A Neural Networks Controller for Mixed Formation of Satellites and Space Debris

In this paper, a mixed formation of satellites and space debris is proposed to solve debris cleaning problem. The relative dynamics model of formation near the L
₂
point of Earth/Moon-Sun is analyzed and the solar radius pressure is used as reconfiguration force. Then, a nonlinear controller with feed forward neural networks is designed which using a hybrid algorithm. The weights between input layer and hidden layer are adjusted by BP algorithm. The weights between hidden layer and output layer are adjusted by gradient descent method. The simulation results show that the method can help control followers achieve the target track (location of space debris), while the change of the solar sail attitude is kept under control.