Aerodynamic-Integrated Trajectory Optimization for Reusable Launch Vehicles: A Successive Convexification Approach Transforming Nonlinear Problems into Linear Programming
- Paper ID
95892
- DOI
- author
- company
LandSpace Technology Corporation Ltd.
- country
China
- year
2025
- abstract
This paper presents a convex optimization-based trajectory planning method developed by LandSpace Technology Co., Ltd. for reusable launch vehicle(RLV) recovery, during the development of its ZhuQue-3 (ZQ-3) liquid oxygen-methane (LOX/LCH4) RLV model. To model the aerodynamic effects, aerodynamic forces are integrated into the governing equations. The proposed method reformulates the nonlinear optimization problem into a sequence of linear programming (LP) subproblems via successive convexification, ensuring solution feasibility through computationally efficient LP solvers. Key innovations include: (a) the systematic integration of aerodynamics, addressing a critical limitation in lossless convexification frameworks that neglect aerodynamic effects; (b) a novel reformulation strategy converting the original problem into LP subproblems instead of conventional second-order cone programming (SOCP) subproblems, providing algorithmic flexibility to accelerate computation through LP/SOCP selection; and (c) a unified multi-phase trajectory optimization framework guaranteeing end-to-end optimization across critical recovery phases. Detailed algorithm derivations are provided with performance rigorously validated via simulations, demonstrating the method's superiority over models that actively neglect aerodynamic effects. The paper further discusses real-time implementation considerations. This work fundamentally addresses aerodynamic integration challenges in trajectory planning, establishes simplified convex structures, and presents a systematically scalable framework for RLV systems.