NN adaptive control of spacecraft attitude

Modern spacecraft demand an attitude control system with very high performance and accuracy, and many new features, such as self-modification, adaptability, more autonomy and fault tolerance. NN adaptive control can be applied to spacecraft attitude control to meet the demanding requirements and to provide many new features. To control spacecraft attitude motion, a recurrent NN is used to train as an identification model and an adaptive controller. The quaternion method can overcome the singularity problem in a traditional Euler differential equation. To accelerate the convergence speed of the NN training course, an extended Kalman filter method is used to train NN weights.