节点文献
一种基于多级神经网络分类器的沉积相识别方法
A Method of Sedimentary Facies Recognition Based on Multilayer NN Classifier
【摘要】 由于油层沉积特性本身的多样性和复杂性,造成了沉积相各类别之间测井曲线形态差异较小,这无疑增加了识别难度。针对这一问题,给出了一种新的沉积相识别方法。该方法通过将神经网络和多级分类器相结合,有效地提高了识别度。最后用实例验证了该方法的正确性。
【Abstract】 Owing to the variety and complexity of the sedimentary characteristics of the layer itself, subtle differences of the logging curves should undoubtedly increase the hardness of recognition. Considering this problem, the paper presented a new recognition method of sedimentary facies. The method effectively raised the recognition rate by combining the error accumulation of the neural network and multilayer classifier. The method was tested by using the real examples at last.
【关键词】 多级分类器;
神经网络;
沉积相识别;
差错累积;
【Key words】 neural network; multilayer classifier; sedimentary facies recognition; error accumulation;
【Key words】 neural network; multilayer classifier; sedimentary facies recognition; error accumulation;
- 【文献出处】 应用科技 ,Applied Science and Technology , 编辑部邮箱 ,2003年03期
- 【分类号】TP183
- 【被引频次】7
- 【下载频次】126