AVIONIC ARCHITECTURE FOR MODEL PREDICTIVE CONTROL APPLICATION IN MARS SAMPLERETURN RENDEZVOUS SCENARIO

摘要

Optimization-based control techniques such as ModelPredictive Control (MPC) are considered extremelyattractive for space rendezvous, proximity operationsand capture applications that require high level ofautonomy, optimal path planning and dynamic safetymargins. Such control techniques require highperformancecomputational needs for solving largeoptimization problems. The development andimplementation in a flight representative avionicarchitecture of a MPC based Guidance, Navigation andControl system has been investigated in the ESA R&Tstudy “On-line Reconfiguration Control System andAvionics Architecture” (ORCSAT) of the Auroraprogramme. The paper presents the baseline HW andSW avionic architectures, and verification test resultsobtained with a customised RASTA spacecraft avionicsdevelopment platform from Aeroflex Gaisler.