Parameter-adaptive estimation and control for a single-link flexible space manipulator carrying a large uncertain payload

End-effector control of flexible space manipulators is particularly difficult when dealing with large payloads of unknown mass. This is a relevant problem in active debris removal missions which involve grasping non-cooperative targets. This paper introduces a simultaneous parameter and state estimation algorithm that addresses the challenge of handling uncertainties from both the elastic deformation of the manipulator links as well as the unknown payload mass properties. The payload parameters are estimated using an indirect model reference adaptive control scheme, and the elastic coordinates are estimated using a Kalman filter, as well as a quasi-static closed-form estimation formula. Both the parameter and state estimates are used to improve controller performance - the estimated elastic coordinates are used to construct a modified control output called the $\mu$-tip rate which is used for task-space feedback control, while the estimated payload parameters are used to improve both the feedforward control as well as the model used in the state estimator. It is shown that the algorithm is able to converge to an accurate estimate of the payload mass and successfully stabilize the elastic deflections at the tip.