Online Trajectory Optimization For The Terminal Stage Of Reentry Hypersonic Vehicles

Hypersonic technology is an emerging field converging aerospace, aviation andother technical advantages. Providing real time and complete information about thefuture flight trajectory to attack the remote target rapidly and precisely is one of thegreatest features of hypersonic technology. The rapid optimization of trajectory is thekey issue to be resolved for its guidance system. The Common Aero Vehicle (CAV)model is researched in this dissertation, while the Model Predictive Control (MPC)algorithm, implemented by convex optimization solver CVXGEN, is used to solve thetrajectory optimization problem of the terminal stage of reentry.Firstly,3DOF reentry equations of motion are formulated by analyzing the forcesof the hypersonic vehicle. The nonlinear mathematical description with constraints forthe late paragraph of reentry vehicle is given, including the optimization index function,the state equation constraint, the terminal constraint, the state and control constraints.Secondly, the basic theory of convex optimization is introduced, including thedefinition and properties of convex function, the standard form of convex optimization.As an online interfacial software and a convex optimization solver, CVXGEN which isgood at solving quadratic convex optimization problems is researched too.Furthermore, the specific process of multi-constrained nonlinear being transformedinto convex optimization problem is given. The theoretical basis is presented to indicatethe feasibility of using linear, time-varying systems approximating nonlinear dynamicalsystems. Then, convex transformation of the trajectory optimization problem which isdescribed on the longitudinal and lateral planes is completed. Based on these, the ModelPredictive Control (MPC) algorithm is introduced, and the relevant mathematicaldescription is given in the end.Finally, the trajectory optimization for the terminal stage of reentry hypersonic issolved. Depended on the specific mission, the trajectory optimization problems areworked out from the longitudinal movement, lateral movement and synthetic movementrespectively. The optimal trajectory curve diagrams are given at last.It can be showed from the simulation results that the multi-constraints nonlineartrajectory optimization problems can be solved accurately through the MPC algorithmimplemented by CVXGEN.