Mars entry navigation performance analysis using Monte Carlo techniques

Abstract

An atmospheric entry and descent full-state navigation filter is developed and presented. Using this filter a navigation performance analysis is performed to examine the effects of various instrument packages and differing sensor scenarios for the entry and descent phase of the Mars Aerial Regional-scale Environmental Survey (ARES) mission. Deterministic simulation in conjunction with Monte Carlo techniques is used to verify navigation performance of an extended Kalman filter. This analysis specifically compares the performance of four feasible instrument packages. examines navigation performance as it varies with changes to initial sensor activation altitude, and examines error sources and covariance trends for this entry and descent scenario. The results from the analysis show that large attitude uncertainty resulting from the LN200 IMU bias causes a breakdown of the filter algorithm dlue to nonlinearities. The addition of a surface relative velocity measurement, to the altimeter measurement provides only marginal position uncertainty improvement and significant velocity and attitude uncertainty improvement. Increasing the initial altitude for sensor activation provides slight improvements in position uncertainty. but large velocity and attitude uncertainty improvements. Finally, it is shown that initial state uncertainty dominates over all other error sources in this navigation analysis. Error growth within the principal states (position, velocity, and attitude) is predominantly a product of the near-constant attitude uncertainty as it transfers from the innocuous roll attitude channel into the more consequential pitch and yaw attitude channels.