Multi-Objective Optimization Applied to Real-Time Command Problem of Spacecraft Thrusters

Abstract

A novel approach to solve the real-time command problem of spacecraft thrusters, called the thruster multi-objective command method, is proposed in this paper. The reaction control system technology uses a set of thrusters in a special setup to simultaneously provide force and torque to the spacecraft. The thruster management function calculates all the candidate solutions that solve the thruster coupling problem. Then, a discrete multi-objective optimization method selects at every control cycle the best combination of thrusters and their firing time duration, which simultaneously optimizes a group of four objectives: the force error, the torque error, the propellant mass consumption, and the total number of pulses. The proposed method is included in a coupled translational and attitude control system applied to the final approach rendezvous scenario. Furthermore, all elements of the guidance, navigation, and control loop are accurately designed and implemented in a simulation framework. Results indicate effectiveness, robustness, and a better performance when compared to the usual single-objective optimization case.