Multi-segment Identification Method for Multi-sensor Multi-channel ARMA Signal

For the multi-sensor multi-channel ARMA (the Auto-Regressive and Moving Average model) signal with unknown model parameters and noise variance, a multi-segment identification method is proposed. The first segment utilize Recursive Instrumental Variable method (RIV), using the method of average local model parameter estimators to get the unknown model parameters online information fusion estimators. The second segments utilize related methods, using the method of average local noise variance estimates to get the unknown noise variance online information fusion estimators. The third segment utilize related methods and Gevers-Wouters algorithm with dead zones to get the MA model information fusion parameter estimators.

Provided by: Science and Development Network (SciDev.Net) Topic: Security Date Added: Mar 2013 Format: PDF

Find By Topic