| With the development of nuclear magnetic resonance logging with drilling technology,it is difficulty to store,handle and transimit large amount of nuclear magnetic resonance logging data with the limited storage space,processing speed,and transmission bandwidth in real time.Meanwhile,increasing the capacity of the memory,raising the processing speed of the instrument and enhancing the transmission rate of the mud pulse simply cannot solve the problems.Therefore,this thesis studies the nuclear magnetic resonance logging while drilling data compression method and mainly finished the following innovative work.Firstly,in order to satisfy the real time storage,management and transmition requirements of the data compression process,a nuclear magnetic resonance logging while drilling data compression method based on KL transform coding is proposed with the data characteristics of centralized distribution,partial smoothing transformation and substantial correlation.Through using continuous raster,the logging data isreorganized efficiently.Meanwhile,the redundancy of the redundant data is removed out thoroughly by applying the KL orthogonal transformation and uniform quantization methods.Moreover,a successive run-length coding and Huffman coding methods are present to remove the encoding redundancy of the quantization coefficients and to achieve the secondary compression of logging data.As results,the proposed KL transform coding method can compensate the disadvantage of normal data compression method.Beside,the real time data processing is fulfilled and the quality of the reconstructed signal is ensured.Simulation and actual logging data processing results showed the validity and feasibility of the method.Secondly,considering the high computation complexity,the insufficient storage space and the poor real-time transmission performance of the data compression process under KL orthogonal transformation method,a nuclear magnetic resonance logging while drilling data compression method based on DCT transform coding is proposed.The related redundancy of redundant datas is removed greatly based on DCT orthogonal transformation.Meanwhile,the secondary compression of logging data can be completed sufficiently.Besides,this method can make up the shortages of high computation complexity,the insufficient storage space and the poor real-time transmission performance of the data compression process under KL orthogonal.Simulation and actual logging data processing results showed that this method could realize data real-time compression at a high speed and reserve the characteristic of the logging data.At last,with the problems of high computation complexity and mismatching of data encode/decode of the floating-point operations under DCT orthogonal transformation method,a nuclear magnetic resonance logging while drilling data compression method based on improved integer DCT transform coding is proposed.With the analysis of the DCT transform principles,an arbitrary order integer DCT transform basis general generation approach is proposed.Meanwhile,an integer DCT transform basis optimal selection method is presented,which is suitable for nuclear magnetic resonance logging data.Besides,the orthogonal transforms is realized based on the calculated integer DCT transform basis.As results,the proposed integer DCT transform coding data compression method can both maintain the log data features and overcome the shortcomings of the DCT transform.Simulation and actual logging data processing results showed that this method could be able to implement storing,processing and transmitting of the logging data with higher speed in real time. |