OPTIMAL TUNING OF SINGLE-AXIS SATELLITE ATTITUDE CONTROL PARAMETERS USING GENETIC ALGORITHM

This paper studies the optimization of control parameters for single-axis attitude control of a rigid satellite using thrusters. It is desired to tune the control parameters to minimize the number of thruster firings. The motion task is defined as an attitude pointing maneuver from arbitrary initial condition to the rest. To this end, a control loop with pulse-width pulse-frequency modulator is suggested. The control actuators are pairs of identical non-ideal thrusters which each one produces torque in one direction around the control axis. A novel approach is proposed to model the dynamics of thruster with a design parameter which shapes the response of the actuator. Nine parameters of the control loop, e.g. feedback gains, modulator parameters, and thruster dynamics, are selected as decision variables. Genetic algorithm is used to find the optimal values of the variables such that the firing is minimized. It is shown that firing minimization requires a sluggish thruster. Moreover, to study the effect of deviation from optimal value on number of thruster firings and fuel consumption, a sensitivity analysis is performed. Based on sensitivity analysis, an optimal range is suggested for each parameter where both number of firing and fuel consumption are minimized.