DocumentCode :
2971154
Title :
The structure and defects recognition algorithm of an aircraft surface defects inspection robot
Author :
Zhiwei, Xing ; Muhua, Chen ; Qingji, Gao
Author_Institution :
Inst. of Aviation Autom., Civil Aviation Univ. of China, Tianjin, China
fYear :
2009
fDate :
22-24 June 2009
Firstpage :
740
Lastpage :
745
Abstract :
The architecture and trajectory tracing control of a climbing robot used for aircraft surface defects inspection are provided in this paper. The robot has been designed with cruciform structure and equipped with jointed legs and suction cup that allow it maneuver over the surface of an aircraft exterior. To enhance the tracing precision and motion stability of the robot, a TLC(trajectory linearization control) control scheme is adopted. And a rivet´s cracks identification algorithm is proposed according to the character of the cracks from a rivet on the surface of aircraft. A star vector algorithm is adopted to extract the feature of rivet with crack defects. And the fuzzy-SVM based algorithm is used to classify the different crack defect around the rivets. The validity of the control scheme and crack defects recognition algorithm is verified by simulation and experiment result.
Keywords :
aircraft; fuzzy set theory; mobile robots; motion control; position control; support vector machines; TLC; aircraft surface defects inspection robot; defects recognition algorithm; fuzzy SVM; fuzzy support vector machine; motion stability; rivet cracks identification algorithm; structure recognition algorithm; trajectory linearization control; trajectory tracing control; Aerospace control; Aircraft; Climbing robots; Feature extraction; Inspection; Leg; Legged locomotion; Motion control; Stability; Surface cracks;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Information and Automation, 2009. ICIA '09. International Conference on
Conference_Location :
Zhuhai, Macau
Print_ISBN :
978-1-4244-3607-1
Electronic_ISBN :
978-1-4244-3608-8
Type :
conf
DOI :
10.1109/ICINFA.2009.5205019
Filename :
5205019
Link To Document :
بازگشت