جستجوی مقالات مرتبط با کلیدواژه "فضاپیما" در نشریات گروه "میان رشته ای"

This paper compares spacecraft attitude control in the presence of disturbance torque using an adaptive backstepping controller and an extended state observer-based backstepping controller. Firstly, the adaptive backstepping controller is designed, in which the unknown parameters in a specific disturbance model are estimated using an adaptive law so that the closed-loop system is stable. Afterward, the backstepping control based on the extended state observer is designed. In this approach, the standard backstepping controller is initially designed, and then disturbances with a completely unknown model are estimated by the extended state observer, and the disturbance is rejected by applying the feed-forward law. The simulation results for two different disturbance models show that the backstepping controller based on the extended state observer demonstrates very good results compared to the adaptive backstepping controller when no disturbance information is available. This is for the reason that the extended state observer estimates the uncertainties more accurately.

Fault-tolerant control is one of the important issues in automatic control. The reason for this importance is the probability of fault/failure occurrence in control subsystems (sensor-actuator-system). Direct access to spacecraft is not always possible, therefore fault-tolerant control has become even more important in space systems. On the other hand, due to the necessity for weight reduction in these systems, employing hardware redundancy has limitations. Therefore, analytical redundancy has gained much attention in such systems. In this paper, reference inputs are corrected based an open-loop control command adjustment. Using simulation, it is shown that without reference input adjustment, the controller will not be able to satisfy mission requirements when actuator faults occur. Then, the proposed method is used and the desired requirements are satisfied. The advantage of the proposed method is that, there is no need for taking the first and second derivatives of the reference inputs and it is possible to obtain these inputs using integration. This will prevent computational problems associated with differentiation.