The Remote Sensing class of spacecraft requires high pointing control, stability and fastmaneuvers for the various remote sensing/science payloads. During different mission phases, severalAttitude Control System and fault detection algorithms onboard the spacecraft require accurate knowledgeof the spacecraft 3-by-3 inertia matrix. Before launch, the inertia matrix of the spacecraft is estimated byadding together the moments of inertia of the individual components of the spacecraft. The moments ofinertia of individual components are computed with respect to the predicted center of mass of the overallspacecraft before being summed. During the mission life the Spacecraft inertia changes due to fueldepletion. It necessitates estimating inertia accurately using on-orbit data. This paper details the ExtendedKalman filter based inertia matrix estimation algorithm formulation, design, analysis and simulation results.The spacecraft is in closed loop controlled mode and the desired sinusoidal three axis persistent torqueexcitation is provided by selecting appropriate reference rates. The three axes computed control torque(realized using reaction wheels) and gyro measured spacecraft rates are used as the inputs for the algorithm.The algorithm estimates the diagonal inertia elements to an accuracy of 1.5% of actual values. Extensivesimulation studies show that estimated inertia is the actual inertia plus the contribution due to the constantwheel torque scale factor. It is proposed not to compensate for the wheel torque scale factor in theestimation algorithm. Since the estimated inertia is used in the feed forward torque computation, the extratorque component gets nullified by wheel scale factor and the desired feed forward torque will be realizedon the Spacecraft. A case study was done for Cartosat-2 mission which is a high resolution earth imagingsatellite providing better than one meter panchromatic imaging capability using 2.5m camera. Considerablereduction in targeting errors was observed on up linking the estimated inertia by the proposed algorithm forfeed forward torque computation.
展开▼
