Course info

Course description:

This course provides the concepts and methods of prognostics and health management (PHM) of engineering system, which bright insights into data driven intelligent fault diagnosis and prognosis technologies. A variety of tools and techniques for developing health management and monitoring of components and systems will be discussed. Topics cover signal analysis and processing, feature extraction, machine learning methods for fault diagnosis, data driven prognostics models and remaining useful life prediction. The course presents students to understand the state-of-the-art health monitoring methods and explore applications in various engineering systems.

 Prerequisite:

None

 Textbook

None.

 References:

1.      Yaguo Lei, Naipeng Li, Xiang Li.  Big Data-Driven Intelligent Fault Diagnosis and Prognosis for Mechanical Systems. Springer, 1st ed. 2023 Edition. ISBN-10:‎ 9811691304

2.      Nam-Ho Kim, Dawn An, Joo-Ho Choi. Prognostics and Health Management of Engineering Systems: An Introduction Springer, 1st ed. 2017 Edition. ISBN-10‏: ‎ 3319447408

3.      Amiya R Mohanty. Machinery Condition Monitoring Principles and Practices. CRC Press, 2014. ISBN-10: 1466593040

 

Course objectives:

1.      Introduce to students the fundamental theories and cutting-edge methods on fault diagnosis and prognosis of engineering systems.

2.      Introduce multi-domain signal processing and feature extraction, intelligent diagnosis models, machine learning methods, and remaining useful life prediction approaches.  

3.      Train students to conduct experimental demonstrations, explore application cases and test the PHM methods.

 

 

Topics covered:

1.        Introduction

·        Introduce the background and address the critical challenges in the field of prognostics and health management;

·        Present the recent development of prognostics and health management;

·        Content covered in the course.

2.        Signal Acquisition

·        Vibration Monitoring

·        Noise Monitoring

·        Thermography

3.        Signal analysis, Processing

·        Time domain signal analysis

·        Frequency domain signal analysis

·        Wavelet package transform

4.        Feature Extraction

·        Empirical model decomposition

·        Variational model decomposition

·        Feature extraction

·        Feature selection

5.        Machine Learning Methods for Fault Diagnosis

·        Artificial neural networks (ANN) based approaches

·        Single layer neural network (SVM, ELM) based approaches

·        Simultaneous faults diagnosis

·        Case studies

6.        Advanced Health Recognition Approaches

·        Deep learning based diagnosis

·        Capsule Network with dynamic pruning

·        Hierarchical cascade forest

·        Case studies

7.        Prognostics Models

·        Remaining useful life prediction

·        Case studies

8.        Experimental Verification

·        Build a simulation experiment platform for electromechanical equipment in the laboratory

·        Simulate the malfunction operation state

·        Data analysis and fault diagnosis