Intelligent Fault Diagnosis in Decentralized Renewable-Energy Systems

Elena Rossi

Intelligent Fault Diagnosis in Decentralized Renewable-Energy Systems

Author(s)	Dr. Elena Rossi
Country	United States
Abstract	The increasing deployment of photovoltaic generation, wind-energy conversion, battery storage, power-electronic converters, and networked microgrids is transforming electrical power systems from predominantly centralized architectures toward decentralized renewable-energy systems. This transition improves flexibility and local energy autonomy but complicates fault diagnosis because power flow may become bidirectional, operating topology can change, inverter-based sources may contribute limited or variable fault current, and decentralized assets can alternate between grid-connected and islanded operation. Conventional fixed-threshold protection can therefore become difficult to coordinate across heterogeneous operating conditions. This study develops an intelligent decentralized fault-diagnosis framework in which local electrical measurements, machine-learning classification, edge-level reasoning, topology information, and peer-to-peer diagnostic coordination are combined to improve fault detection and isolation. Because no primary experimental dataset was supplied, the study is explicitly designed as a simulation-based methodological investigation. A reproducible synthetic dataset representing 180 hypothetical diagnostic units was generated across three architectures: conventional threshold-based diagnosis, centralized artificial-intelligence diagnosis, and decentralized intelligent diagnosis. Six operational indicators were assessed: fault-detection accuracy, fault-classification accuracy, diagnosis latency, robustness to topology change, communication efficiency, and false-alarm rate. These variables were combined into a Fault Diagnosis Index (FDI). Mean FDI values increased from 63.3 under threshold-based diagnosis to 76.1 under centralized AI and 88.3 under decentralized intelligent diagnosis. The decentralized architecture produced the highest modeled detection accuracy (95.2%), classification accuracy (92.7%), topology-change robustness (87.7%), and communication efficiency (84.9%), while reducing mean diagnosis latency to 1.7 s and the false-alarm rate to 4.6%. These values are synthetic and must not be interpreted as measured field performance. The contribution of the study lies instead in a transparent framework that treats intelligent fault diagnosis as a distributed coordination problem rather than merely a classification problem.
Keywords	intelligent fault diagnosis; decentralized renewable energy; microgrid protection; distributed energy resources; artificial intelligence; machine learning; fault detection; fault classification; edge intelligence; smart grid
Field	Engineering
Published In	Volume 2, Issue 1, January-February 2021
Published On	2021-02-03

View / Download PDF File

About AJAMIR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+1-872-302-3851	editor@ajamir.com

American Journal of Advanced Multidisciplinary Innovation and Research

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Intelligent Fault Diagnosis in Decentralized Renewable-Energy Systems

Share this