Accurate and effective attitude determination and control is crucial for space missions such as earth observations, large-scale distributed sensing and target tracking [1–4]. Among the various control theories and methods, Proportional Integral Derivative (PID) control method has been widely used to the attitude stabilization and tracking missions of spacecraft due to its simple structure, low cost and easy implementation [5]. However, in light of the pretty complex space environment and system structure for the advanced satellites, some inherent limitations existed in the PID controllers are often encountered such as typical parameter tuning and performance degradation in the presence of actuator saturation and faults [6, 7]. Thus how to tune the controller parameter for improving the control performance of the conventional PID control method is a hotspot both in the theory and engineering. With the fast development of artificial intelligence technology, one promising approach is to apply fuzzy logic system or neural network (NN) to develop adaptive PID control schemes [8]. For example, a PID-type sliding mode manifold was constructed for the attitude system of reentry vehicle with its gain parameters being adjusted online via using fuzzy logic system and radial basis function NN in [9]. In order to achieve faster convergence time and higher performance, an adaptive fuzzy PID controller was developed for the attitude system of geostationary satellite in [10].