基于粗糙集和Petri网的油层含油识别研究

油藏受天然气的侵入,就会产生油气分异的现象:天然气会萃取出油藏中的轻组分,造成凝析油藏或轻质油藏;此外,油藏中的重组分脱出,形成沉淀,直接导致储层的渗透率降低,从而影响油气的分布规律。针对油藏中油气分布规律复杂的问题,在传统油层含油判别分析的基础上,提出了一种基于粗糙集和Petri网的油藏建模方法。应用粗糙集的知识约简对油层含油识别相关的岩层厚度、泥质含量等6个指标进行属性选择,提取最简规则,建立Petri网模型,根据Petri网的并行推理达到简洁高效的含油识别。仿真实验结果表明,采用粗糙集与Petri网判断的油气分布规律与现场数据高度接近,精度高,识别速度快,正确率高。可见将粗糙集与Petri网组合用于油层油气识别是有效的。

Recognition at Oil and Gas Distribution in Oilfield Based on Rough Sets and Petri Nets

If the reservoir is invaded by foreign natural gas,it will produce oil and gas differentiation:on the one hand,the light component of the reservoir will be extracted by natural gas to result in condensate reservoir or light oil reservoir;on the other hand,the recombination of the reservoir will be separated to result in solid precipitation.These precipitates will reduce the permeability of the reservoir and even affect the distribution of oil and gas.Aiming at the complex distribution law of oil and gas in oil deposit,based on the discriminant analysis of oiliness in traditional oil layer,we put forward a modeling approach for oil deposit on the basis of rough set and Petri net.By applying the simple knowledge in rough set,attribute optimization could be realized towards the thickness of stratum,shale content and other 4 indexes relevant to the oiliness identification of oil layer,and the minimum identification rule could be acquired for establishing Petri net model.Through parallel inference of Petri net,the concise and high-efficient oiliness identification could be realized.The simulation shows that the distribution law of oil and gas judged after adopting rough set and Petri net is highly close to the practical situation,with fast the identification speed high accuracy.It is obvious that using the combination of rough set and Petri net to identify the oil and gas in oil layer is effective.