A preliminary study on wellbore flow interpretation of fiber optic vibration signals based on K-means clustering algorithm

Abstract

Article Highlights A distributed fiber optic logging automatic identification technique based on K-mean clustering algorithm model is proposed. The technique uses distributed fiber-optic acoustic sensors to collect wellhead data by injecting profile logging, and the analysis results are consistent with the production reservoir absorption intensity results. The wellbore flow interpretation of optical fiber vibration signal based on K-means clustering algorithm can effectively classify distributed optical fiber data and solve the content problem in production layer analysis. Better accuracy than distributed temperature logging.

Keywords

• Distributed fiber optic acoustic wave sensor
• Fiber optic vibration signal
• Nonlinear least squares fitting
• K-means clustering algorithm
• Accuracy