Optimization of Angular-Momentum Biases of Reaction Wheels

RBOT [RWA Bias Optimization Tool (wherein “RWA” signifies “Reaction Wheel Assembly”)] is a computer program designed for computing angular momentum biases for reaction wheels used for providing spacecraft pointing in various directions as required for scientific observations. RBOT is currently deployed to support the Cassini mission to prevent operation of reaction wheels at unsafely high speeds while minimizing time in undesirable low-speed range, where elasto-hydrodynamic lubrication films in bearings become ineffective, leading to premature bearing failure. The problem is formulated as a constrained optimization problem in which maximum wheel speed limit is a hard constraint and a cost functional that increases as speed decreases below a low-speed threshold.

The optimization problem is solved using a parametric search routine known as the Nelder-Mead simplex algorithm. To increase computational efficiency for extended operation involving large quantity of data, the algorithm is designed to (1) use large time increments during intervals when spacecraft attitudes or rates of rotation are nearly stationary, (2) use sinusoidal-approximation sampling to model repeated long periods of Earth-point rolling maneuvers to reduce computational loads, and (3) utilize an efficient equation to obtain wheel-rate profiles as functions of initial wheel biases based on conservation of angular momentum (in an inertial frame) using pre-computed terms.

This work was done by Clifford Lee and Allan Lee of Caltech for NASA’s Jet Propulsion Laboratory. NPO-42011

This Brief includes a Technical Support Package (TSP).

Optimization of Angular-Momentum Biases of Reaction Wheels (reference NPO-42011) is currently available for download from the TSP library.

Download it now!

 

This Brief includes a Technical Support Package (TSP).

Optimization of Angular-Momentum Biases of Reaction Wheels (reference NPO-42011) is currently available for download from the TSP library.

Login first to download.