scholarly journals Adaptive pose and inertial parameters estimation of free-floating tumbling space objects using dual vector quaternions

2017 ◽  
Vol 9 (10) ◽  
pp. 168781401771421 ◽  
Author(s):  
Xianghao Hou ◽  
Chuan Ma ◽  
Zheng Wang ◽  
Jianping Yuan
Author(s):  
Jianping Yuan ◽  
Xianghao Hou ◽  
Chong Sun ◽  
Yu Cheng

Estimating the parameters of an unknown free-floating tumbling spacecraft is an essential task for the on-orbit servicing missions. This paper proposes a dual vector quaternion based fault-tolerant pose and inertial parameters estimation algorithm of an uncooperative space target using two formation flying small satellites. Firstly, by utilizing the dual vector quaternions to model the kinematics and dynamics of the system, not only the representation of the model is concise and compacted, but also the translational and rotational coupled effects are considered. By using this modeling technique along with the measurements from the on-board vision-based sensors, a dual vector quaternion based extended Kalman filter for each of the two small satellites is designed. Secondly, both of the estimations from each small satellite will be used as inputs of the fault-tolerant algorithm. This algorithm is based on the fault-tolerant federal extended Kalman filter strategy to overcome the estimation errors caused by the faulty measurements, the unknown space environment and the computing errors by setting the appropriate ratios of the two estimations from the first step dual vector quaternions extended Kalman filter. Together with the first and second steps, a novel fault-tolerant dual vector quaternions federal extended Kalman filter using two formation flying small satellites is proposed by this paper to estimate the pose and inertial parameters of a free-floating tumbling space target. By utilizing the estimation algorithm, a good prior knowledge of the unknown space target can be achieved. Finally, the proposed dual vector quaternion federal extended Kalman filter is validated by mathematical simulations to show its robust performances.


Author(s):  
Xianghao Hou ◽  
Jianping Yuan

Estimating the parameters of an uncooperative space target is essential to the on-orbit service missions. A good parameter estimation can provide sufficient prior knowledge for the further operations. This paper proposes a novel dual vector quaternions based adaptive extended two-step filter (DVQ-AETSF) to estimate the pose and inertial parameters of a free-floating tumbling space target. Firstly, both of the rotational and translational motions are modeled by the dual vector quaternions (DVQ). Then, by using the DVQ-based system model, the DVQ-AETSF is designed. The proposed DVQ-AETSF mainly consists of a traditional Kalman filter prediction procedure in the first step and an adaptive regularized Newton iteration technique in the second step. The new proposed two-step filter aims to deal with the high nonlinearities in the measurements equations. By using the proposed DVQ-AETSF, both of the pose and initial parameters of a free floating tumbling space target under large errors of initial guesses and high measurement noise can be well estimated. Finally, the proposed DVQ-AETSF is validated by mathematical simulations to show its performance.


Author(s):  
Xianghao Hou ◽  
Jianping Yuan

Based on the dual vector quaternions, this paper modeled both the kinematics and dynamics of the disabled satellites. In addition, considering the complex space environment may lead to the measurements failure of the target, a novel robust federal Kalman filter based algorithm (DVQ-REKF) is proposed by this paper. By utilizing the designed algorithm, both the pose and inertial parameters can be estimated in the same time when the measurements by each of the monocular camera on the service satellite failed for a certain period of time. Finally, the simulation shows the validity of the designed DVQ-REKF.


2012 ◽  
Vol 12 (04) ◽  
pp. 1250061 ◽  
Author(s):  
SILVIA FANTOZZI ◽  
PIETRO GAROFALO ◽  
ANDREA GIOVANNI CUTTI ◽  
RITA STAGNI

The aim of the present study was to identify the phases of gait and the joints where the "ground reaction vector technique" (GRVT) can represent an acceptable alternative to the use of inverse dynamics (ID), when considering subjects with a lower-limb amputation. First, an analytical investigation of the ID of the three joints of the lower limb is given, distinguishing the gravitational, the inertial and the ground reaction contributions. The first two contributions require inertial parameters estimation; for this purpose, literature anthropometric data are typically used, both for the unimpaired and prosthetic limb, as accurate specific inertial parameters for the prosthetic limb are difficult to obtain from companies or require time consuming estimation. This assumption potentially leads to errors in the three-dimensional (3D) joint moment estimation. Second, the results of two case studies, a trans-femoral amputee with two different prostheses and a trans-tibial amputee, showed that the GRVT can explain the most part of the net joint moment for the ankle and the knee in the whole stance phase, and for the hip in the first part of the stance, leading to a similar clinical evaluation without any assumptions on inertial parameters.


Sign in / Sign up

Export Citation Format

Share Document