论文部分内容阅读
为解决设计BP神经网络时所遇到的网络参数和连接权重难以确定,在随机扰动下不能达到最佳效果,学习时间较长难以满足系统实时性要求等问题,引入进化神经网络.根据舰载导航的要求及特点,对基于进化规划的BP神经网络进行设计,并将之应用于舰载导航系统中.仿真结果证明,该方法加快了神经网络的收敛速度,解决了BP神经网络存在的问题.并在舰载导航系统外观测数据不可得时,规避卡尔曼滤波所存在的问题,从而保证了卡尔曼滤波器的正常运行,进一步提高了舰载导航系统的精度.
In order to solve the problems that the network parameters and connection weights encountered in the design of BP neural network are hard to be determined, the best results can not be achieved under stochastic disturbance, the learning time is long and it is difficult to meet the real-time requirements of the system, the evolutionary neural network Navigation requirements and characteristics of BP neural network based on evolutionary programming, and apply it to shipborne navigation system.The simulation results show that this method accelerates the convergence speed of neural network and solves the problems of BP neural network In addition, the problem of Kalman filter is avoided when the external observation data of shipborne navigation system is not available, so as to ensure the normal operation of Kalman filter and further improve the precision of shipborne navigation system.