粒子群-支持向量机在油田开发综合经济评价中的应用
周默1，龚勇军2
(1 中国石油伊拉克公司北京 100011；2中国寰球工程公司北京 100012)
关键词：油田开发；综合评价；开发方案；粒子群算法；支持向量机；
中图分类号：F275 文献标识码：A
摘要：通过建立一个可靠、合理、有效的油田开发综合评价模型来避免或消除油田开发过程中各种不合理情况的出现，是一个有效技术途径和可操作手段。本文尝试通过理论和实证分析两方面来构建适合于油田开发项目的粒子群-支持向量机油田开发综合评价模型，并以某油田为例，进行实证分析。实证研究结果表明，通过引入粒子群算法可有效减少支持向量机的计算时间，提高计算效率，使得模型的可靠性、合理性和科学性得到更好保证，同时提高模型的评价精度，具有良好的学习推广价值和评价能力。
PSO - supportvector machine in
oilfielddevelopment Comprehensive Economic Evaluation
Zhou Mo1,Gong Yongjun2
(1 Chinese oil companies in Iraq,
Beijing 100011；2China HQCEC Beijing 100012)
Keywords: Oilfield development;comprehensive evaluation; development programs; particle swarm optimization;support vector machine;
Abstract: The establishment ofa reliable, reasonable and effective field development comprehensive evaluationmodel is an effective technical way and operational measure to avoid oreliminate all kinds of unreasonable conditions in the oilfield developmentprocess. Both theoretical and empirical analysis were adopted to build anintegrated assessment model suited for the oil field development project in thePSO-support vector machine oilfield development, taking a reservoir for an exampleto carry out the empirical analysis. The empirical results show that theintroduction of particle swarm algorithm can effectively reduce the computationtime SVM and improve the computational efficiency, making the model’sreliability, rationality and scientificity better guaranteed, while improvingthe model’s evaluation accuracy. It possesses a good promotional value forlearning and evaluation.