Autonomous Satellite Rendezvous via Hybrid Feedback Optimization

A research team including Oscar Jed R. Chuy, Matthew T. Hale, and colleagues has published a breakthrough paper on arXiv detailing a novel Hybrid Feedback Optimization (HFO) system for autonomous satellite rendezvous, proximity operations, and docking (ARPOD). The system addresses two critical challenges in space operations: measurement uncertainties that make future states difficult to predict, and the limited processing power of satellite onboard computers compared to terrestrial systems. The researchers developed an asymptotically stabilizing controller for the marginally stable Clohessy-Wiltshire equations that model satellite dynamics, then integrated discrete-time gradient descent optimization directly into the control loop.

The technical approach creates a hybrid system where optimization happens continuously in response to real-time measurements, rather than relying on pre-computed trajectories. The team proved their model is well-posed with complete, non-Zeno solutions that converge exponentially fast to a ball around the rendezvous point. Most impressively, simulations demonstrated up to 98.4% reduction in disturbance magnitude across various scenarios. This represents a significant advancement over traditional control methods, potentially enabling more reliable satellite servicing, debris removal, and formation flying without constant ground station intervention. The work has been submitted to Nonlinear Analysis: Hybrid Systems for 2026 publication.