Space Robot Sensor Noise Amelioration Using Trajectory Shaping.

Area of Science:

• Robotics
• Control Systems Engineering
• Aerospace Engineering

Background:

• Space robots require lightweight designs, leading to low natural frequencies that interfere with attitude control.
• Control-structural interactions are a significant challenge, exacerbated by low-quality, noisy sensors.
• Traditional step function trajectories are discontinuous and not always optimal for these systems.

Purpose of the Study:

• To explore alternative input trajectories for controlling flexible space robots.
• To investigate the efficacy of sinusoidal trajectories in mitigating control-structural interactions.
• To compare the performance of different trajectories under varying noise conditions.

Main Methods:

• Derivation of equations of motion for a flexible appendage, rigid body, and reaction wheel system.
• Development of a benchmark feedback controller for rigid body modes and additional filters for flexible modes.
• Autonomous generation of sinusoidal trajectories and implementation of feedforward whiplash compensation for comparison.

Main Results:

• The sinusoidal trajectory method demonstrated a 97.39% improvement over baseline step trajectories in the absence of random errors.
• Sinusoidal trajectories were found to be superior in minimizing control errors under ideal conditions.
• However, traditional step trajectories performed better than sinusoidal ones when sensor noise and random errors were present.

Conclusions:

• Sinusoidal trajectories offer a significant improvement for controlling flexible space robots in noise-free environments.
• The choice of input trajectory is critical and depends on the presence of sensor noise and external disturbances.
• Further research may explore adaptive or robust trajectory generation methods to handle real-world space conditions.