Sea clutter prediction based on linear and non-linear models
Yingchun Wang · Journal of Dalian Maritime University · 2010
Linear and nonlinear prediction theories were used to establish prediction models based on the nonlinear uncertainty characteristics of sea clutters so as to predict them accurately.Autoregressive(AR) and non-linear model based on radial basis function(RBF) and Volterra series filter(VSF) were adopted to predict the simulated and real sea clutter data by using logistic chaotic signal and the real sea clutter data collected by IPIX radar.Results show that nonlinear prediction method is more suitable for the chaotic backgrounds because of their larger difference between prediction errors,and the nonlinear model for prediction and setting threshold for detection can be used;while as for IPIX radar data,since there is less difference between their prediction results,likelihood ratio test detection can be chosen.