GPS-based on-board orbit determination of a satellite using extended \(H_\infty\) filtering algorithms (Q1769095)

In this paper the extended \(H_\infty\) filtering algorithms for the design of the GPS-based on-board autonomous navigation system for a low earth orbit (LEO) satellite are introduced. The dynamic process models for the estimation of position, velocity and acceleration from the GPS measurements are established. The nominal orbit of the small LEO satellite is determined by using the 7th--8th order Runge-Kutta algorithms. Three filtering approaches are applied to smooth the orbit solutions, respectively, based upon the simulated GPS pseudo range observables using the Satellite Navigation Tool Box. The simulation shows that the observed orbit errors obtained by using the extended \(H_\infty\) filtering algorithms can be reduced to a lower level than the observed orbit errors in the sense of RMS within 12 h of tracking time by using the \(H_\infty\) filtering algorithms and the extended Kalman filtering algorithms under the appropriately designed parameters. Based upon the position errors predicted by the three filtering algorithms after the last observation, we find that the extended H infin filtering algorithm provides the least position errors of the user satellite.