Agile Attitude Control and Singularity Avoidance/Escape by the SDRE Method Using a Biased State-Dependent Weighting Matrix

In recent years there has been an increasing need to improve satellite attitude control performance in terms of agility and attitude accuracy in large-angle attitude maneuvers. To achieve such control performance, single-gimbal control moment gyros (SGCMGs) should be mounted as modern-type actuators. Conventionally, based on the torque command calculated by the attitude control system of the satellite, SGCMGs were controlled by solving inverse kinematics through a pseudo inverse matrix steering law. However, in such a control system structure, it may be difficult to obtain the desired torque required by the attitude control system because of the singularity problem of SGCMGs. Furthermore, with respect to implementation, since the condition number of the Jacobian matrix of SGCMG becomes extremely large in the singularity, the numerical calculation error of the pseudo inverse matrix increases greatly. Therefore, we propose an overall control system that can solve above-described problems and the state-dependent Riccati equation (SDRE) control system that integrates the satellite and SGCMG system. The proposed optimal control system, which does not solve the pseudo inverse matrix, can realize gimbal angle guidance by gimbal angle feedback and singularity avoidance/escape using the biased weighting matrix. In the numerical simulation, the usefulness of proposed system is shown in comparison with the conventional system.