Design of sliding mode controller based on radial basis function neural network for spacecraft autonomous proximity

Since the dynamic model of spacecraft has the characteristics of non-linear, kinematic couplings, uncertainties and nonstationary disturbance, it has become a challenging problem to accurately control the relative position and attitude of the spacecraft. A radial basis function neural network (RBFNN)-based sliding mode controller (SMC) is proposed for trajectory tracking of spacecraft autonomous proximity in this paper. Firstly, a six degree-of-freedom (DOF) relative motion dynamics model is developed for close proximity operations. The modified Rodrigues parameters are applied to solve the problem of singularity. Then, a SMC that does not require accurate model information is designed. RBFNN is used to adaptively eliminated the model uncertainty impacts on the system. Finally, the stability of the relative motion dynamics is proved via Lyapunov stability theory. Simulation results illustrate that the method can attenuate the attitude and position errors, reduce the chattering of the input and decrease the overshoot of the control torque effectively.