摘要
针对非线性预测问题,提出了小波神经网络算法.该算法采用权重贡献率分析法和关键神经节点法分析权重,精进模型,利用具有优良渐进性的递推预报误差法训练小波的尺度因子和平移因子,并提出了一种网络的改进算法.通过对导航设备的仿真预测,该算法优于同等规模的BP神经网络,其收敛速度快,预测精度高.
Aiming at the problem of non-linear prediction,the arithmetic of wavelet neural network is presented.The contribution rate of weight and key neural node arithmetic are used to analyze weights and improve models,and scale factor and displacement factor are studied by the predicting error method with an excellent recursive character excellent,besides,an improved arithmetic of network is given.By the simulation of navigation equipment,it is superior to the traditional BP neural network,and it is fast in its convergence speed and has a good predicting precision.
出处
《光电技术应用》
2005年第2期44-46,共3页
Electro-Optic Technology Application
关键词
小波神经网络
非线性预测
BP神经网络
导航设备
wavelet neural network
non-linear prediction
BP neural network
navigation equipment
