Real-time breakover detection during pickup weight step for friction test using machine learning techniques
AssigneeNational Oilwell Varco, L.P.
InventorsFrancisco Ocegueda-Hernandez, Ricardo Vilalta
Life of this patent
Estimating the term…
Loading patent…
AssigneeNational Oilwell Varco, L.P.
InventorsFrancisco Ocegueda-Hernandez, Ricardo Vilalta
Estimating the term…
Detecting a breakover event during a friction test includes obtaining a hookload measurement for each of a series of time windows and generating a linear model and a nonlinear model from the plurality of hookload measurements. During run time, from the nonlinear model, an inflection point is identified from the nonlinear model, where the inflection point is determined to have occurred at a particular time window. A hookload value associated with the linear model, and a hookload value associated with the nonlinear model is determined for the particular time window. A breakover event is determined to have occurred at the particular time when the hookload value associated with the linear model at the particular time window exceeds the hookload value associated with the nonlinear model at the particular time window.
A method for detecting a breakover event, comprising, during a friction test:
A non-transitory computer readable medium comprising computer readable code for detecting a breakover event, the computer readable code executable by one or more processors to:
A system for detecting a breakover event comprising:
Dependencies are inferred from the claim text.