Filtering of linear heredity systems with delays in the input
Year: 1981
Type: dissertation
Abstract: evolution equations on Hilbert spaces are studied in connection with stochastic linear systems described through linear delay differential equations.The abstract evolution equations are delay free.It is established that the abstract evolution equations provide an equivalent description of the linear system.This provides the bridge to solving certain filtering problems for the linear system by application of known 'infinite dimensional filtering' techniques to the evolution equations which are of standard form.The results reported here supply the first Kalman filter, to our knowledge, for systems having delays in the signal noise.For the case that no delays are present in the output, it is shown that the Kalman gains are generated by the unique solution of a set of partial differential equations.When observation delays are present, again we obtain partial differential equations for the Kalman gains but only under some smoothness assumptions.Examples are given, illustrating the relevance of our mathematical model to engineering systems.
Source: Spiral (Imperial College London)
Author: Monima Shenstone Briggs
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2547585324
