Data mining is the process of extracting implicit but potentially useful information from incomplete, noisy, and fuzzy data. Data mining offers excellent nonlinear modeling and self-organized learning, and it can play...Data mining is the process of extracting implicit but potentially useful information from incomplete, noisy, and fuzzy data. Data mining offers excellent nonlinear modeling and self-organized learning, and it can play a vital role in the interpretation of well logging data of complex reservoirs. We used data mining to identify the lithologies in a complex reservoir. The reservoir lithologies served as the classification task target and were identified using feature extraction, feature selection, and modeling of data streams. We used independent component analysis to extract information from well curves. We then used the branch-and- bound algorithm to look for the optimal feature subsets and eliminate redundant information. Finally, we used the C5.0 decision-tree algorithm to set up disaggregated models of the well logging curves. The modeling and actual logging data were in good agreement, showing the usefulness of data mining methods in complex reservoirs.展开更多
The first generation coherence algorithm(namely C1 algorithm) is based on the statistical cross-correlation theory, which calculates the coherency of seismic data along both in-line and cross-line. The work, based on ...The first generation coherence algorithm(namely C1 algorithm) is based on the statistical cross-correlation theory, which calculates the coherency of seismic data along both in-line and cross-line. The work, based on texture technique, makes full use of seismic information in different directions and the difference of multi-traces, and proposes a novel methodology named the texture coherence algorithm for seismic reservoir characterization, for short TEC algorithm. Besides, in-line and cross-line directions, it also calculates seismic coherency in 45° and 135° directions deviating from in-line. First, we clearly propose an optimization method and a criterion which structure graylevel co-occurrence matrix parameters in TEC algorithm. Furthermore, the matrix to measure the difference between multi-traces is constructed by texture technique, resulting in horizontal constraints of texture coherence attribute. Compared with the C1 algorithm, the TEC algorithm based on graylevel matrix is of the feature that is multi-direction information fusion and keeps the simplicity and high speed, even it is of multi-trace horizontal constraint, leading to significantly improved resolution. The practical application of the TEC algorithm shows that the TEC attribute is superior to both the C1 attribute and amplitude attribute in identifying faults and channels, and it is as successful as the third generation coherence.展开更多
基金sponsored by the National Science and Technology Major Project(No.2011ZX05023-005-006)
文摘Data mining is the process of extracting implicit but potentially useful information from incomplete, noisy, and fuzzy data. Data mining offers excellent nonlinear modeling and self-organized learning, and it can play a vital role in the interpretation of well logging data of complex reservoirs. We used data mining to identify the lithologies in a complex reservoir. The reservoir lithologies served as the classification task target and were identified using feature extraction, feature selection, and modeling of data streams. We used independent component analysis to extract information from well curves. We then used the branch-and- bound algorithm to look for the optimal feature subsets and eliminate redundant information. Finally, we used the C5.0 decision-tree algorithm to set up disaggregated models of the well logging curves. The modeling and actual logging data were in good agreement, showing the usefulness of data mining methods in complex reservoirs.
基金Project(2013CB228600)supported by the National Basic Research Program of ChinaProject(2011A-3606)supported by the CNPC "12.5" Program of China
文摘The first generation coherence algorithm(namely C1 algorithm) is based on the statistical cross-correlation theory, which calculates the coherency of seismic data along both in-line and cross-line. The work, based on texture technique, makes full use of seismic information in different directions and the difference of multi-traces, and proposes a novel methodology named the texture coherence algorithm for seismic reservoir characterization, for short TEC algorithm. Besides, in-line and cross-line directions, it also calculates seismic coherency in 45° and 135° directions deviating from in-line. First, we clearly propose an optimization method and a criterion which structure graylevel co-occurrence matrix parameters in TEC algorithm. Furthermore, the matrix to measure the difference between multi-traces is constructed by texture technique, resulting in horizontal constraints of texture coherence attribute. Compared with the C1 algorithm, the TEC algorithm based on graylevel matrix is of the feature that is multi-direction information fusion and keeps the simplicity and high speed, even it is of multi-trace horizontal constraint, leading to significantly improved resolution. The practical application of the TEC algorithm shows that the TEC attribute is superior to both the C1 attribute and amplitude attribute in identifying faults and channels, and it is as successful as the third generation coherence.
