ELEC Oral Comprehensive Exam for Doctoral Candidacy by Xianchao Guo

Thursday, January 29, 2026 at 9:00am to 11:00am

Topic: Artificial Intelligence-Empowered Cascading Failure Prediction in Internet of Things Systems

Abstract

Cascading failures occur when a single incident triggers a chain of successive disruptions across interconnected components. In complex Internet of Things (IoT) systems, accurate prediction of such cascading failures is essential to their timely mitigation, thereby preventing widespread systemic disruptions. This study addresses these challenges through three major contributions. First, we investigate the reliability of artificial intelligence (AI) methods used for short-term forecasting based on IoT sensor data. A novel reliability index that incorporates accuracy, efficiency, and security is developed to provide a quantitative tool for selecting reliable AI models for deployment in critical applications such as smart grids. Second, we propose a comprehensive AI-driven framework that encompasses codified data generation, methodical model comparison, meticulous hyperparameter optimization, and holistic performance evaluation based on the proposed reliability index for predicting failure propagation paths. The proposed framework can support industry stakeholders in selecting and deploying suitable AI methods for cascading failure prediction in real-world IoT systems. Third, to address the complex topological characteristics inherent in IoT systems, we develop an end-to-end principal neighborhood aggregation (PNA)–multilayer perceptron (MLP) architecture for cascading failure diagnosis. Experimental results demonstrate that the proposed architecture significantly outperforms the state-of-the-art AI models, offering a robust tool for enhancing resilience against cascading failures. This proposal also outlines directions for future dissertation research in enhancing the performance of AI models through reliability-focused design, modeling and evaluation. 

Advisor:  Dr. Liudong Xing, Commonwealth Professor, Department of Electrical & Computer Engineering, UMass Dartmouth 

Committee Members:

• Dr. Hong Liu, Commonwealth Professor, Department of Electrical & Computer Engineering, UMass Dartmouth
• Dr. Lance Fiondella, Professor, Department of Electrical & Computer Engineering, UMass Dartmouth
• Dr. Gregory Levitin, Senior Expert, Reliability Department, NOGA-Israel Independent System Operator, Israel 

Note: All ECE Graduate Students are encouraged to attend. All interested parties are invited to attend. Open to the public. 

*For further information, please contact Dr. Liudong Xing

Lester W. Cory Conference Room, Science & Engineering Building (SENG), Room 213A : Zoom Link: https://umassd.zoom.us/j/93501062273 Meeting ID: 935 0106 2273 Passcode: 933143
liudong xing
508.999.8883
lxing@umassd.edu
https://umassd.zoom.us/j/99573148168