Stochastic Interpretation of Robust Estimation: Risk Sensitivity
The game theoretic robust estimation problem defined in Chapter 3 is a deterministic one. That is, all the input disturbances have deterministic bounds. Nevertheless, as was mentioned in Chapters 2 and 3, this class of problems is closely related to a class of stochastic estimation problems, namely, risk sensitive optimal estimation , , . In risk sensitive estimation, the noise is assumed to be white, Gaussian, and the cost criterion is the expected value of the exponential of a weighted sum of the squared state estimation errors.
KeywordsProbability Density Function Kalman Filter Estimation Problem Robust Estimation Input Disturbance
Unable to display preview. Download preview PDF.