An Improved BP Neural Network Algorithm And Its Application

Artificial neural networks are made up of a large number of artificial neurons,which abstractly mimics the neural network structure of animal brain and can handle and store information like human brain.A lot of practice has proved that this kind of information processing system has the characteristics of non-linear,non-limitation,nonconformity and nonconvexity,and can find out the optimal solution of complex problem at high speed.Artificial neural network has shown great potential in modern science and technology.Based on the development process,basic principle and common application of artificial neural network(ANN),this paper introduces BP neural network with error back propagation method(BP algorithm)as learning algorithm.This kind of artificial neural network has good approximation,which can approximate arbitrary nonlinear function with arbitrary precision.The structure is simple and the performance is excellent,but there is a problem that the operation speed is slow and it is easy to fall into the local minimum.In order to solve this problem,this paper attempts to improve the BP algorithm by changing the direction of error gradient and using linear search.In order to verify the improvement effect,the acoustic logging data of a drilling in Henan Oilfield are compared with the traditional BP algorithm and the improved algorithm in determining the porosity of the rock.The simulation results show that the traditional BP algorithm needs 488 iterations to complete the operation when the data from the same port is drilled.The improved algorithm only needs 70 iterations to complete the operation,and the improved algorithm Speed has been significantly improved.The results show that the porosity of the formation is closer to that of the drilling geological record,and the operation accuracy is also higher.