论文部分内容阅读
介绍一种磁场的新模型,它基于四元代数形式的神经网络。任一种系统的特征都可用简单的几个由综合图解法描述的参数来表征。设计出的识别算法.可使这些参数与实验数据相吻合。用可以设计水下航行体(灭雷具及其组件)的CHAMP3D软件实现了这些方法,并给出了澄清传感器对识别影响的一些数值结果。
A new model of magnetic field is introduced, which is based on the neural network of quaternion algebra. The characteristics of either system can be characterized by a few simple parameters that are described by the integrated graphical method. The designed recognition algorithm can make these parameters agree with the experimental data. These methods are implemented with CHAMP3D software that can design underwater vehicles (destroyers and their components) and give some numerical results that clarify the effect of sensors on identification.