论文部分内容阅读
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。
The water plugging process plays a key role in the oil stabilization of the oil field. The water plugging effect prediction technology is an important part of water plugging technology evaluation. In recent years, a great deal of explorations and experiments have been carried out on carbonate reservoirs, and some precipitation and oil-enhancing effects have been achieved. Due to the complex reservoir characteristics of carbonate reservoirs, it is more difficult to predict the effect of water plugging. In this paper, the effectiveness of water blocking effect, carried out the characterization of parameters. Comprehensive application of a variety of statistical learning algorithms, evaluation of the accuracy of the algorithm, the selection of a suitable support vector machine algorithm for prediction of water blockage effectiveness. The established analysis method can accurately predict the effectiveness of water blocking technology and provide the decision-making basis for the selection of water blocking wells.