Spacecraft Rendezvous and Docking Using Adaptive Neuro-Fuzzy Inference System

Please enable javascript in your browser.

Tallebi, Morteza | 2017

494 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 51555 (45)
University: Sharif University of Technology
Department: Aerospace Engineering
Advisor(s): Asadian, Nima
Abstract:
In this research an ANFIS controller will be developed to rendezvous and dock with a non-cooperative spacecraft; the controller must make the spacecraft reach the velocity and nearby position of the target spacecraft. Due to uncertainty in modeling and changes in the moment inertia tensor, the controller must be able to adapt with these changes and uncertainty, these changes may happen because of robotic arm or solar panel movings. Robustness of the controller will be evaluated by applying orbital disturbances and modeling errors through monte-carlo simulations. To model the spacecraft the specifications of a known spacecraft will be used but some specifications may be studied parametricaly. Sensors will be modeled as ideal sensors. The control result will be computed as continuous force and moments being function of time. For this application, other controllers like: non-linear, fuzzy, optimal are used. ANFIS is a novel method to be used for this application. It is estimated that because of ANFIS being an adaptive method it will be good replacement for other methods for situations where uncertainties and changes may occur
Keywords:
Spacecraft Rendezvous and Docking ; Neuro-Fuzzy Controller ; Non-Cooperative Sattelites ; Monte Carlo Simulation ; Adaptive Neuro-Fuzzy Inference System (ANFIS)

No TOC