河流-三角洲储层沉积微相模糊推理网络的识别

针对河流-三角洲储层沉积微相自动识别问题，根据取芯井分析资料和专家解释结果，确定研究区块的沉积微相类型和地质规律，建立了标准模式库和专家经验库，选择和提取判别特征指标的基础上，构建了可同时处理定性专家知识和定量数据的加权模糊推理神经网络进行微相判别的方法．考虑过渡性沉积相在识别中存在的多解性，在小层对比基础上，参照邻井同层微相识别结果，依据区块地质规律采用模糊逻辑推理方法确认和修正微相识别类型，保证平面沉积相和小层单井相的一致性．通过对实际资料处理，其方法的符合率达到84．1％．

The Recognition of the Fuzzy Reasoning Net in River-delta Reservoir Sedimentary Microfacies

To solve the problem of automatic recognition in river-delta reservoir sedimentary microfacies, this papaer puts forward a new method. First, it ascertains the type of the section-block sedimentary microfacies and the geological rule according to the analysis data from the cores well and the result explained by expert, and it establishes the standard mode warehouse and expert experience warehouse. Then, it selects, extracts and differentiates the characteristic indexes. It constructs the fuzzy reasoning NN with the weighting factor to deal with qualitative expert knowledge and quantitative data to differentiate microfacies. It is considered that the transitional sedimentary microfacies have multiplicity of solutions. On the basis of the small layer contrast and consulting the recognition results with the microfacies in the same layer of the adjacent well, it assures the consistency of the plane sedimentary and the small layer well sedimentary. It affirms and amends the recognition type of microfacies with the fuzzy reasoning method according to the geological rule. Its coincidence rate gets to 84.1% through dealing with the practical data for 23 wells in Daqing oil field Sabei developing area.