Title:
Failure Prediction of Double Shaft Gas Turbine using Uncertainty Aware Explainable AI

Author:

Ahmad kamal Mohd Nor

Supervisor:

Department:

Mechanical Engineering

Year:

2024

URL:

Abstract:

Error-proof prediction is currently a major interest in machine learning (ML) based-gas turbine (GT) failure prognostics applications, indicated by the rise of probabilistic, ensemble, physics-AI, and...
View more

Multimedia Content:

01

Methodology

Key Methodology Steps 1. Data Collection & Preprocessing • Data from an 18.8 MW twin-shaft industrial gas turbine recorded over one year (8737 hours). • Removed sensor errors and filtered data with...
View more
Methodology Image
02

Citation Analysis

1. Applied Energy (Elsevier) Scope: Advanced energy systems, predictive maintenance, and AI applications in power generation. Relevance: Your work focuses on gas turbine life prediction using Baye...
View more
03

Impact

1. Industrial Impact Improved Reliability: Enables accurate prediction of Remaining Useful Life (RUL) for gas turbines, reducing unexpected downtime. Cost Savings: Optimizes maintenance schedules,...
View more
04

Benefits to Society

1. Scientific and Technical Benefits Novel Approach: Combines Bayesian machine learning with Explainable AI (XAI) for gas turbine prognostics, introducing transparency and uncertainty quantificatio...
View more
05

Applications

1. Gas Turbine Failure Prediction System Application: Predicting Remaining Useful Life (RUL) of twin-shaft gas turbines under fouling conditions. Use Case: Offshore oil & gas platforms and power p...
View more
06

Infographic

Summary of this research
Infographic Image
07

Patents / Products

1. Gas Turbine Failure Prediction Framework Patent: A method for predicting Remaining Useful Life (RUL) of twin-shaft gas turbines using Bayesian LSTM integrated with Explainable AI (XAI). Applica...
View more
08

Experts

1. AP Ir. Dr. Masdi Muhammad Institution: Universiti Teknologi PETRONAS (UTP), Malaysia Contribution: Main supervisor for research on uncertainty-aware explainable AI for gas turbine failure predi...
View more