Extended Kalman Filter Application For Launch Vehicle Trajectory

Abstract

Real time applications have been implemented out of intuitive modeling but due to inefficient estimation techniques there were deviations in the results from expected processes. To avoid these deviations for better performance modeling, many mathematical state estimation methods are in use. This paper focuses on a state estimation method for monitoring and control of a real time system. The existing models for state estimation, as well as the challenges associated with adopting such a method in practice, have been reviewed. This paper also deals with the estimation of state vector in nonlinear systems with the assumptions that the measurement errors, as well as the number of outliers that could occur within a given time window, are bounded. This paper presents estimation model for various stages of a satellite launch vehicle trajectory, for a lift off from an idealized spherical, airless, non-rotating earth. Tracking  RADAR measures slant range, azimuth and elevation of the vehicle and yields an output which may be corrupted by noise. Estimation of vehicle stages performance is done by Extended Kalman Filter, which focuses on control of a process from a priori information and updating it. A Monte Carlo computer simulation is used.

Keywords: Theory of Estimation, Nonlinear Models, Kalmanfilter, Extended Kalman Filter, Apriori Information, Launch Vehicle