Moving Horizon Estimator for Space Vehicle Dynamics with Measurement Noise in Close Propinquity Operation

Abstract

Analyzing the impact of measurement noise on space vehicle dynamics is vital for understanding its implications in proximity missions. Despite researcher's focus on addressing position and velocity challenges during cooperative operations, few studies explore measurement noise effects in dynamic environments. This study introduces a methodology, primarily utilizing the Model Predictive Control-Based Moving Horizon Estimator (MPC-MHE) with Euclidean navigation constraints, to optimize the estimation of unmeasured state variables in space vehicle dynamics. Results show the efficacy of the moving-horizon estimator inaccurate estimation, outperforming LQ-MPC and IDVD in assessing unmeasured states with minimal control effort under measurement noise. The proposed technique is further compared with LQ-MPC/DH, LQ-MPC/RH, and NMPC in three scenarios, revealing superior performance in time to dock and control effort metrics. The method demonstrates an ARPI of 42.33 and 41.70 in Case A, 34.51 and 33.65 in Case B, and 32.90 and 41.25 in Case C for time to dock and control effort, respectively. Moreover, the moving horizon estimator exhibits a 95 coefficient of determination for unmeasured state estimates compared to ground truth data, highlighting its effectiveness in mitigating measurement noise challenges in space vehicle dynamics.

Keywords

• Model predictive control
• Moving horizon estimation
• Debris avoidance
• State estimation
• Autonomous rendezvous