Online waviness and material removal prediction in aerospace MRO robotic polishing
Polishing is a common task in aerospace industry, providing better surface quality, and therefore increasing the structure intensity and system working efficiency. Manual polishing often gives low repeatability and it is a tedious task for the workers. On the other hand, traditional robotiz...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
2013
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/54682 |
| _version_ | 1811688916629061632 |
|---|---|
| author | Wang, Runfeng. |
| author2 | Er Meng Joo |
| author_facet | Er Meng Joo Wang, Runfeng. |
| author_sort | Wang, Runfeng. |
| collection | NTU |
| description | Polishing is a common task in aerospace industry, providing better surface
quality, and therefore increasing the structure intensity and system working
efficiency. Manual polishing often gives low repeatability and it is a tedious task for
the workers. On the other hand, traditional robotized polishing has a problem of
ineligible deviation between actual and desired robot path. Therefore, force control
technologies which provide large tolerances of positioning errors are introduced to
the polishing process. However, the problem is that, offline measurement of surface
quality reduces the efficiency significantly. Thus, there is an urgent need of certain
techniques that can estimate the surface quality online, and help with more robust
process control, which is also the objective of this project.
The basic idea is to build a data-driven model, which reveals the relationship
between online monitoring signals and surface quality parameters. The model is
trained offline based on the data obtained from a series of experiments, but through
validation, it could be used for online prediction of surface waviness and material
removal. Inputs of the model are the features extracted from the raw signals of
dynamometer, current sensor and AE sensor, and outputs are the surface quality
parameters, including surface waviness, and material removal. To train the model,
these output parameters are measured offline after each experiment.
The key research issues involved in this project include sensing techniques,
signal processing techniques, feature extraction and correlation modelling study. |
| first_indexed | 2024-10-01T05:39:49Z |
| format | Thesis |
| id | ntu-10356/54682 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T05:39:49Z |
| publishDate | 2013 |
| record_format | dspace |
| spelling | ntu-10356/546822023-07-04T15:33:43Z Online waviness and material removal prediction in aerospace MRO robotic polishing Wang, Runfeng. Er Meng Joo School of Electrical and Electronic Engineering DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation Polishing is a common task in aerospace industry, providing better surface quality, and therefore increasing the structure intensity and system working efficiency. Manual polishing often gives low repeatability and it is a tedious task for the workers. On the other hand, traditional robotized polishing has a problem of ineligible deviation between actual and desired robot path. Therefore, force control technologies which provide large tolerances of positioning errors are introduced to the polishing process. However, the problem is that, offline measurement of surface quality reduces the efficiency significantly. Thus, there is an urgent need of certain techniques that can estimate the surface quality online, and help with more robust process control, which is also the objective of this project. The basic idea is to build a data-driven model, which reveals the relationship between online monitoring signals and surface quality parameters. The model is trained offline based on the data obtained from a series of experiments, but through validation, it could be used for online prediction of surface waviness and material removal. Inputs of the model are the features extracted from the raw signals of dynamometer, current sensor and AE sensor, and outputs are the surface quality parameters, including surface waviness, and material removal. To train the model, these output parameters are measured offline after each experiment. The key research issues involved in this project include sensing techniques, signal processing techniques, feature extraction and correlation modelling study. Master of Science (Computer Control and Automation) 2013-07-19T06:45:11Z 2013-07-19T06:45:11Z 2012 2012 Thesis http://hdl.handle.net/10356/54682 en 122 p. application/pdf |
| spellingShingle | DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation Wang, Runfeng. Online waviness and material removal prediction in aerospace MRO robotic polishing |
| title | Online waviness and material removal prediction in aerospace MRO robotic polishing |
| title_full | Online waviness and material removal prediction in aerospace MRO robotic polishing |
| title_fullStr | Online waviness and material removal prediction in aerospace MRO robotic polishing |
| title_full_unstemmed | Online waviness and material removal prediction in aerospace MRO robotic polishing |
| title_short | Online waviness and material removal prediction in aerospace MRO robotic polishing |
| title_sort | online waviness and material removal prediction in aerospace mro robotic polishing |
| topic | DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation |
| url | http://hdl.handle.net/10356/54682 |
| work_keys_str_mv | AT wangrunfeng onlinewavinessandmaterialremovalpredictioninaerospacemroroboticpolishing |