Switching Neural Network Control for Underactuated Spacecraft Formation Reconfiguration in Elliptic Orbits

A switching neural network control scheme, consisting of the adaptive controller and sliding mode controller, is proposed for underactuated formation reconfiguration in elliptic orbits with loss either radial or in-track thrust. By using inherent coupling system states, technique then adopted to estimate unmatched disturbances design achieve high precision. The works active region, composed linearization errors external perturbations are approximated by basis function networks. outside upper bound approximation estimated adaptation law. stability closed-loop proved via Lyapunov-based approach. numerical simulation results have demonstrated rapid, high-precision robust performance compared linear controller.