论文部分内容阅读
论述了小波分析和神经网络在惯性导航系统初始对准中的重要作用 ,给出了惯导系统初始对准的线性和非线性模型。基于小波变换 ,运用小波神经网络对目前该问题进行了仿真研究。仿真结果表明 ,对于不同的误差模型 ,小波神经网络采用不同的基函数可以很好地对非线性系统进行逼近 ,近似精度高 ,而且网络规模比BP、RBF等神经网络规模要小 ,计算量少 ,收敛速度快。
The importance of wavelet analysis and neural network in the initial alignment of inertial navigation system is discussed. The linear and nonlinear models for the initial alignment of inertial navigation system are given. Based on the wavelet transform, the wavelet neural network is used to simulate the current problem. The simulation results show that for different error models, the wavelet neural network adopts different basis functions to approximate non-linear systems well and the approximation accuracy is high. Moreover, the network size is smaller than the BP neural networks such as BP and RBF, , Convergence speed.