Date of Award
1-23-2026
Document Type
Thesis
Degree Name
Engineering, MSE
First Advisor
Ehsan Naderi
Committee Members
MdMahmudul Hasan; Zahid Hossin
Abstract
The increasing complexity of current power systems, resulting from the integration of distributed generators and renewable energy sources, necessitates intelligent and adaptive fault detection schemes. Traditional protection using impedance and phasor analysis is usually weak when operating in nonlinear and transient operating conditions. Consequently, the tools of Data-driven fault classification and decision-making have gained strength under artificial intelligence (AI) and machine learning (ML) to improve grid reliability. This thesis is a proposal of an automatic fault detection and classification system based on AI applied to a smart mini-grid setting built in MATLAB/Simulink. A complete set of voltage and current data was generated by modeling a variety of fault conditions: single line-to-ground, line-to-line, two line-to-ground, three-phase, and no-fault conditions with a variety of fault resistances. Four intelligent classifiers, including adaptive neuro-fuzzy inference system (ANFIS), decision tree (DT), long short-term memory (LSTM), and support vector machine (SVM), were trained with the help of preprocessed signals recorded. A strategy of hybrid validation was used to test real-time and offline performance. ANFIS and DT models were incorporated into the Simulink environment to be tested in real-time and predict faults online as a part of the simulation. In the meantime, the predictive performance is tested using the unlabeled datasets to validate the LSTM and SVM models. The experimental findings showed that ANFIS, DT, LSTM, and SVM had a high classification accuracy of 96%, 98%, 97%, and 98% among twelve faults, respectively. The findings confirm that the incorporation of AI classifiers in the simulation and offline environment increases the robustness of fault detection and computational efficiency. The specified method is a unique combination of real-time model embedding and offline deep learning analysis, which eliminates the gap between the theoretical study and its implementation. The framework created can hence offer a reliable and scalable base of AI-based protection systems in next-generation smart microgrids.
Rights Management
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Recommended Citation
Sanikommu, Aravind, "Ai-Driven automatic Fault Detection Systems: Revolutionizing Modern Smart Grids" (2026). Student Theses and Dissertations. 1146.
https://arch.astate.edu/all-etd/1146
