Communication Assisted Protection Scheme Based on Artificial Neural Networks for Multi-Microgrid

Abstract

Designing a properly coordinated protection scheme for a multi-microgrid is very challenging because of their distinct characteristics. A microgrid can be operated in a grid-connected mode through direct connection to utility or through another grid-connected microgrid. It can also operate in an islanded mode or get connected with another independent microgrid. The various operation modes and topologies under which the system may operate bring major challenges such as the bi-directional flow of power and substantial variation in fault current. Such characteristics make protection schemes of conventional radial distribution systems unreliable options. In this paper, a centralized communication-assisted protection scheme based on artificial neural networks is proposed. The scheme operates in a cascaded process. In the first stage, a central protection controller is responsible of identifying and isolating the microgrid or tie line which has the fault. In the second stage, the local protection controller will be activated due to detection of islanding condition. It will then identify and isolate the faulty line accordingly. Identification of fault location is accomplished through neural networks trained with massive amount of three phase voltage and current measurements of all buses and lines during different fault scenarios using MATLAB/Simulink environment. The proposed scheme utilizes IEC 61850 based standardized communication to monitor the multi-microgrid and send the trip commands. Finally, the performance evaluation of the proposed scheme in terms of end-to-end delays including neural networks computational delay and communication network delay through extensive simulations is also presented which proves the effectiveness of the proposed protection scheme.INDEX TERMS Communication assisted protection, distributed energy resources, microgrid, neural networks.

Fault identification, classification, and localization in microgrids using superimposed components and Wigner distribution function

“…The obtained signals are then used to calculate the superimposed components. The superimposed components of the three-phase current and voltage signals are obtained by subtracting the pre-fault/normal signal from the post-fault current/voltage signal (Awagan et al, 2021;Qusayer and Hussain, 2024). Mathematically, superimposed components can be expressed by Eq.…”

“…The authors in (Chauhan et al, 2022) developed a protection method by using the rate-of-rise of fault current and an adaptive current threshold to provide quick and precise fault detection in MGs. The authors in (Qusayer and Hussain, 2024) presented a communication-assisted protection scheme for multi-microgrid using neural networks. The scheme identified and isolated faults in different operation modes and topologies.…”