摘要
提出一种量子LM(Levenberg Marquardt,LM)神经网络与粗糙集相结合的智能识别方法,以替代传统的统计识别方法和工程应用中以单一智能控制为基础的识别方法.基于LM神经网络的技术方案可以整理测井定位数据,提高预测的准确性;量子计算具有并行和类映射的优势;通过削减冗余信息和简化信息量,粗糙集可以降低量子LM神经网络的复杂性,缩短数据处理时间,削减神经网络的负担.通过在石油储层识别实践中的应用证明:该方法可以有效提高计算速度和识别精度,降低成本.
An intelligent identification method of oil reservoir based on quantum Levenberg-Marquardt （LM） neural network and rough set is presented, substituting traditional statistical identification methods and the method based on a single intelligent control in engineering applications. The technical solutions based on LM neural network can organize the logging location data to improve the accuracy of prediction. Quantum calculation has the advantages of parallel and class mapping. By reducing the redundant information and simplifying the information content, rough set can reduce the complexity of quantum LM neural network, shorten the data processing time and lighten the bur-den of neural network. The method is applied to the practice of oil reservoir identification, and it is proved that it can improve the calculating speed and identification precision, and reduce the cost.
出处
《天津师范大学学报（自然科学版）》
CAS
2012年第3期45-50,共6页
Journal of Tianjin Normal University：Natural Science Edition
基金
中国博士后基金资助项目(20090450750)
关键词
量子LM神经网络
石油储层识别
粗糙集
测井数据
智能识别
quantum LM neural network
oil reservoir identification
rough sets
logging data
intelligent identifica-tion
