Cost–Information Tradeoffs in Active Diagnosis of Rocket Airbrake Effectiveness Under Atmospheric Uncertainty

Abstract

Rocket airbrakes are commonly used to control apogee during coast, but a weaker-than-expected response does not always mean the airbrake has failed. Changes in drag, atmosphere, wind, sensing, timing, and actuator behavior can produce similar effects. In this study, we use an OpenRocket- based one-dimensional coast model to determine when reduced airbrake effectiveness can be separated from these uncertainties. A predefined set of diagnostic pulses is tested using measurements available during flight, and each pulse is evaluated by the information it provides and the modeled apogee it costs. The method was evaluated on a frozen 2025 ERA5 holdout and a separately defined 2026 ERA5 dataset. In 2026, the screened short pulse resolved 6/84 trials at a mean apogee cost of 0.684 m, while a longer fixed pulse resolved 12/84 at 2.047 m. An offline search of every feasible pulse resolved 15/84 trials. Of the 78 cases unresolved by the online method, 69 also remained unresolved offline, showing that the main limitation was the amount of diagnostic information available under the tested conditions.