Title :
Piecewise monotone filtering in discrete-time with small observation noise
Author :
Fleming, W.H. ; Ji, D. ; Salame, P. ; Zhang, Q.
Author_Institution :
Div. of Appl. Math., Brown Univ., Providence, RI, USA
fDate :
10/1/1991 12:00:00 AM
Abstract :
A discrete-time model for filtering with small observation noise is considered. A piecewise linear observation function is considered with two intervals of monotonicity. A sequential quadratic variation test is found to detect intervals of linearity of the observation function. Diffusion approximations to certain discrete processes are made to estimate the mean times for reaching a decision and the error probabilities
Keywords :
error statistics; filtering and prediction theory; piecewise-linear techniques; probability; diffusion approximation; discrete-time model; error probabilities; linearity; observation noise; piecewise monotone filtering; Error probability; Filtering; Filters; Linearity; Partial differential equations; Performance evaluation; Piecewise linear approximation; Piecewise linear techniques; Sequential analysis; Testing;
Journal_Title :
Automatic Control, IEEE Transactions on
