Real-Time Optimal Guidance and Control Strategies for Space Maneuver Vehicles

This chapter presents a real-time optimal guidance strategy as well as an integrated guidance and control algorithm for the space maneuver vehicle skip entry problem. To produce an optimal guidance command, model predictive control-based (MPC) techniques are applied. Since the MPC solves the online optimal control problems at each sampling instant, the computational cost associated with it can be high. In order to decrease the computational demand due to the optimization process, the two-nested gradient method proposed in Chap. 5 is used and embedded in the MPC scheme. Simulation results show that it can effectively improve the computational performance of the constructed MPC-based guidance schemes. As for the integrated guidance and control algorithm, the proposed computational framework employs a bi-level structure incorporating optimal trajectory design and deep neural network-based real-time control. In the upper level, a set of optimal flight trajectories with accumulated aerodynamic heating minimization is generated by sequentially applying a desensitized trajectory optimization algorithm. Subsequently, the generated trajectory ensemble is provided to the lower level, where several deep neural networks are constructed to learn the structure of the optimal state-control relations and to produce optimal control actions in real time. A detailed simulation study was carried out to verify the real-time applicability as well as the optimality of the proposed integrated design.