The application of nonlinear filtering to fault detection in linear systems
M Davis - IEEE Transactions on Automatic Control, 1975 - ieeexplore.ieee.org
M Davis
IEEE Transactions on Automatic Control, 1975•ieeexplore.ieee.orgThe application of nonlinear filtering to fault detection in linear systems Page 1 TECHNICAL
NOTES AND CORRESPONDENCE 251 Ko(Z+ 1)= QHT[HQHT+ R]-’ (25) INTRODUCTION For
linear systems perturbed by Gaussian noise the Kalman filter (26) provides a robust means of
estimating the state from the observations. However, if the system is liable to faults then the
filter must be reparametrized when these occur or else large errors will be incurred. Thus, the
many respects. Specifically, we note that although the basic estimation the state and for …
NOTES AND CORRESPONDENCE 251 Ko(Z+ 1)= QHT[HQHT+ R]-’ (25) INTRODUCTION For
linear systems perturbed by Gaussian noise the Kalman filter (26) provides a robust means of
estimating the state from the observations. However, if the system is liable to faults then the
filter must be reparametrized when these occur or else large errors will be incurred. Thus, the
many respects. Specifically, we note that although the basic estimation the state and for …
Using nonlinear filtering theory, the equations satisfied by the optimal state estimator and fault detector for a linear system are derived. These suggest an implementable scheme based on reparametrization of the Kalman filter.
ieeexplore.ieee.org