A discussion was devoted to designing adaptive nonlinear dynamic inversion controllers. The equations of motion of the combined satellite/fuel system were derived, and the Rigid–liquid coupling system equations were constructed. Due to the strong coupling nonlinearity of spacecraft control system with fuel sloshing, neural networks were applied to create an adaptive NDI (Nonlinear Dynamic Inversion) controller. The results in the form of tracking performance for the adaptive NDI controller were shown using a model of SloshSat, which demonstrates that the proposed NDI controller with neural network is capable of excellent reference tracking in case of fuel shloshing.