An ICA-based gas recognition method in the analysis of oil-gas two-phase flow pressure signal in a dispenser system
Jun Liu · Modern Manufacturing Engineering · 2011
Vapor recovery is a new technology relating to energy saving and environment protection in many gas stations.The traditional point holds that the vapor liquid ratio is exactly 1:1.Because of the nondeterminacy of the real environment and the volatility of the fuel,the ratio is generally above 1:1,which leads to the incompleteness of the oil vapor.To achieve the object of determining the relatively accurate vapor liquid ratio,the analysis of oil pressure signal during fueling is considered.FastICA algorithm is used to detangle the original signals acquired so as to extract the independent signal of the oil pressure.Then,amplitude spectrum result of the independent component is obtained through fast Fourier transition.At last,by analysis,the features changing rule of two-phase flow with different air proportions is generalized out and used to recognize the air proportion.Furthermore,it can be regarded as a reference to determine the setting value of the proportion valve in the fuel vapor recovery system.