摘要
神经网络具有良好的学习特性,小波变换有良好的时频局部化性质,将二者结合在一起构成小波神经网络兼有神经网络和小波变换的优点。本文提出了解决虚拟仪器系统非线性校正问题的小波神经网络算法。最后通过一个应用实例表明,采用小波神经网络建立软校正模型,不仅可以使系统获得高精度,而且在相同的误差条件下,其收敛速度也要远远快于传统的BP神经网络。
Neural network has good learning characteristics, and wavelet transform has good localization characteristics both in time-domain and frequency-domain. The wavelet neural network (WNN) can be obtained by combining, which has better characteristics comparing with neural network and wavelet transform. In this paper, wavelet neural network algorithm of solving the problems on the nonlinear correction of virtual instrument system is put forward. A application example is given to demonstrate that the correction model of WNN may cause the system to obtain the high accuracy. Moreover. the convergence rate of WNN is also faster than the speed of BP neural network under the same error condition.
出处
《微计算机信息》
北大核心
2005年第12S期157-159,共3页
Control & Automation
基金
四川省教育厅青年基金资助项目(2004B024)
关键词
非线性校正
小波神经网络
软校正
nonlinear correction
wavelet neural network
soft correction