Application of Wavelet Neural Networks to Forecast Deformation Monitoring
Chen Si-sheng · Hydrographic Surveying and Charting · 2009
Wavelet neural network has good localization characteristics both in time-domain and frequency-domain,and also has the self-learning function of neural network.The wavelet series,which are got by wavelet resolution in shift and scale transform,have the characteristics of the average approximating functions by wavelet resolution.So it can be used to forecast deformation monitoring.There are two new parameters added into wavelet neural network,which are scale factor and shift factor.Therefore,it has more degree of freedom than wavelet resolution to make its approximating function more flexible and effective and make its fault tolerance better.