Microgrid confrontations and smart resolution

Abstract

Hybrid microgrids are emerging as an alternate solution for connecting distributed AC/DC energy resources. Effective fault detection and response are highly essential for the microgrid controller (MGC) for protection of the microgrid. The conventional schemes of protection cannot be applied in microgrid because of dynamic conduct and unconventional topology of the microgrids. It is highly essential to develop an appropriate scheme for detection and classification of faults for the effective protection of microgrids. In this paper, a novel and smart solution based on the application of an intelligent machine learning (ML) fine tree algorithm is applied to the hybrid microgrid controller. This algorithm resulted in effective detection & classification of faults which in turn was used for separation of faulty segment. The intelligent model obtained with the proposed algorithm performed well and fault detection accuracy has been showcased for various fault scenarios. The overall fault detection accuracy obtained is 98%. Severity of faults and associated confrontations are also discussed in this work. Performance efficacy of the proposed ML based protection algorithm for MGC is substantiated in MATLAB environment.