Optimal control of the attitude maneuvering for Razaksat class satellite based on rigid and flexible model

Abstract

The increase in demand for performance for satellite capabilities has pushed the design of the system to be more and more power consuming. This is the case for RazakSAT-2, which is a new satellite program that will be equipped with bigger solar panel to generate sufficient power. Thus, this translates to a higher flexibility in the satellite. Satellite mission is known to be highly sensitive to the flexible motions and it is time constrained. Hence, understanding the behavior of the system is required to solve the time constrain flexibility problem. The Floating Reference Frame is applied to obtain the mathematical model of the system which consists of three solar panels. In addition, the model for the actuator is also developed for a four-reaction-wheel system and the Eigen-axis Quaternion Feedback control is also derived. The obtained model is simulated using the MATLAB and ANSYS software for verification of the model. The obtained Percentage Root Mean Square Error falls between 2.015% to 4.841% which is low. Hence, this signifies that the model is sufficient to describe the dynamic of the system. From the model, the control of the minimum time optimal control is developed to minimize the time to achieve desired orientation while minimizing the amplitude of the flexible solar panel. GPOPS toolbox is applied to obtain the optimal control solution. The optimal control is shown to decrease the maneuver time by 3.49% to 25.11% depending on the Eigen Axis of the rotation compared to the conventional Eigen-axis Quaternion Feedback controller. This phenomenon is contributed by two factors. Firstly, the optimal control is able to fully utilize the all the capacity of the reaction wheel while the Eigen-axis Quaternion Feedback controller is plagued by the pseudo-inverse limitation which allows a maximum 35% increases in performance. Secondly, the application of optimal control allows the trajectory to deviate from the effective Eigen axis to achieve faster maneuver by utilizing the torque that is unavailable to the effective Eigen axis maneuver. In terms of the performance of the rigid and flexible model in the optimal control, it shown that the flexible motion converges at 10.53% faster for the flexible model. The primary factor that affects the maneuver time is the natural frequency of the system. The effect of the natural frequency is observed in this section and is shown that maneuver times increase when the natural frequency decreases. For future works, additional parameters such as the stiffening effect, external disturbances and the imbalance mass distribution on the rigid and flexible due to the deflection are studied. This can contribute to a more refined flexible model that would further increase the accuracy of the model.