Autonomous contour tracking industrial robot using neuro fuzzy
A lot of work has been done to automate this process in order to realize self learning capability. Non contact contour tracking has potential in robotic glasssealing, welding and painting applications. In this aspect, the robot learns the contour surface to be tracking autonomously. However, one of...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2013
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/id/eprint/33088/1/EaswandySaiianMFKE2013.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | A lot of work has been done to automate this process in order to realize self learning capability. Non contact contour tracking has potential in robotic glasssealing, welding and painting applications. In this aspect, the robot learns the contour surface to be tracking autonomously. However, one of the problems in contour following technique is how to allocate sample according to curve complexities. For a simple curve, it will have less sample points, while more complex one will be represented by more number of points. In order to realize the above-mentioned objectives, neuro fuzzy contour tracking analysis has been studied in this work. Two inputs and one output ANFIS has been created and it will be used to improve the curve error and noise level. There are more points on complex curve and lesser points on simple or flat curve. If the slope is sharp and the difference between the previous and current slope gradient is large, then it represents a complex curve which requires more number of sampling points. Based on this consideration, twenty-five fuzzy rule have been established and using ABB revolute robot. The analysis shows that the contour tracking using neuro fuzzy can be improved. |
|---|