The Optimal Mars Entry Guidance with External Disturbance Using Neural Network Solution

摘要

In this study, a novel H_∞ guidance strategy based on optimal control theory is proposed to guide the spacecraft landing on the Mars precisely. Initially, the error system of entry phase with the external disturbance is derived, and the method of tracking the nominal entry trajectory is adopted. Considering the difference of magnitudes' order between several state variables concluding the altitude, the velocity and range, normalization strategy is used to avoid numerical problem. Subsequently, the H_∞ optimal control theory is introduced, and the difficulty of the entry phase is that the Hamilton-Jacobi-Isaacs (HJI) equation with control constraint must be solved. So the neural network (NN) is introduced to get the solution of HJI equation approximately. Finally, simulation is done to demonstrate the H_∞ guidance strategy.