Improving the performance of spacecraft attitude control with a reaction wheel actuator and considering sensor noise

Improving the performance of satellite attitude control with reaction wheel actuator and considering the sensor noise in this article, robust optimization method has been used to reduce the effect of sensor noise on the performance of the satellite attitude control with reaction wheel actuator. In this regard, the mean absolute pointing error has been chosen as the main performance index. The optimization based on the genetic algorithm and the Monte Carlo method of successive iterations have been used to include the effect of noise and obtain the control coefficients. First-order dynamics have been used to model the reaction wheel, taking into account the practical limit of the maximum production torque. Also, a modified proportional-integral-derivative (PI-D) controller with a observer method has been used to control the satellite. Gaussian white noise is added to angular velocity and angular feedback signals after passing through a first order filter. To compare the results, the control coefficients for the same simulation conditions have been obtained for two approaches: robust optimization and deterministic optimization. The performance in terms of the noise power spectral density function has been investigated for two approaches. The comparative results show that the control system adjusted by the robust optimization method, its performance criterion is more resistant in the face of noise and has less changes, while the performance criterion of the deterministic optimization method has more changes in Exposure to noise. The results indicate the preference of using the robust optimization approach for the control system exposed to noise.