基于数据挖掘的三次采油技能培训案例的设计及实现

发布时间：2018-04-17 11:05

本文选题：三次采油 + 技能培训　；参考：《东北石油大学》2017年硕士论文

【摘要】：随着三次采油工业化规模的日益扩大,历史数据的逐年累积,人工智能技术的广泛使用,利用数据挖掘的方法在海量数据中发现有意义知识供员工学习已成为三次采油现场的迫切需求。目前,在三次采油员工技能培训过程中,针对三次采油开发管理平台的使用及专业技能的学习还没有一套完整的培训体系,在培训过程中更是缺少案例的支撑。这种状况会导致以下问题:第一,没有案例样本,员工在使用平台的过程中没有目的性地学习,导致员工不能从根本上掌握业务信息,使员工培训周期变长,业务技能增长缓慢;第二,三次采油业务种类繁多,没有案例辅导,员工不能反复学习巩固,致使员工业务技能知识掌握的不全面。针对以上问题,提出了基于数据挖掘的三次采油技能培训案例模型。首先,根据业务衡量指标提取原始项目集合,提出基于关联度分析的有效特征筛选方法,建立各个实体特征之间的关联关系,计算关联度,筛选出有效特征。在均值化方法处理数据的过程中,引入时间序列,选取多种时间粒度来截取距离数据,通过计算关联度、摒弃无效影响特征和降低数据维度来完成数据的准备过程。其次,利用决策树算法构造决策树,生成规则,并存到规则知识库中。接下来,建立三次采油技能培训案例模型,设计案例样本结构,划分三次采油岗位,匹配规则知识库与岗位之间的映射关系,完成基于岗位划分的三次采油技能培训案例的设计。最后,将案例应用于三次采油员工技能培训的过程中,解决员工培训周期长,业务技能增长缓慢,业务技能知识掌握不全面的问题。基于三次采油技能培训案例的设计及实现的基础上,通过案例教学实施前后员工各方面能力的对比,证实案例教学在三次采油员工技能培训过程中具有良好的应用价值。
[Abstract]:With the increasing scale of industrialization of tertiary oil recovery, the accumulation of historical data year by year, and the extensive use of artificial intelligence technology,Using the method of data mining to find meaningful knowledge in massive data for employees to learn has become an urgent need for tertiary oil production field.At present, there is not a complete training system for the use of the tertiary oil recovery development management platform and the study of professional skills in the process of three oil recovery staff skills training, and there is a lack of case support in the training process.This situation can lead to the following problems: first, without a sample case, the employee does not learn purposefully in the process of using the platform, resulting in the employee unable to grasp the business information fundamentally and making the employee training cycle longer.The growth of business skills is slow; second, there are many types of oil production business, no case counseling, staff can not learn and consolidate repeatedly, resulting in the lack of comprehensive knowledge of the staff's business skills.Aiming at the above problems, a case model of three times oil recovery skill training based on data mining is proposed.Firstly, the original item set is extracted according to the business measurement index, and an effective feature selection method based on the correlation degree analysis is proposed. The correlation relationship between the various entity features is established, the correlation degree is calculated, and the effective feature is screened out.In the process of processing data by means of average method, time series is introduced, and a variety of time granularity is selected to intercept the distance data. By calculating the correlation degree, the invalid influence feature is abandoned and the data dimension is reduced to complete the preparation process of the data.Secondly, decision tree algorithm is used to construct decision tree, generate rules, and co-exist in rule knowledge base.Then, the case model of three-time oil recovery skill training is established, the sample structure of the case is designed, the post of tertiary oil recovery is divided, and the mapping relationship between the knowledge base of matching rules and the post is established.Complete the design of three oil recovery skill training cases based on job division.Finally, the case is applied to the process of three oil recovery staff skills training to solve the problems of long training period, slow growth of business skills and incomplete knowledge of business skills.Based on the design and implementation of three oil recovery skill training cases, it is proved that case teaching has a good application value in the process of three oil recovery skills training by comparing the abilities of employees before and after the implementation of case teaching.
【学位授予单位】：东北石油大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.13;TE3-4;C975

【相似文献】