Robust control in non-vector space for X-ray polarimetry observation

The vision-based measurement and control method is an important key technology for X-ray polarimetry observation of space deployable mechanism. Compared with the traditional image based visual servoing (IBVS) method, the control method based on non-vector space has a unique advantage, because it can avoid extracting and tracking reliable features in harsh space environment. At present, the study of non-vector space control mainly focuses on the stabilization controller design; whereas, considering the serious imaging noise in space, a new robust control theory in non-vector space needs to be proposed. Towards to this goal, a set difference is first defined to describe the difference between two sets, which could be also used to represent the noise term in set space. Then, the robust problem in set space is discussed and reformulated, and a non-vector space robust control method is proposed to damp out the noise of pixel intensity. Finally, the comparisons of the simulation and experiment results between the existing stabilization controller and the designed robust controller are analyzed, which demonstrate the performance of the new robust controller.