Solución analítica de un filtro de Kalman estacionario para la observación de deriva en modelos de emisiones de NOx en motores diesel de automoción

Abstract

An augmented state-space model for drift correction is proposed adding an extra-state for cancelling drift on a given model or sensor output. A Kalman filter is used for drift observation. The model is Linear Time Invariant and noise covariances are considered constant. Under these assumptions, filter is steadystate and an analytical solution to the Riccati equation can be derived. Current paper gives the analytical solution to the Kalman gain and covariance matrix from using the iterative filter equations.