测井岩性识别新方法研究

A novel method for well logging lithologic identification

为了更好地解决测井岩性识别问题,引入了一种基于粒子群优化的支持向量机算法.通过实际测井资料和岩性剖面资料进行学习训练支持向量机,并利用粒子群优化算法对支持向量机参数进行优化,建立了测井岩性识别的支持向量机模型.应用该方法对准噶尔盆地某井的测井岩性进行识别,并将该方法的识别结果与BP神经网络方法的识别结果进行了比较,结果表明该方法优于BP神经网络方法,具有识别正确率高、收敛速度快、推广能力强等优点.

A novel support vector machine based on particle swarm optimization（PSO-SVM） is proposed for better solving the well logging lithologic identification problem.An identification model for well logging lithologic is established using the data of actual well logging and lithologic profile by training the SVM,which is optimized by PSO algorithm.The proposed method is applied to certain well in Junggar Basin and the experimental results show it has higher identification precision,faster convergence speed and better generalization effect than BP neural network based approach.