Research on Interlayer Recognition Based on Intelligent Optimization Algorithms and Convolutional Neural Networks

Due to the particularity of interlayer logging curves in reservoirs, as well as the traditional interlayer identification has many problems, such as small amount of data, poor adaptability, low discrimination among interlayer types and strong subjectivity, an interlayer identification method has been proposed based on the intelligent optimization algorithm and Convolutional Neural Networks (CNN) in this paper. In this method, the parameters and structure of the model are optimized by genetic algorithm, particle swarm algorithm and artificial fish swarm algorithm, so that the network model can be adjusted based on the basic architecture and characteristics of the data set. The experimental results show that through the optimization process, three intelligent optimization algorithms are used to build the model structure, and the obtained CNN can avoid artificial selection of the model structure, and evolve a model that is more suitable for the data set, thereby reducing the number of model parameters, shortening the training time and improving the accuracy.