| In recent years, Electrical logging has become an important means of exploring resistivity distribution of oil-gas formation in invaded profile, but in order to obtain a fine interpretation on resistivity distribution of invasion profile, we must recourse to the arrayed logging tools and its data processing methods.Compare to the information those tools can provided, the prensent inversion methods are rather finite. The inversion problems of Array data processing Logging are rather nonlinear and multi-value, and traditional algorithms can't keep the balance of stability and efficiency. The main idea of this paper is apply an improved Particle Swarm Optimization(PSO) algorithm to the inverse of Array data processing Logging, and seek after new efficient inversion methods.We mainly researched on the following aspects in this paper:Firstly, electrical logging is introduced simply. Descript methods of processing Electrical logging data and strategy of deal with heterogeneity of two-dimensional inversion of log data.Secondly, we analyzed the principle of PSO algorithm, classification the improved PSO algorithms according to different methods. This article was inspired by the physics and proposed two improved PSO algorithms: FCPSO algorithm and m-PSO algorithm. Testing results demonstrate that those algorithms we proposed are better than PSO. We also compared m-PSO with others improved PSO algorithms, and results show m-PSO algorithm is a more rapid and precise algorithm.At last, the preferable m-PSO algorithm was used in the inversion of Array data processing Logging Plentiful inversion results of EPRL show that m-PSO algorithm can accurately inverse the resistivity distribution of invasion profile with fewer particles than other algorithms, and effectively keep the balance of stability and efficiency of the inverse. m-PSO algorithm is a mothod that can be effectively used in the inversion of Array data processing Logging and it has a well foreground of application. |
PDF Full Text Request