Auto-Calibration and Fault Detection and Isolation of Skewed Redundant Accelerometers in Measurement While Drilling Systems.

Autor: Seyed Moosavi SM; Department of Electrical Engineering, Science and Research Branch, Islamic Azad University, Tehran 14778-9385, Iran. m.moosavi@srbiau.ac.ir., Moaveni B; School of Railway Engineering, Iran University of Science and Technology, Tehran 16846-13114, Iran. b_moaveni@iust.ac.ir., Moshiri B; Control and Intelligent Processing, Center of Excellence, School of Electrical and Computer Engineering, University of Tehran, Tehran 14395-515, Iran. moshiri@ut.ac.ir., Arvan MR; Department of Electrical Engineering, Malek-Ashtar University of Technology, Tehran 15875-1774, Iran. arvan@mut.ac.ir.
Jazyk: angličtina
Zdroj: Sensors (Basel, Switzerland) [Sensors (Basel)] 2018 Feb 27; Vol. 18 (3). Date of Electronic Publication: 2018 Feb 27.
DOI: 10.3390/s18030702
Abstrakt: The present study designed skewed redundant accelerometers for a Measurement While Drilling (MWD) tool and executed auto-calibration, fault diagnosis and isolation of accelerometers in this tool. The optimal structure includes four accelerometers was selected and designed precisely in accordance with the physical shape of the existing MWD tool. A new four-accelerometer structure was designed, implemented and installed on the current system, replacing the conventional orthogonal structure. Auto-calibration operation of skewed redundant accelerometers and all combinations of three accelerometers have been done. Consequently, biases, scale factors, and misalignment factors of accelerometers have been successfully estimated. By defecting the sensors in the new optimal skewed redundant structure, the fault was detected using the proposed FDI method and the faulty sensor was diagnosed and isolated. The results indicate that the system can continue to operate with at least three correct sensors.
Competing Interests: The authors declare no conflict of interest.
Databáze: MEDLINE
Nepřihlášeným uživatelům se plný text nezobrazuje