| The development of logging technology needs to be accompanied by the progress of related technologies.At present,logging is moving towards the direction of big data and high rate.But the transmission rate has not kept up with the current development needs,the frequency of signal generator is difficult to be quickly improved,underground direct data storage and data transmission,the signal delay is too high.Adding data compression module to the system while drilling is a better solution.When the transmission bandwidth can not be improved quickly,compressing and transmitting the data is equivalent to indirectly improving the transmission bandwidth and improving the real-time performance of the system.It can be used for the downhole staff to make timely analysis and judgment.This paper mainly studies how to make a suitable compression scheme for LWD system,and verifies the feasibility of the scheme through experiments,and makes a comparative analysis.Through the analysis of logging data,it is found that there are many and complex downhole data,so it is impossible to make a compression scheme for each kind of logging data.Only by analyzing the commonalities of logging data,can we work out a scheme to meet these commonalities.Logging data can be divided into two types according to whether the change is stable or not,one is stationary logging signal,the other is non-stationary signal.In this paper,the compression scheme is developed from three directions,one is for the stable logging signal with high real-time requirements,the other is for the logging data with high compression ratio.In addition,from the perspective of signal sampling,the compressed sensing is studied to reduce the sampling data points and improve the overall real-time performance.The main work and innovation are as follows:1.Through the analysis of the traditional differential pulse code modulation(DPCM)algorithm,it is found that although the algorithm is simple and easy to use,and has a certain compression ability,but in the case of large data fluctuations,the reconstruction effect will decline significantly,the quality of data reconstruction is unstable,and it is easy to misjudge.An improved algorithm is proposed in this thesis.A difference statistics module is added before the coding,and the code length is dynamically adjusted according to the probability and difference distribution of logging data.The experimental results show that the algorithm stability and compression performance are improved,and the maximum compression ratio is 2.4 under the premise of satisfying the error.2.The well logging data compression method based on wavelet transform is studied.Due to the powerful decomposition ability of wavelet,there is a large space for data compression.But the overall complexity of the algorithm is too high,and the computing pressure of the downhole processor is too high.The traditional wavelet is changed to 9 /7 wavelet based on lifting,and the embedded coding is adopted to improve the flexibility of the system.Moreover,lifting wavelet uses integer operation,which will not cause floating-point overflow like traditional wavelet transform.The overall reconstruction accuracy is better than ordinary wavelet transform,and the maximum compression ratio can reach 30.3.The logging data compression method based on compressed sensing is studied.For sparse representation and reconstruction algorithm,the over complete matrix which can be trained is selected to improve the sparsity of logging signal;The reconstruction algorithm chooses the Bayesian learning algorithm which is closer to the logging model.The maximum compression ratio can reach 4. |
PDF Full Text Request