Improved hp-adaptive Mesh Refinement Algorithm and Its Application

Aiming at rapid trajectory optimization problems of hypersonic re-entry vehicle,an improved hp-adaptive mesh refinement algorithm of rapid trajectory optimization was proposed.Multi-intervals optimal control problem was transformed into NLP by using an hp-LGR pseudospectral method.The accuracy of the solution was assessed by the residual of the discredited differential-algebraic constraints at sample points.For the inaccurate intervals,the curvature of the trajectory was used as the criterion of choosing pseudospectral method or mesh refinement method to improve accuracy.The calculation precision was satisfied by adjusting the number and the degree of the polynomial of subintervals iterativel.Simulation results show that the solution of higher precision can be obtained with less calculation cost by the proposed method,and it is applicable to calculating trajectory optimization under multi-constraints.