The Application Of Deep Learning For Condition Monitoring And Fault Diagnosis Of Industrial Processes

Rabbani, Arfyan and Schwung, Andreas and Rusyadi, Rusman (2018) The Application Of Deep Learning For Condition Monitoring And Fault Diagnosis Of Industrial Processes. Bachelor thesis, Swiss German University.

[img]
Preview
Text
Arfyan Rabbani 11401063 TOC.pdf

Download (607kB) | Preview
[img] Text
Arfyan Rabbani 11401063 1.pdf
Restricted to Registered users only

Download (223kB)
[img] Text
Arfyan Rabbani 11401063 2.pdf
Restricted to Registered users only

Download (1MB)
[img] Text
Arfyan Rabbani 11401063 3.pdf
Restricted to Registered users only

Download (862kB)
[img] Text
Arfyan Rabbani 11401063 4.pdf
Restricted to Registered users only

Download (965kB)
[img] Text
Arfyan Rabbani 11401063 5.pdf
Restricted to Registered users only

Download (217kB)
[img]
Preview
Text
Arfyan Rabbani 11401063 Ref.pdf

Download (332kB) | Preview

Abstract

In this paper a condition monitoring and fault diagnosis in industrial processes is made with the deep learning algorithm, with the unlabelled data provided by the Tennessee Eastman Process it is better to use unsupervised learning type of deep learning. Different types of autoencoders (normal autoencoder (AE), denoising autoencoder (DAE), deep autoencoder (Deep AE), and variational autoencoder (VAE) is chosen as the type of deep learning algorithm. Monitoring graphs is created by reconstruction of the data into a new statistic H 2 and the Squared Prediction Error (SPE) robustly. Then the control limit formed by kernel density estimation. This method demonstrated a better result with the VAE, especially with the barely detectable faults from the test data set, such as 3, 5, 9, 10, 11, 15, 19, 20 and 21. VAE shows the overall robustness in the H 2 statistic and SPE reconstruction.

Item Type: Thesis (Bachelor)
Uncontrolled Keywords: Deep Learning; Process Monitoring; Autoencoders; Denoising Autoencoder; Deep Autoencoder
Subjects: Q Science > Q Science (General) > Q325.5 Machine learning
T Technology > TJ Mechanical engineering and machinery > TJ163.12 Mechatronics
Divisions: Faculty of Engineering and Information Technology > Department of Mechatronics Engineering
Depositing User: Astuti Kusumaningrum
Date Deposited: 09 Nov 2020 14:00
Last Modified: 09 Nov 2020 14:00
URI: http://repository.sgu.ac.id/id/eprint/827

Actions (login required)

View Item View Item