人工神经网络在压裂选井及选层中的应用

根据中原油田砂岩油藏 2 0 0多口压裂井及压裂层的静、动态地质资料和压裂施工参数 ,对影响压裂效果的单因素、多因素和主参数进行了分析。采用人工神经网络技术 ,建立了砂岩油藏压裂选井及选层人工智能系统 ,并将其用于中原油田复杂断块压裂选井及选层识别。利用该系统可以估算压裂后单位厚度油层的日增产油量 ,并能预测压裂效果 ,可为压裂选井及选层提供科学依据。

APPLICATION OF THE ARTIFICIAL NERVE NETWORKS TO THE TARGET SELECTION OF HYDRAULIC FRACTURING

An expertise knowledge database is established on the basis of analyses on the static and dynamic geologic data and fracturing parameters obtained from more than 500 wells that have been fractured in sand reservoirs of Zhongyuan oilfield. The single and multi factors and chief parameters which affect fracturing results are analyzed. An artificial intelligent system for selecting targets of hydraulic fracturing is developed by the use of the artificial nerve network technique. Adopting the artificial nerve network technique, this artificial intelligent system has been used for selecting wells and target formations to be fractured in sand reservoirs of complex faulty blocks fracturing in Zhongyuan oilfield. It can estimate daily increment per meter oil bearing layer, predict post-fracturing effect, provides scientific basis for selecting wells and formations to be fractured. The use of this system can decrease the risks resulted from the traditional selection by experiences over many years and increase the fracturing effeciency.