Typical mission maneuver planning is taken up with designing the nominal burn sequence. For example, in a normal Geosynchronous Transfer Orbit (GTO), one is concerned with the effect on the trajectory from each burn, assuming all the previous have performed nominally. Principal effort is devoted to evaluating whether the resulting trajectory meets visibility, radio frequency interference requirements, solar constraints and so forth. Customarily one assumes that prevailing orbit determination uncertainty, engine performance and attitude control uncertainties are sufficiently small that problems encountered can be handled as they arise in flight. This assumption is exacerbated by the underlying belief that coherently handling all the sources of uncertainty requires too much computing time, too many runs of too many alternate cases to make the effort worthwhile. This need not be the case. It is precisely because the prevailing uncertainties are reasonably small that differential techniques can be applied to address the ensemble of prevailing uncertainties in the same computation used to propagate the planned nominal trajectory. One can readily determine, for example, what range of post burn longitude drift rates to expect. Equally quickly one can decide how to adjust nominal targets to comfortably handle the expected range of orbit, attitude and burn uncertainty. This paper discusses the technique and applies it to trajectories seen in recent missions. It is shown how the result is a more robust burn plan at remarkably low cost in the development process.
展开▼