required criteria and risk indicators based on the related standards of power system operation. In this approach, a new frequency response modelling system, named Extended System Frequency Response (ESFR) model and new aggregated load modelling system are proposed. This approach does not only consider all factors which contribute to frequency performance of power system simultaneously, but also is capable to consider advanced components of electric power systems. This modelling system is designed in consistent with the new generation of advanced power system simulators. In the next step, Genetic Algorithm (GA) as an Artificial Intelligent (AI) method is used for designing an optimal and integrated UFLS system. The technical implementation of this step leads to the creation of a new methodology for coupling two software or simulators together. This approach is applied to create a junction between the advanced power system simulator and the GA provider. This method does not only decrease the simulation time dramatically, but also makes the remote communications possible between two or more software. Finally, an AI system, namely Artificial Neural Network (ANN), is used in a hybrid structure to execute the GA UFLS system design as an online Wide Area Protection (WAP) system. The results of the first step show the high capability of the proposed frequency response modelling system. The new approach of under frequency protection system design shows clear advantages over the conventional methods. Finally, the performance of ANN is promising as a new generation of intelligent WAP systems.