Extraction of Time-Domain Characteristics and Selection of Effective Features Using Correlation Analysis to Increase the Accuracy of Petroleum Fluid Monitoring Systems

In the current paper, a novel technique is represented to control liquid petrochemical and petroleum products passing through transmitting pipe. A simulation setup, including an X-ray tube, detector, pipe, was conducted by Monte Carlo N Particle-X version (MCNPX) code examine two-by-two mixture of four diverse (ethylene glycol, crude oil, gasoline, gasoil) in various volumetric ratios. As feature extraction system, twelve time characteristics were extracted from received signal, most effective ones selected using correlation analysis present reasonable inputs for neural network training. Three Multilayers perceptron (MLP) networks applied indicate volume ratio three kinds products, fourth product can be feasibly achieved results aforementioned networks. this study, increasing accuracy placed on agenda, RMSE < 1.21 indicates high accuracy. Increasing predicting ratio, which due use appropriate as input, important innovation why proposed system used efficient method oil industry.