Z.W. Wang, Q.X. Cao, N. Luan, and L. Zhang
In-pipe robot, autonomous localization, production rule, consistentdata fusion, maximum likelihood estimate
This paper proposes a novel in-pipe localization technique for industrial in-pipe robots to implement long-distance subsea pipeline maintenance. The proposed localization technique is that the in- pipe robot first performs rough in-pipe localization by an eddy current sensor detecting girth welds along pipelines, and then achieves precise in-pipe localization between adjacent girth welds by multi-odometer detecting movement distance. To eliminate the bad influence caused by the awful in-pipe working environment on the localization method, the production rules are presented to perform fault-tolerance for those undetected girth welds to ensure a reliable rough localization phase. Besides, to improve localization precision, an optimal multi-sensor data fusion algorithm is introduced to process the redundant movement distance data from the multi-odometer. The localization technique is performed on the developed in-pipe robot. Tests on experimental pipeline system and beach oil pipelines validate the applicability and effectiveness of the localization technique.
Important Links:
Go Back
