Time-Synchronized Estimator-Based ADP for Spacecraft Optimal Pose Tracking

This article addresses the optimal attitude-position integrated (pose) tracking for spacecraft proximity operations without exact knowledge of dynamics parameters. To tackle this challenge, this work proposed a time-synchronized estimator-based adaptive dynamic programming (ADP) to eliminate reliance on exact parameters knowledge in optimal pose tracking problems. Specifically, the concept of time-synchronized convergence is introduced into the estimator design to ensure that mass and inertia estimating errors converge synchronously within a finite time. This synchronization is crucial for online solving the optimal tracking problem. Subsequently, the estimator-based ADP is developed under the dual quaternion framework. This approach demonstrates that optimal pose tracking can be achieved through online learning, without dependence on the exact mass and inertia parameters. Finally, a series of typical simulations and experiments are illustrated to demonstrate the effectiveness and superiority of our technical findings.