Shale gas reservoir recognition method based on self-organizing competitive neural network

Abstract

The invention discloses a shale gas reservoir recognition method based on a self-organizing competitive neural network. The method comprises the following steps that a, data of drilled shale gas horizontal wells in the same block are synthesized and classified; b, regional data are used for conducting optimization with while-drilling parameters as training samples; c, the while-drilling parameters are subjected to normalization processing; d, a regional SOM neural network model is established; e, reservoir recognition is conducted on the forwards-drilled shale gas horizontal wells as forecast samples through the established model. According to the shale gas reservoir recognition method based on the self-organizing competitive neural network, the model capable of effectively recognizing shale gas reservoirs is established by means of SOM neural network analysis with the optimized while-drilling parameters as discrimination parameters according to known well logging interpretation and formation testing results, so that while-drilling interpretation of the shale gas reservoirs is realized, and the problems that shale gas reservoir logging interpretation methods are insufficient and multiple parameters are non-linear are solved.

Description

technical field [0001] The invention relates to a method for establishing a shale gas reservoir identification model based on a self-organizing competitive neural network (namely, a SOM neural network), and belongs to the field of petroleum and natural gas exploration and development. Background technique [0002] Shale gas is natural gas stored in shale in free and adsorbed states. It is gradually becoming a hot spot in oil and gas exploration. Domestic shale gas horizontal well development is about to be carried out on a large scale. For the interpretation method of shale gas while drilling, for example, the study on the logging response characteristics and identification methods of shale gas reservoirs disclosed in the 18th issue of "Science and Technology Herald" in 2012. [0003] However, due to the particularity of shale gas reservoirs, in addition to the comprehensive logging parameters, there is only one logging-while-drilling parameter of gamma while drilling, and t...

AI Technical Summary

Problems solved by technology

According to the known well logging interpretation and oil testing results, the present invention uses the preferred parameters while drilling as the discriminant parameters, and establishes a model that can effectively identify shale gas reservoirs through SOM neural network analysis, so as to realize the identification of shale gas reservoirs. Interpretation while drilling solves the problems of lack of logging interpretation methods and multi-parameter nonlinearity in shale gas reservoirs

Images