In state space representation, with Kalman filtering, say I want to get the errors in states. I have errors in outputs. Can I get the errors in states?
Laya, I think you mean to describe noise in the state equation. Then, for the sake of optimality, you need to start with the process in continuous time, include the noise components associated with physical sources, and approximate it to be white Gaussian for the Kalman/Bucy filter. For the Kalman filter which is a discrete-time estimator, you need to convert the state linear differential equation to discrete time.