A method for predicting colored noise by introducing prediction of nonlinear time series is presented. By adopting three kinds of neural networks prediction models, the colored noise prediction is studied through chan...A method for predicting colored noise by introducing prediction of nonlinear time series is presented. By adopting three kinds of neural networks prediction models, the colored noise prediction is studied through changing the filter bandwidth for stochastic noise and the sampling rate for colored noise. The results show that colored noise can be predicted. The prediction error decreases with the increasing of the sampling rate or the narrowing of the filter bandwidth. If the parameters are selected properly, the prediction precision can meet the requirement of engineering implementation. The results offer a new reference way for increasing the ability for detecting weak signal in signal processing system.展开更多
文摘A method for predicting colored noise by introducing prediction of nonlinear time series is presented. By adopting three kinds of neural networks prediction models, the colored noise prediction is studied through changing the filter bandwidth for stochastic noise and the sampling rate for colored noise. The results show that colored noise can be predicted. The prediction error decreases with the increasing of the sampling rate or the narrowing of the filter bandwidth. If the parameters are selected properly, the prediction precision can meet the requirement of engineering implementation. The results offer a new reference way for increasing the ability for detecting weak signal in signal processing system.
