神经网络在石油钻井投资分析中的应用.pdfVIP

神经网络在石油钻井投资分析中的应用 摘 要 人工神经网络是由大量的神经元互联组成，模拟大脑神经处理信息的方式并对 信息进行并行处理和非线形转换的系统。通过样本信息对神经网络的训练，使其具 有与大脑相类似的记忆、辨识能力。按照网络的拓扑结构和运行方式，神经网络可 分为没有反馈的前向网络和相互结合性网络。前向网络由输入层，中间层（隐含层） 和输出层组成，中间层还可有若干层，每层的神经元只接受前一层的输出。本文通 过人工神经网络预测技术模型的实践和研究，针对钻井过程中的影响钻井投资的因 素，选用了拟合比数学模型方法更逼近、无需对输入样本作任何事先的假设即可从 训练样本中发现隐含的信息的人工神经网络作为石油钻井预测方法。 关键词：神经网络，BP 网络，石油钻井，投资预测 I 神经网络在石油钻井投资分析中的应用 Abstract Artificial neural network is made up of a large number of neurons interconnected to simulate brain neural way, it deals with the parallel processing of the information and nonlinear transformation system. Through the training for the neural network by the information of the sample , it is made to have memory and distinguishing ability that is similar to brain. According to the network’s topological structure and operational way, the neural network can be sort of no feedback back －propagation network and associative network. Back －propagation network is made up of input layer, the intermediate layer (imply layer) export layer, intermediate layer can have several layers, each layers of neurons only accepts previous layer’s outputs. Through studying and practising the artificial neural network predicted model, this text is aiming at the factor that influence the well drilling investment. Select artificial neural network that more fitted than the model method of mathematics and no need to input sample to find the imply message from the training sample as any assumption. Key words ：Neural network, BP network, petroleum well drilling, investment prediction II 目 录 第一章 前 言