摘要
描述了一种光栅信号的神经网络细分方法 ,阐述了细分原理和神经网络的训练算法。给出了仿真实验 ,结果表明 ,使用该方法只要用少量的训练样本即可达到较高的细分精度 ,使分辨率得到很大提高 ,简化了硬件设计 。
This paper describes a method of subdividing grating signals to which a neural network is applied The subdividing principle,as well as the algorithm of training the neural network,is expounded A simulation experiment is given,and the results of experiment show that the subdividing accuracy is very high so long as to use se veral training samples So,not only the hardware designing is greatly simplified,but also the resolution and re liablity of system are remarkably improved
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2003年第3期264-267,271,共5页
Chinese Journal of Scientific Instrument
关键词
光栅信号
分辨率
神经网络
测控系统
仿真
细分原理
Grating signal Resolution Neural network Measuring and controlling system Simulation
