论文部分内容阅读
介绍了小波神经网络（ＷａｖｅｌｅｔＮｅｕｒａｌＮｅｔｗｏｒｋ）的结构和算法，并将其应用于色谱数据的滤噪与压缩。在小波神经网络中，采用了Ｂｕｂｂｌｅ母小波和一维搜索变步长共轭梯度优化方法。结果表明，小波神经网络将原始数据压缩至１∶２０时，仍能得到很好的重建谱图。重建的色谱图不仅能保留原始的谱图特征而且有光滑滤噪的效果。对于色谱数据的存储与平滑滤噪具有重要意义。
The structure and algorithm of Wavelet Neural Network are introduced and applied to the filtering and compression of chromatographic data. In wavelet neural network, Bubble mother wavelet and one-dimensional search variable step conjugate gradient optimization method are used. The results show that the wavelet neural network still can get good reconstructed spectrum when the original data is compressed to 1:20. The reconstructed chromatogram not only preserves the original spectral features but also smoothes the noise. For the storage of chromatographic data and smoothing noise is of great significance.