百年来，石油工业促进了全球经济的发展，但资源的有限性却成为制约世界各国经济发展的战略问题。我国的石油产业经过近50年的发展，多数油井进入开发的中后期，油层有效含油厚度变薄，储量丰度越来越低，从而导致低产井数量逐年增多。随着企业信息化的推进，各石油生产企业先后建成服务于油井开发的多种数据库，这使得运用数据挖掘技术研究低产油井的生产方案成为可能。 本文以降低生产能耗为目标，以数据挖掘技术为基础，进行了低产油井生产方案研究。将VC++、数据挖掘技术、油井生产方案研究三者结合，提出了油井数据挖掘系统设计和开发的三层体系结构，实现了低产油井生产管理系统平台的搭建。在对某采油区53口油井的数据进行分析的基础上，对于原油凝固点低、含蜡量低的低产油井，运用数理统计的方法制定油井的开、关机工作时间即采取间歇抽油的方案；对于含蜡量高、不适宜间歇工作的油井，设计了连续抽油方案，即按照冲次最低为标准，运用灰色关联分析-最大熵算法分析油井生产参数，挖掘符合标准的冲次值，抽油机在保证不停机工作的情况下调整抽油冲次，不影响原油产量的同时，延长抽油杆的使用寿命，降低生产成本。 低产油井生产管理系统通过SQL2000完成数据库的设计，其中油井原始生产数据、数据处理时的推导数据和最终分析的结果数据均存储于数据库中。借助于VC++6.0平台，实现与数据库的连接、数据挖掘算法，以及低产油井生产管理系统操作界面设计。决策者通过简单操作，了解不同井况的油井所适宜的生产方案，为提高生产效率提供了可靠依据。

In the past centry, the oil industry has promoted the development of global economy, and become a strategic issue restricting the economic development of countries of the world because of its limited resources. Our country’s oil industry goes through nearly 50 years of development, and most of the wells have entered the development of middle and late period. The reservoir effective thickness of the oil-bearing gets thin, and reserves abundance are getting lower and lower, leading to the increase in number of low-yielding wells year by year. With the advance of enterprise information technology, the oil production companies have built up a variety of database in the service of the well development, making the production program of using data mining technology to research low-yielding wells possible . In order to reduce production energy consumption in this paper,the low-yielding wells production program based on the data mining technology is researched. Combining the VC++, data mining technology and the oil well production plan,the three layer architecture of the well data mining system design and development are put forward, realizing the construction of the low-yielding well production management system platform.On the basis of the analysis of the data of 53 wells in a oil production area, for low-yielding wells with low crude oil freezing point and low wax content, mathematical statistics methods are used to formulate wells’ open off work time namely adopting the intermittent pumping scheme. For wells with high wax content or not suitable for intermittent work, the continuous pumping scheme is designed, namely according to the impact for minimum standard, using the grey correlation analysis-maximum entropy algorithm to analyze oil well production parameters, mining the conforming value of flush times. The pumping unit can adjust pumping speed without stoping work, which does not affect the crude oil production and can extend the service life of sucker rod, and reduce the production cost at the same time. Low-yielding well production management system database is designed by SQL2000. The original oil well production data, the derivation of data during the data processing and the final results of the analysis are stored in the database. With the aid of VC++ 6.0 platform, the connection with the database, the data mining algorithm, and the low-yielding oil well production management system interface design are realized.Users can understand the different well conditions of oil wells by suitable production plan through a simple operation, which provides a reliable basis for improving the production efficiency.