Nonlinear Optimization Positioning Method for Planetary Landing Navigation Using Shadow Constraints

Precise landing navigation on small celestial bodies poses significant challenges due to low-gravity environments, rugged terrains, and harsh illumination conditions. This paper proposes a tightly coupled multi-source navigation framework integrating inertial sensors and navigation cameras, enhanced by shadow constraints, to achieve high-precision planetary landing. A nonlinear sliding-window optimization strategy is developed, where landmark position errors are treated as additional state variables, and an adaptive residual weighting scheme is introduced to improve the robustness and convergence rate of the estimator. Furthermore, a novel tightly coupled shadow measurement model is formulated to exploit the geometric regularity of spacecraft shadows under parallel sunlight. Simulation results demonstrate that proposed method significantly enhances positioning accuracy compared to traditional approaches, achieving sub-meter level precision during the final descent phase.