【摘要】 堵水工艺对于油田的稳油控水起到了关键的作用,堵水措施效果预测技术是堵水工艺评价的重要组成部分。近年来,针对碳酸盐岩储层进行了大量的探索和试验,取得了一定的降水增油效果。由于碳酸盐岩油藏复杂的储集特征,使得堵水效果预测难度加大。本文针对堵水效果的有效性,开展了表征参数的研究。综合应用多种统计学习算法,评价了算法的准确性,优选出适合于堵水有效性预测的支持向量机算法。建立的分析方法,可以准确地预测堵水工艺的有效性,为堵水井选择提供决策依据。更多还原

【Abstract】 Water plugging technology plays an important role in oil-stabilizing and water-controlling of oilfield and predicting water plugging effects is a significant component in technology evaluation.A large number of explorations and practices had been carried out in carbonate reservoir and gained some achievement.Carbonate reservoir characteristicsmake prediction difficult for water plugging effect.In this paper,characterizing parameters are optimized at first.Using statistical learning methods comprehensively,conclusion is made that support vector machine is appropriate for effectiveness prediction of water plugging.The analysis methodology is proved for predicting technology effectiveness accurately and can offer decision support for the target wells selection.更多还原