Model-Based/Model Predictive Control Design for Free Floating Space Manipulator Systems
Evangelos Psomiadis, Evangelos Papadopoulos
Abstract
The rapid increase in satellites and space debris mandates advanced capabilities for on-orbit operations. The hostile-to-human environment and the required high accuracy and robustness of on-orbit operations render Space Manipulator Systems (SMS) the appropriate choice. This work proposes an easily applicable, computationally inexpensive, nonlinear, and robust Cartesian control law for spatial Free-Floating SMS (FFSMS). The controller consists of two fundamental parts. The first is a Model-Based (MB) controller, which linearizes the system and guarantees prescribed performance. The second is a Model Predictive Controller (MPC), which integrates the model and provides optimal performance with parametric uncertainty, noise, and disturbances compensation. Input and output constraints are integrated into the latter to improve its performance. Numerical simulations for a planar model using Matlab/Simulink and MSC Adams highlight the MB/MPC’s increased accuracy in comparison to a regular MB/PID controller, during a task that requires moving a captured object in the presence of parametric uncertainty, disturbances, and sensor noise. Monte-Carlo simulations substantiate the higher accuracy achieved by the MB/MPC.