RESOLVING RADARSAT TORQUE ROD MAGNETIC DISTURBANCES PROBLEM

Y.V.Kim, G. Deraspe

Canadian Space Agency

6767 route de l'Aéroport

St-Hubert, Québec, Canada   J3Y 8Y9

ABSTRACT

 

After almost seven years in operation, the RADARSAT Attitude Control System (ACS) is still capable of providing very accurate attitude stabilization of the satellite with respect to its orbital frame. Successful operation of the satellite payload, Synthetic Aperture Radar (SAR), is demonstrated by the high-resolution images of high quality that is provided to the customers. The images are also proof that the stabilization in pointing mode is within the mission requirements tolerance and the errors do not exceed the specified level of 0.1^o (3s in each axis). This accuracy is achievable using the Primary Attitude Determination Method (ADM1), which uses the Sun Vector and the Local Geodetic Vertical Vector directions measured by Sun Sensor (SS) and Horizon Scanner (HS) to obtain three angles of the satellite attitude: Roll, Pitch and Yaw and the estimate of their derivatives. It provides an appropriate input to the control law and activates the actuators: Reaction Wheels and Magnet Torquers (MT).

However, after redundant HS2 failure (June 1996), using of ADM1 mode has been based on single failure devise: HS1, which since August 2000 has been indicating of potential motor failure signs.

In the case of HS1 failure RADARSAT ACS will automatically transit in Tertiary Attitude Determination Method (ADM3), which uses Sun Vector, provided by the SS and Earth Magnetic Field Induction Vector measured by Magnetometer (MAG) for 3 Euler Angles of RADARSAT attitude derivation. Despite this mode was designed as a redundant ADM and is less accurate as ADM1, it might provide a moderate accuracy level (within sigma = 0.3 deg) that would be sufficient for providing of SAR data processing possibility and RADRSAT mission continuation.

In order of preparation to potential HS1 failure and ADM3 functioning evaluation, special Flight Test of ADM3 in closed ACS control loop was performed in January 2001. It was discovered that some special effect takes place in ADM3. The effect is resulted in short period (few minutes duration) oscillation of ATT errors with maximum Pitch angle amplitude about 5 deg in equator crossing areas. After theoretical analysis the effect was explained by existence of a parasite feedback in ACS closed control loop created by Torque Rods Magnetic Residuals applied to MAG input in the periods between the Torque Rod Coils active current control. This effect was classified as a parametric resonance effect in the control loop because of periodical (with orbital period) change of the feedback gain. After a big number of simulations an appropriate adjustment
for the control loops bandwidths (internal for momentum control and external for attitude control) were found to dump the effect due to the control loop filtering capability futures improvement.
RADARSAT Attitude Control Processor (ACP) Control gains were changed correspondingly by new ACP Control Table uploading. New Flight Test performed after the table uploading have confirmed expected significant improvement of RADRSAT attitude stabilization being controlled by ADM3 mode.  It provides RADARSAT mission continuation in case of HS1 failure.

The method of Torque Rod Magnetic Residual Disturbances damping can be considered as general effective method for typical design of satellite ACSs.