ESA GNC Conference Papers Repository

Metop-A is the first European operational satellite for meteorology flying in a Low Earth Polar Orbit, and the first satellite operated by EUMETSAT in this type of orbit. A repeat orbit of 412 revolutions every 29 days has to be maintained within 5 km around the nominal ground-track and the local time of the descending node has to be kept within 2 minutes of 9:30. Moreover, frozen eccentricity conditions have to be ensured. Those conditions are foreseen in order to optimise the calibration and operations of the on-board instruments. Therefore, it is necessary to perform regularly (around every 4 months) very small in-plane manoeuvres (of a few centimetres per second) for ground-track control and, more rarely (around every 8 months), large out-of-plane manoeuvres (several metres per seconds) for local-time control. The Metop-A AOCS design uses a unique mode for the thrusting phases, which is optimised for relatively large manoeuvres (around 0.5 m/s) in order to ensure reasonable performances throughout the entire operational range. Also, the thruster mounting was driven by the need of minimizing the thruster plume impingement on the instruments carried on-board. However, these solutions have a non-negligible impact in the execution accuracy of the in-plane manoeuvres and thus on the capability to achieve precisely the desired orbital target. Two effects have to be considered: firstly, the large time separation between manoeuvres amplifies the impact of any over/under-performance in the ground-track evolution; secondly, unexpected radial and transverse parasitic forces affect the eccentricity achieved and local-time drift. The latter is directly observable in the ground-track evolution. A detailed modelling of the platform response during the execution of a manoeuvre is thus necessary to evaluate precisely the forces generated in the three axes, and to have them properly taken into account in the manoeuvre targeting. The selected thruster mounting results in two main effects: a large deviation of the main thrust from the along-track direction; and an significant cross-coupling between the torque control thrusters, causing large activations of the attitude control thrusters during the manoeuvre execution and noticeable variation in the off-modulation. Both these effects can be taken into account through a representative solid-body model of the propulsive system. The thruster pulses fired during the manoeuvre excite the solar array flexible modes. The extremely short duration of the manoeuvre means that the excitations will only be damped during the stabilisation phase which follows the manoeuvre, by means of torque control thruster actuations. In order to consider this effect accurately, a detailed characterisation of the platform behaviour against a fully representative satellite simulator was needed; different positions of the solar panel and different durations of the propulsion phase were analysed. This paper presents the activities performed by the flight control and flight dynamics teams, supported by satellite manufacturer experts, to characterise the satellite behaviour and to enhance the algorithm generating the manoeuvre telecommands. The in-flight experience acquired during the first year of manoeuvres proves that, due to these improvements, a much higher level of accuracy was possible for short manoeuvres, without requiring any modification either to the on-board software or to the controller gains.