Dynamic classification committee machine logging fluid identification method and system
A dynamic classification and fluid recognition technology, applied in character and pattern recognition, instruments, earthwork drilling and production, etc., can solve problems such as low application efficiency, limited performance improvement of fluid recognition models, and limited performance of experts themselves, and achieve improved combination strategies , enhance stability and generalization ability, improve training efficiency and accuracy
Active Publication Date: 2020-09-22
CHINA UNIV OF GEOSCIENCES (BEIJING)
View PDF5 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, limited by the performance of the experts themselves, the performance of the fluid recognition model built by the committee machine is limited, and the actual application efficiency is low.
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
example
[0048] The process of using the dynamic classification committee machine for logging fluid identification is actually a process in which the data set sequentially passes through the input layer, the gate network, the expert layer, the combiner, and the output layer to obtain the optimal training model. Using the optimized model, the logging data of other wells or reservoirs can be automatically processed to realize fluid type discrimination.
[0049] In the machine input layer of the Dynamic Classification Committee, 10 wells in the Bashijiqike Formation in the Kuqa Depression on the northern margin of the Tarim Basin were selected, and a total of 1696 sets of logging data and corresponding test data were constructed. In the training set, the logging data are preferably natural gamma ray logging (GR), array induction resistivity logging (AT90), deep-shallow resistivity difference (ΔAT), compensated density logging (DEN), neutron density logging ( CNL) and acoustic logging (DT)...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The embodiment of the invention provides a dynamic classification committee machine logging fluid identification method and system. The method comprises the steps of constructing a dynamic classification committee machine DCCM; inputting first input data for model training into the gate network for pre-learning, and dividing sub-learning tasks through the gate network to obtain a sub-data set; inputting the sub-data set subjected to data structure simplification into the expert layer for training to obtain a sub-model; optimizing the sub-models by adopting the combiner to obtain a target fluididentification model; and inputting second input data to be predicted into the target fluid identification model to obtain the fluid type output by the target fluid identification model. According tothe method of the invention, the expert training efficiency and accuracy are improved, the stability and generalization ability of the final training model are enhanced, and the intelligent discrimination of the fluid type of the complex reservoir is realized.
Description
technical field [0001] The invention relates to the field of geophysical well logging, and more specifically, to a dynamic classification committee machine logging fluid identification method and system. Background technique [0002] Fluid identification is an important content of geophysical logging interpretation and is of great significance to oil and gas exploration and development. Existing logging technologies include conventional logging and imaging logging. Conventional logging fluid identification methods mainly include overlapping graphs, cross graphs, porosity difference ratios, and sensitive parameter methods. Response relationships among fluid identification plates are established. Compared with conventional logging, imaging logging such as array acoustic wave and nuclear magnetic resonance can better identify reservoir fluid characteristics. In addition, there is another type of method is the artificial intelligence algorithm, which has good nonlinear approxi...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62E21B49/00
CPCE21B49/00G06F18/23211G06F18/214G06F18/2415Y02A90/30
Inventor 谭茂金白洋
Owner CHINA UNIV OF GEOSCIENCES (BEIJING)