State Estimation And Dynamic Analysis Of Electrodynamic Tethered Deorbiting System

Last decades' researches reveal that electrodynamic tethered systems have huge potential for application in spacecraft deorbiting in the future. This dissertation deals with the nonlinear state estimation and de-orbiting control of electrodynamic tethered systems. The key works are as follows.Concerning that on-orbiting tethered system faces with real-time state estimation, the nonlinear dynamics model and the measurement equation of an electrodynamic tethered system are given first. Based on Gaussian harmonic method, the perturbation force and torque is obtained through the help of the precise geomagnetic field model. As a result, the de-orbiting dynamics model is modified by orbital elements. Based on Bayesian estimation theory, then, the real-time state estimation is respectively implemented by the following three Filters, i.e., an extended Kalman Filter, an unscented Kalman Filter, and an cubature Kalman Filter, with different abilities to achieve state estimation. For non-Gaussian noise in space, it is hard to get an ideal performance by traditional Filters, so an algorithm in conjunction with Gaussian mixture approximation and cubature Kalman Filter is developed to address the estimation over flicker noise, whose performance has been confirmed by comparison with the algorithms of other filers. In order to avoid the complexity of the algorithm mentioned above, particle Filter algorithm is taken into account here. The results suggest that the unscented particle Filter algorithm, compared with Kalman Filters, is able to achieve state estimation with simpler structure. Based nonlinear model prediction control with state estimation algorithm, finally, the de-orbiting process of electrodynamic tethered system is studied.