A Neural Network-based Approach for Electrodynamic Control of Two-Satellite Formation Flying

Document Type : Research Paper

Authors
Department of Theoretical and Applied Mechanics, Faculty of Mathematics and Mechanics, Saint Petersburg State University, Saint Petersburg, Russia
10.22055/jacm.2026.49318.5888
Abstract
The development of small satellites drives the need for new propellantless motion control techniques. The utilization of environmental forces represents a possible solution, thereby motivating scientific research in this area. This paper examines satellite formation flying in the geomagnetic field. A neural network-based adaptive control method is utilized to address the problem of ensuring bounded relative motion of an electrically charged satellite by means of the Lorentz force. Based on the charge variation, a control algorithm is developed to compensate for launch system errors and achieve the formation keeping. Using the nonlinear equations of motion, numerical simulations confirm the operability and performance of the proposed algorithm.
Keywords
Subjects


Articles in Press, Accepted Manuscript
Available Online from 17 November 2026