Digital filtering methods, with applications to satellite attitude determination in the presence of modelling errors. Part I: theory

The report reviews a number of digital filtering methods and gives examples of their application. The methods include batch processing methods and rec\irsive methods, not only in the classical formulations but also in square-root formulations. The latter formulations have the purpose to prevent certain numerical inconveniences. Much attention is...

Accurate spacecraft attitude estimation with the ESA starmapper

The ESA starmapper is an optical-electronic instrument with an array of photo-sensitive slits, to "be strapped down to spinning spacecraft. Recorded times of passage of known stars over these slits, together with mathematical models for spacecraft dynamics and starmapper instrument, provide the basis for recursive estimation of the spacecraft...

Detection-filter representations for Markov jump diffusions

The problem considered is filtering for Gaussian observations of linear differential systems that are driven by both Wiener processes and marked Poisson point processes. Well-known representations of the MMSE-filter for such a Markov jump-diffusion are a differential for the evolution of its conditional density or differentials for all its...

Some aerospace applications of estimation theory

In this report, the application of estimation theory to three problems from aerospace research is described. The applications considered deal with spacecraft attitude estimation, software reliability modelling and estimation, and with a satellite-based navigation algorithm. Maximum likelihood estimation is used in the software reliability case...

A Two-Stage Bayesian optimisation for Automatic Tuning of an Unscented Kalman Filter for Vehicle Sideslip Angle Estimation

This paper presents a novel methodology to auto-tune an Unscented Kalman Filter (UKF). It involves using a Two-Stage Bayesian Optimisation (TSBO), based on a t-Student Process to optimise the process noise parameters of a UKF for vehicle sideslip angle estimation. Our method minimises performance metrics, given by the average sum of the states’...

Online Edge Flow Imputation on Networks

An online algorithm for missing data imputation for networks with signals defined on the edges is presented. Leveraging the prior knowledge intrinsic to real-world networks, we propose a bi-level optimization scheme that exploits the causal dependencies and the flow conservation, respectively via <italic>(i)</italic> a sparse line...