Paired-Outcome Learning for Selective Spacecraft Attitude Capture under Torque Limits

Key points

• Faster settings can reduce average tracking error while worsening individual maneuvers, making selective modification of an existing controller a baseline-relative learning problem.1
• A paired-outcome method combines three mechanisms: physically interpretable continuation policies with different speed envelopes, separate prediction of task failure and successful-pair tracking gains, and context-dependent rejection that executes the unchanged baseline.1
• The selective controller reduces aggregate integrated attitude error by 4.67% relative to the branch baseline, with 17 improvements, 31 exact ties, and no observed regressions.1
• Gains occur in principal-axis, near-axis, braking-boundary, and selected noise/drift maneuvers; general coupled cases retain the baseline.1

Sources (1)

• [1]Paired-Outcome Learning for Selective Spacecraft Attitude Capture under Torque Limits
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 17, 04:11 PM

  Torque-limited spacecraft attitude capture requires adapting the maneuver speed to uncertain braking authority.

  Faster settings can reduce average tracking error while worsening individual maneuvers, making selective modification of an existing controller a baseline-relative learning problem.

Extractive summary: sentences quoted from the sources.