基于径向基函数神经网络在非地震数据中的应用

针对径向基函数（RBF）神经网络的非线性特点，利用已控点来训练RBF网络，而达到预测未知非地震数据控点的目的。综合已知点和预测控制点，把得到的规则数据体大致对应相应空间进行排布用以全空间成像，最后利用相关软件对处理后的非地震数据进行了三维数据的成像，从而可以显示全息的三维信息，该方法显示出很强的处理问题的能力，同时该仿真结果也表明了该方法的有效性和可行性。

Applications of Radial Basis Function Neural Network in Non-Seismology Data

In allusion to the non - linearity feature of the RBF neural network, and in order to forecast the unknown non - seismology data reference point, we can use the assignment of reference point to train the RBF neural network, thereby integrating the assignment with the unknown reference point. Finally these non - seismology data have been imaged through the software. So we can get better information than before. This method represents the strong ability of dealing with data, at the same time, the simulation results reveal that the method has availability and feasibility.