论文部分内容阅读
人工智能在试井中的应用是一个新的课题,国内外都刚起步。本文描述了试井油藏模型识别专家系统(RMIESWT)原型的研制成果。以压力导数为基础,用计算机模拟人类专家的视觉判断过程,用符号表达曲线,实现推理,是一种很好的智能模拟技术。除了压力恢复数据以外,还利用辅助资料来识别油藏模型,这对于那些实测数据不完整的情况,特别有效。选用基于规则的通用型开发工具CM.1作为专家系统环境。采用分层式结构构造知识库,逻辑关系清晰易懂,非常有利于规则的增添、删除和修改。终端用户同RMIESWT的对话全部是菜单式中文显示,易于被现场使用单位所接受。
The application of artificial intelligence in well test is a new subject, both at home and abroad have just started. This article describes the development of the prototype of the well testing reservoir model identification expert system (RMIESWT). Based on the pressure derivatives, it is a very good technique to simulate the visual judgment of human experts by computer and express the curve by symbols to realize reasoning. In addition to pressure-recovery data, secondary data are used to identify the reservoir model, which is particularly effective for cases where the measured data are incomplete. Use rule-based general-purpose development tools CM.1 as an expert system environment. The use of hierarchical structure of knowledge base, the logical relationship is clear and easy to understand, is very conducive to the addition of the rules, delete and modify. The dialogue between end user and RMIESWT is all menu-based Chinese display, which is easy to be accepted by the site user.