Article ID: | iaor2002508 |
Country: | Spain |
Volume: | 5 |
Issue: | 1 |
Start Page Number: | 143 |
End Page Number: | 158 |
Publication Date: | Jan 1997 |
Journal: | TOP |
Authors: | Romera Rosario |
Partial information for stochastic control systems with state equations linear in the input are considered. The observation noise process is independent Gaussian and the case of ϵ-contamination is treated. A robustified version of the Kalman filter gives the update state in the contaminated observations case. The optimal control is obtained and for the quadratic cost a closed solution is given.