Research On Downhole MWD Imaging Data Compression Method

MWD imaging technology can be used to identify the change of stratum interface and carry out geo-steering work.MWD imaging instruments can acquire a large amount of downhole data.Limited by the low capacity of downhole memory and the low data transmission rate of conventional mud pulse telemetry system,it is impossible to realize the long-term storage and real-time transmission of a large amount of MWD imaging data,which restricts the development of MWD imaging technology.Data compression technology provides a new way for efficient storage and transmission of MWD imaging data.Combining with the characteristics of downhole MWD imaging data and for the engineering application of geo-steering,aiming at improving the data compression efficiency,this thesis studied the downhole MWD imaging data compression method,and mainly finished the following innovative work:Firstly,in order to improve the transmission efficiency of MWD imaging data,combining with the characteristics of centralized distribution,gentle change and strong correlation of MWD imaging data,a hybrid coding method based on DPCM and integer DCT transform is proposed to compress MWD imaging data.This method reduces the correlation redundancy between data by predictive coding,removes the correlation redundancy in data by integer DCT transform,and removes the statistical redundancy between data by Huffman coding,thus improving the data compression efficiency step by step.Simulation and actual MWD resistivity imaging data processing results showed the validity and feasibility of the method.This method can be used as a general method for MWD imaging data compression.Secondly,in order to further improve the compression efficiency of the proposed hybrid coding method for MWD imaging data,a data rearrangement method is proposed from the perspective of enhancing data correlation.After differential processing of original data by using predictive coding,taking the average distance of rearranged elements as the optimization criterion,constrained with continuous and no crossing rearranging path and data adjacency,the method reorganizes the obtained 1-D difference data into a 2-D matrix with the best correlation.While maintaining the correlation of adjacent data,the correlation between non-adjacent data is enhanced and thus improves the efficiency of removing correlation redundancy by transform coding.Simulation and actual MWD resistivity imaging data processing results showed that under the premise of ensuring the reconstruction quality of MWD imaging data,this method significantly improved the compression efficiency of the existing hybrid coding method.Finally,for the practical need of geo-steering application,a MWD imaging data compression method for peak and subsection slope characteristics is proposed.By analyzing and summarizing these two types of characteristics,a method for extracting peak and subsection slope characteristics of MWD imaging data is proposed.Differential coding is performed on the peak value and its corresponding sector number,and combining the characteristics of alternating positive and negative values of segmented slope,a segmented slope compression method based on k-means clustering and run-length coding is proposed.Categories quantization of segmented slope using k-means clustering method,the clustering center of each kind of segmented slope data set is used as the quantified level to quantify all the data in this class,and run-length coding of quantized segmented slope is performed.Thus,large compression of characteristic data is realized.Simulation and actual MWD resistivity imaging data processing results showed that the method could achieve efficient compression of MWD imaging data on the premise of fully retaining the original data characteristics,and it can be applied in geo-steering application.