Using a fractional-order controller to improve the performance of a spacecraft attitude control in the presence of uncertainties and disturbances

Abstract

In this paper, an investigation into the control of a three-axis rigid satellite attitude control system utilizing a fractional order Proportional-Integral-Derivative (PID) controller is carried out, even in the presence of disturbances and uncertainties. The satellite’s attitude is controlled using a reaction wheel actuator with first-order dynamic model. Uncertainties are considered across various aspects, including the satellite’s moment of inertia, the actuator model, and the amplitude and frequency of disturbances. The external disturbances are modeled with both fixed and periodic components, and uncertainty is integrated into the disturbance model. For comparison, an integer order controller is also employed under the same conditions. This study aims to enhance the robustness and accuracy of satellite attitude control by optimizing fractional-order PID controllers and comparing their performance with integer-order controllers under uncertainty and disturbances. The objective is to evaluate their effectiveness in reducing pointing error, overshoot, and improving overall stability in dynamic space environments. The optimization problem is formulated with the mean absolute pointing error of the satellite during pointing maneuvers as the objective function. Controller gains, for integer and fractional order controllers, are determined using the Particle Swarm Optimization (PSO) method. Additionally, Euler’s moment equations and numerical solutions are utilized to implement three-axis coupled control of the satellite in the presence of disturbances. The performance index is thoroughly examined with regards to controller time response, as well as the standard deviation of uncertainties and disturbances. The findings from this research demonstrate that the fractional order controller exhibits superior accuracy and robustness compared to integer order controllers when confronted with uncertainties and disturbances. Based on the results, the fractional-order PD controller demonstrates, on average, a 69% reduction in overshoot, an 82% reduction in the sum of absolute error, 20% less settling time, 77.5% faster angular velocity, and 8.5% less control effort compared to the corresponding values achieved by the integer-order PD controller.