A numerical simulation noise reduction algorithm for removing spectrum overlap noise and an adaptive threshold recognition mud pulse signal algorithm in MWD

Abstract

Measurement While Drilling(MWD) is a signal transmission technology widely used in geological exploration and oil drilling. With the development of geological drilling, the requirements for the types of downhole parameter and transmission rate are increased, and the difficulty of ground signal processing is also increased. Ground signal processing mainly includes signal noise reduction and signal recognition. At present, the reduction of mud pump noise whose spectrum overlaps the mud pulse signal has not been resolved. The noise reduction algorithm based on a single pressure signal can not solve the problem that the noise overlapped with the signal in the frequency domain. In pulse recognition, the fixed threshold method is usually used to recognize useful pulse signals. Due to the interference of the complex downhole working conditions, there are a large number of missed and false detections in pulse recognition. This paper analyzes the Time Delay differential Algorithm for mud noise reduction and the effectiveness of the Time Delay differential Algorithm through numerical simulation. And then, this paper discusses the influence of algorithm parameter setting on the processing result. In the useful pulse signal recognition, the paper uses the one-dimensional signal OTSU method to design an adaptive threshold for the signal recognition of Frequency-shift keying (FSK) and Positive mud pulse signal. The result represents that the effective pulses are extracted by the adaptive threshold which avoids the shortcomings of the current signal recognition algorithm.