基于分布式检测和时间相关神经网络的训练模拟器设计

Design of Training Simulator Based on Distributed Detection and Time-dependent Neural Network

某型导弹发射设备训练模拟器涉及大量输入输出信号的检测和控制,为了提高硬件系统和软件系统的工作效率,确保仿真过程的实时性,根据对输入信号分别按照空间和时间进行分组的思想,提出了基于分布式检测的硬件系统结构和基于时间相关神经网络的软件算法。这些方法已成功的应用于某型导弹发射设备训练模拟器的设计与研制,并取得了显著的效果。

The detection and control of a great deal of input-output signals were involved in the training simulator for a type of missile launching device, To improve the efficiency of the hardware and software, and to assure the real-time simulation, the hardware structure based on distributed detection and the software algorithm based on time-dependent neural network were proposed according to the idea of grouping the input signals in terms of space and time. These methods were used successfully in the design and development of a training simulator. The application result shows that the methods have high efficiency.