Real-time entry guidance framework via an optimization-oriented quadratic programming formulation

Conventional entry guidance methods, including drag-profile-based approaches and predictor–corrector schemes, often require extensive parameter tuning and may fail to satisfy terminal conditions under extreme entry scenarios. This paper presents a real-time entry guidance framework formulated as a quadratic programming (QP) problem through an optimization-oriented reformulation. Considering the highly nonlinear dynamics and the constrained feasible solution space inherent in entry guidance problems, a cost function is formulated to incorporate both terminal constraints and input profiles. The path constraints are directly enforced as upper and lower bounds on radial distance, and terminal conditions are relaxed using slack variables. Smooth control profiles are achieved by penalizing the squared differences between consecutive control inputs. The proposed method is validated through extensive simulations, demonstrating high terminal accuracy and robustness under significant uncertainties in aerodynamic and atmospheric parameters. Comparative analyses indicate that the QP-based guidance outperforms conventional predictor–corrector methods, particularly when vehicle responses are limited by slow and constrained bank-angle dynamics. Overall, the proposed method provides a computationally efficient and reliable framework for real-time atmospheric entry guidance.