A logging identification method of water-flooded layer of low permeability sandstone based on PSO-ELM
Logging identification of water-flooded layer is of great significance to the deployment of oilfield development scheme and the enhancement of oil recovery.The flooding type of reservoir in a block of Luliang Oilfield in Xinjiang is mainly sewage flooding,and the logging response characteristics are complex and varied,it is difficult to identify the water-flooded layer effectively by traditional chart identification method.Based on the data of logging,geology,oil testing,and on the basis of logging response characteristics analysis of water-flooded layer,this paper proposed a new water-flooded layer identification method using improved Particle Swarm Optimization(PSO)algorithm and Extreme Learning Machine(ELM).First,correlation coefficient was used to select six main controlling factors,RD,RS,GR,SP,DEN and AC.Secondly,the improved Particle Swarm Optimization was used to optimize the parameters of the ELM model.Finally,the optimized model was used to predict the water-flooded layer in the study area.The results show that the identification coincidence rate of water-flooded layer can reach 91.7％by using PSO-ELM,which is better than the application effect of ELM and traditional identification chart,providing a new idea for logging identification of water-flooded layer.
NETL
NSTL
万方数据
