Application of identifying fluid properties based on GA-BP neural network in carbonate reservoirs

摘要

It is diffcult to identify fluid properties in carbonate reservoirs with conventional well-log data during the period of oil field exploration. In order to establish an effective method for distinguishing gas/oil/water-bearing zone, a new recognizing approach combined with gas surveying and well-log data has been given in this paper. This approach is based on BP neural network, which is optimized the connection weights and thresholds value and restrainted the learning process by genetic algorithm(GA) using the global optimization characteristic. The result of identification is consistented with the well test in XXX oil field in Pre-Caspian Basin in Kazakhstan. It is proved that the approach is effective and practicable.