Spacecraft controller tuning using Particle Swarm Optimization

In this paper, a new evolutionary algorithm based optimal design approach for spacecraft attitude stabilization is presented. The plant (spacecraft attitude model) is modeled by a second order non-linear multi-input multi-output system. Euler´s equation of rotational dynamics and modified Rodrigues parameter (MRP) jointly defines the plant. The control law is synthesized in an output feedback structure based on a Lyapunov scheme. The virtues of this method extend from simplicity and inherited robustness to practical considerations like the use of high gain filters to estimate the rate of change of output. However, this method suggests a set of parameters that drastically influence the performance of the controlled system. The fact that there is no straightforward correlation between these parameters and the performance of the system may impede its utilization. In this work, we propose the use of particle swarm optimization (PSO) algorithm as a tool to infer the best parameters. The PSO optimization is performed in time domain using a simple cost function which minimizes the absolute value of the tracking error while varying the controller gains. Simulation results are included to show the effectiveness of the proposed approach.