新疆石油地质

新疆石油地质 ›› 2008, Vol. 29 ›› Issue (3): 382-384.

• 应用技术 • 上一篇    下一篇

支持向量回归机在气井产能预测中的应用

童凯军, 单钰铭, 李海鹏, 李幸运   

  1. 成都理工大学油气藏地质与开发工程国家重点实验室, 成都610059
  • 收稿日期:2007-10-29 发布日期:2020-08-19
  • 作者简介:童凯军(1984-), 男, 安徽安庆人,在读硕士研究生,油气田开发,(Tel)13832694075(E-mail)tongkaijun714@126.com.

Application of Support Vector Regression Machine to Productivity Prediction in Gas Well

TONG Kai-jun, SHAN Yu-ming, LI Hai-peng, LI Xing-yun   

  1. State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation, Chengdu University of Technology, Chengdu,Sichuan 610059, China
  • Received:2007-10-29 Published:2020-08-19

PDF

264

摘要： 气井产能预测是气藏工程研究中用于指导气田合理生产的重要工作和任务, 它在气田的整体评价和高效开发进程中具有很强的预见性和主动性。以测井解释结果为基础, 引入近年来预测效果较好的支持向量回归机技术, 建立了气井产能预测的基本模型, 用来对无动态资料的气井进行产能预测。实例分析表明, 该方法预测精度与传统的神经网络技术方法相比有明显提高, 它是一种较为适用和可靠的气井产能预测评价方法。

关键词: 气井, 产能预测, 气藏, 支持向量机

Abstract: The gas well productivity prediction is one of the main tasks to guide the reasonable production of a gas field in gas pool engineering study. It has intensely forecastive and initiative in the whole evaluation of a gas field and in the process of high efficiency development. Based on the well log interpretation, this paper introduces the support vector regression machine (SVRM)which has preferable forecasting effect to set up the forecasting model for gas well productivity, being suitably applied to prediction of the gas well productivity without dynamic data. The case study shows that this method has more accuracy than traditional nerve network and is a successful evaluating method for gas well productivity prediction. Meantime, it offers a new way for fast quantitative assessment of oil-gas reservoirs.

Key words: gas well, productivity prediction, gas pool, support vector machine

中图分类号: