Title:
Energy Performance Based Fault Detection and Prediction of Centrifugal Gas Compressor using Hybrid Machine Learning Model

Author:

Mukhtiar Ali Shar

Supervisor:

Department:

Mechanical Engineering

Year:

2024

URL:

Abstract:

In the oil and gas industry the centrifugal gas compressors are not operating at the optimum performance level, more often with incipient faults or energy performance degradation that result in high e...
View more

Multimedia Content:

01

Methodology

1. Overall Approach The study developed a Hybrid Machine Learning (ML) Model integrated with Energy Efficiency Indicators (EEI) and Energy Performance Indicators (EnPI) to detect and predict faults i...
View more
02

Citation Analysis

1. Energies (MDPI) Scope: Covers energy systems, predictive maintenance, machine learning applications in energy efficiency, and industrial optimization. Relevance: Your work aligns with energy pe...
View more
03

Impact

1. Societal Impact Improved Safety in Industrial Operations: The research enables early fault detection in centrifugal compressors, reducing the risk of unexpected breakdowns, explosions, or hazard...
View more
04

Benefits to Society

1. Societal Benefit Safer Industrial Operations: The hybrid ML model helps detect compressor faults early, preventing equipment failures that could lead to safety hazards in oil and gas plants. Co...
View more
05

Applications

1. Industrial Fault Detection and Diagnosis Application: Real-time fault detection system for centrifugal gas compressors. Use Case: Used in oil and gas plants, refineries, and petrochemical indus...
View more
06

Infographic

Summary of this research
Infographic Image
07

Patents / Products

1. Hybrid Machine Learning Fault Detection Engine Product Type: AI-driven diagnostic software module Use Case: Integrated into industrial monitoring systems or compressor control units to detect m...
View more
08

Experts

1. Dr. Emilio Mastriani Institution: University of Palermo, Italy Contribution: Co-authored research on hybrid clustering models for predictive maintenance of high-pressure industrial compressors....
View more