Inverse Deduction of Parameters of Ring of Zonal Testing Instrument Based on RBF Neural Network Model

Abstract

According to the demand of exploiting the poor-thin oil layer in Dagang oilfield, Zonal testing Instrument designed. The sealing function which is respectively accomplished through the ring is the key techniques of Zonal testing Instrument. For the aim of designing the sealing function, firstly, the sealing structure of ring is introduced. Then in order to calculate contact stress of test model with combinations of different levels of parameters corresponding to different compactions of sealing ring structure, FEM model of sealing ring was established and analyzed. At the same time with normalized different levels of parameters of sealing ring structure for input targets and simultaneously normalized results of model test for output variables, parameters of ring structure were inversely deducted with RBF neural network model; With the use of these parameters were carried out with ANSYS software in this paper, shows that it is feasible that unsaturated parameters are inversely deduced with RBF neural network mode.