Performance evaluation of nonlinearity and non-Gaussianity measures in state estimation
Jindřich Duník, Ondřej Straka, Ángel F. García‐Fernández · 2017
The paper deals with the state estimation of nonlinear stochastic dynamic systems. The stress is laid on the assessment of the estimate error, which is caused by the violation of the estimator design assumptions. The assessment is based on measures comparing estimators actual working conditions and the assumptions under which the estimators have been proposed. In particular, the measures of nonlinearity and non-Gaussianity are discussed. The measures are briefly introduced and selected typical representatives are detailed with respect to their implementation. Performance of the measures is evaluated in the framework of the Gaussian filters in a numerical study.