A data rearrangement method and system for data compression while drilling

Abstract

The invention discloses a data rearrangement method for data compression of imaging while drilling, which comprises: acquiring data arranged in the form of one-dimensional data sequence containing N 2 For the original measurement data of elements, using the correlation constraints, according to the sequence order i of each element, rearrange the original measurement data into a sequence arranged in the form of an N*N matrix and generate corresponding multiple two-dimensional data rearrangement paths; According to the distance of the two-dimensional data rearrangement path, the optimal data rearrangement path with the shortest rearrangement element distance is selected from the two-dimensional data rearrangement path, so as to determine the two-dimensional matrix data sequence corresponding to the optimal data rearrangement path, Provides input data that also has the best correlation between non-adjacent data for the data compression process. On the premise of low calculation amount, the present invention maintains the correlation between adjacent data in one-dimensional sequence, strengthens the correlation between non-adjacent data, improves the efficiency of data compression, and prolongs the storage time of underground data.

Description

technical field [0001] The invention relates to the technical field of logging-while-drilling, in particular to a data rearrangement method and system for compressing measurement-while-drilling data. Background technique [0002] Logging While Drilling (LWD) utilizes logging instruments deployed in drill collars to measure petrophysical parameters in real time while drilling, and transmits the measurement results to the ground for processing in real time with the help of a data telemetry system. It is mainly used in Drilling engineering optimization and formation evaluation. During the operation of tools while drilling, a variety of logging tools are attached to the drill collar, such as acoustic wave, resistivity, NMR, gamma, etc. The joint interpretation of the logging results of multiple tools can improve the accuracy of formation evaluation. With the development of LWD technology, downhole sensors are becoming more and more abundant, and the amount of data acquired is g...

AI Technical Summary

Problems solved by technology

There are relatively few researches on LWD data compression methods in China, mainly including the following four categories: (1) LWD data compression method based on LZW (Lempel Ziv Welch) algorithm, but the LZW compression process needs to continuously query the dictionary and Sequentially output compression code, slow speed, poor timeliness; (2) LWD data compression method based on wavelet transform, this method reduces the complexity of the SPIHT algorithm, improves the running speed of the algorithm, and designs the algorithm Parallel implementation structure, but the wavelet transform only exists reversible transform when it meets the admissible condition of the wavelet function, which has limitations; (3) LWD data compression method based on discrete cosine transform (Discrete Cosine Transform, DCT), which (4) The logging-while-drilling data compression method based on predictive coding (Differential Pulse Code Modulation, DPCM), and the hybrid coding method based on DPCM and DCT transformation applied to imaging data compression while drilling , this method obtains a higher compression ratio, and at the same time, the computational complexity of the algorithm is low, which is suitable for use in downhole instruments

[0004] The compression principles and performances of the above methods are not the same, but they all directly process the one-dimensional raw measurement data

However, the data obtained by logging tools in the actual engineering operation process has spatial correlation between the data, that is, when the one-dimensional data is used to represent, there is a strong correlation between the current data and the data far away, and the one-dimensional representation destroys the This correlation, the correlation between data, has not been fully exploited in existing methods

Images