基于正交基函数的神经网络色谱曲线自适应在线滤波和压缩

Self-Adaptive Compression and Filter of Chromatographic Curve by Orthogonal Neural Network

由于正交基函数神经网络采用Givens正交学习算法 ,不是迭代算法 ,因而速度快 ,加之其无初始权值选择问题 ,不会陷入局部极小 ,所以本文选择此神经网络用于色谱数据的自适应在线滤波和压缩。实验结果表明 ,本方法滤波效果好 ,压缩比高 ,可实时完成 ,非常适用于色谱数据的在线处理需要。

This paper presents a new method of compression and filter for chromatographic curve by a orthogonal neural network. By adopting a non-iterative Givens learning algorithm, the new network learning algorithm can be learned rapidly and it can avoid false local minima and the initialization of weights and other parameters.This method has not only high compression ratio, but also has ability of adaptive filtering noise.