摘要
以专家知识和现场经验为基础，将专家系统技术应用于稠油开发领域的蒸汽吞吐中，开发了蒸汽吞吐的专家系统（ＥＳＳＳ）。通过访问现场专家，搜集现场数据及查阅文献等手段，收集了大量有关蒸汽吞吐工艺的知识。将这些知识进行整理、分析、综合，建立了蒸汽吞吐专家系统的知识库。在知识库中，以产生式规则来表示知识，利用树形结构组织知识库；采用面向目标的推理技术。该专家系统包括三个基本的功能模块，能用来筛选宜于采用蒸汽吞吐开发的稠油油藏，选择基本注汽工艺参数以及进行注汽前、后的指标预测。该系统有简单、直观的菜单提示，操作方便简单。利用部分现场资料进行了检验，与现场资料吻合较好。
ased on the knowledge of experts and field experience, an expert system for steam stimulation (ESSS) has been developed by the application of expert system technology of Artificial Intelligence to steam stimulation for the recovery of heavy oil. The ESSS includes a knowledge base, an inference engine and a general data base, which are the basic elements to develop an expert system. A great deal of knowledge about steam stimulation technology have been collected by visiting field experts, collecting field data and thumbing through literature and a knowledge base has been constructed. The system uses the production rules to represent knowledge, utilize the tree shape structure to organize the knowledge base, employ goal oriented reasoning technology, backward algorithm and depth first searching. The ESSS makes use of simple links list as general data base. The ESSS is composed of three function modules which can be used to screen reservoir suitable for steam stimulation, to select basic steam injection parameters and to predict part of recovery index. There is simple visual menu prompt in system for operating convenience. Testing by part of field data, it is agreed with field data well which shows that the system is reliable and practical.
关键词
蒸汽吞吐
专家系统
稠油开发
注蒸汽
steam stimulation, expert system, reservoir screen, selection of steam injection parameters, index prediction
