A hybrid discretization strategy for successive convex programming in skip entry trajectory optimization

Abstract

When applying successive convex programming to solve the skip entry trajectory optimization problem, it is challenging for a single discretization method to ensure both convergence accuracy and stability simultaneously. To address this issue, a hybrid discretization strategy is proposed. First, from the perspective of whether state propagation is utilized, discretization methods can be categorized into two types, and the properties of both are analyzed theoretically. Second, the strategy begins to iterate using discretization without state propagation. Once the designed switch condition is satisfied, the method transitions to discretization with state propagation to continue iterating until convergence. The precision and stability of the proposed strategy are verified through simulations involving various discretization combinations and different discrete switching criteria.