Adaptive Neuro-Controller for three axes attitude control of innovative satellite

Abstract

There exists so many disturbance torques in space which may deviate the satellite from the desired attitude. To overcome the effects of the disturbance torques some stabilization has to be provided to the satellite. This paper describes the development of a nano-satellite Attitude Control System (ACS) which uses Adaptive Neuro-Controller (ANC) based on Hybrid Multi Layered Perceptron (HMLP) network. The objective of this paper is to analyze the time response of ANC in order to improve the efficiency of the three-axes attitude stabilization. The nano-satellite plant that was used in this simulation is called Innovative Satellite (InnoSAT). The performance of ANC controller was compared with Adaptive Parametric Black Box (APBB). Both controllers used Model Reference Adaptive Control (MRAC) as a control scheme and Weighted Recursive Least Square (WRLS) as an adjustment algorithm. The function of this algorithm is to adjust the controller parameters to minimize the error between the plant’s output and the model reference’s output. The simulation results indicated that ANC controller has significant improvement over APBB controller for varying operating conditions such as varying gain, noise and disturbance torques.