A new interpretation of lattice structure convergence rates
Victor DeBrunner, K. Pondugula · 2003
The authors examine the lattice structure and its use as a model structure for the identification of signals and systems. The structure is examined in a state-space context, and the parameter information content developed earlier is explored in this useful parameterization. The behavior of a recursive prediction error method (RPEM) lattice identifier, as observed by J.J. Shynk (1987), is explained. The analysis indicates why lattice structures are generally a very good structure on which to base signal or system identification.>